WASHINGTON. D.C.  20460
                               January 27, 1981                  OFFICE OF
                                                          AIR. NOISE. AND RADIATION
 Dr.  Richard Procunier
 Noise Program Chief
 Environmental Protection Agency
 215  Fremont Street
 San  Francisco, CA 94105

Dear Dr. Procunier:

     I have enclosed for  your  information the final draft of the forthcoming EPA
Guidelines for Noise Impact Analysis.   Also enclosed is an accompanying  Docket
Analysis.  We are very grateful  for your viewpoints and comments submitted in
response to our request for comments in June 1978.  We apologize for the delay in
revising  the Guidelines, but we have attempted to thoroughly address each  issue
raised in the Docket.

   As you may recall, the purpose of these Guidelines is to provide decision  makers
in both  the public  and private sectors with  analytic  procedures which  can  be
uniformly used to express and  quantify impacts from noise, so that such impacts
can be readily understood and  fully considered within the comparative evaluations
which  constitute noise environment  decisions.   The  procedures contained  in this
document are applicable  to the preparation of  environmental noise assessments,
and  are  entirely consistent  with  analytic procedures in current  use  by EPA.  It
should be carefully noted  that adherence to the procedures within  the Guidelines is
strictly voluntary.   The Guidelines are neither mandatory nor  regulatory  in  intent.
We  believe that  the Guidelines offer a  valuable  aid  to decision  makers for
quantifying and evaluating noise environment changes.

   The Guidelines as enclosed are based on the deliberations of the Committee on
Hearing, Bioacoustics and Biomechanics (CHABA) Working Group  69, from  1972 to
1976,  in response  to  a request  by  the U.S.  Environmental  Protection Agency.
Recommended guidelines for preparing  environmental impact  statements on  noise
were published by the National Academy of Sciences in early 1977.  The procedures
in that document reflected a  compromise among practicality, economy,  desired
accuracy, and specificity.  Subsequently, a request for  comments on  the CHABA
recommended procedures was  solicited by EPA in the Federal Register in  June
1 978, to  which you raised questions or submitted remarks.

    We have since carried out  a detailed, step-by-step analysis of the issues raised
during the comment period in order to improve the overall accuracy, usage, and
general readability of the document. Conceptually, the enclosed final draft report
contains the same basic procedures delineated in the CHABA guidelines  published
in 1977.   The enclosed document  does incorporate,  however, some  notable
refinements in both the  general audible noise  and high energy  impulsive  noise
assessment  methodologies.   In addition, the  document now -contains  improved
example cases demonstrating some applications of the recommended procedures.

    In light of the changes and additions noted above, we extend to you a further
opportunity to comment on the final draft  version of the Guidelines.  Please feel
free to raise any additional  queries on the manner in  which certain  issues were
addressed, and to  indicate to us any  last  minute  refinements or considerations
which  you feel would contribute  to a  more  accurate  and usable  document.
Accordingly, we are requesting that if you wish to submit any further comments to
us on the Guidelines,  that we  receive these comments no later  than March 20,

   Please send your comments to:

      Jeffrey Goldstein
      Environmental Protection Agency
      Office of Noise Abatement and Control
      Washington, D.C. 20460 (ANR-471)
      Telephone:  (703)557-0592
   It is our intention to finalize the Guidelines in the next four to six months, and
make copies available  to the public through an appropriate notice in the Federal
Register at that time.

   We very much appreciated your prompt  and helpful assistance in reviewing and
commenting upon the  CHABA recommended procedures, and we look forward to
receiving any further thought you may have  in regard to the enclosed Guidelines.
                                 Yarles LTElkins
                               Deputy Assistant Administrator
                                 for Noise Control Programs


                                             December 1980
           DRAFT VERSION



Section                             Title                               Page

1.                  Introduction                                           1
 1.1                  Purpose of the guidelines                            1
 1.2                  Overview of the approach                             3
 1.3                  Structure of the guidelines                          7
   1.3.1                Preliminaries                                      8
   1.3.2                General audible noise                             12
   1.3.3                Special noises                                    13
   1.3.4                Vibration                                         14
   1.3.5                Potential changes in population                   14
   1.3.6                Examples                                          15
 1.4                  Other considerations                                16
   1.4.1                Projects which reduce noise                       17
   1.4.2                Temporary projects                                17
   1.4.3                Uncertainties in the analysis                     18

2.                  General audible noise                                 19
 2.1                  Basic screening procedures                          20
   2.1.1                Measures for the decription of general
                         audible noise                                    23
   2.1.2                Determining the yearly day-night sound  level      25
   2.1.3                Determining the population affected by  the
                         noise of the proposed project                    30
 2.2                  Health and welfare effects range                    33
   2.2.1                Human noise exposure criteria                     33
   2.2.2                Quantification of the noise impact                40
 2.3                  Severe health effects range                         52
   2.3.1                Human noise exposure criteria                     54
   2.3.2                Quantification of the impact                      61
 2.4                  Environmental degradation                           63
 2.5                  Treatment of temporary projects                     64
 2.6                  Example application                                 66

3.                  Special noises                                        71
 3.1                  High energy impulse noise                           71
   3.1.1                Description of high energy impulse sounds         72
   3.1.2                Human noise exposure effects of high energy
                         impulse noise                                    75
   3.1.3                Structural damage criteria for impulse  noise      80
 3.2                  Infrasound                                          85
   3.2.1                Description of infrasound                         85
   3.2.2          '      Human noise exposure effects of infrasound        85
 3.3                  Ultrasound                                          87
   3.3.1                Description of ultrasound                         87
   3.3.2                Human effects of ultrasound                       87
 3.4                  Noises with information content                     88

4.                  Vibration                                             89
 4.1                  Human effects of vibration                          89
   4.1.1                Description of building vibration                 90
   4.1.2                Human vibration exposure criteria                 90



         CONTENDS (Continued)

                Title                               Page

    Quantification of the impact                     94
  Structural effects of vibration                    97

Summary of noise impact analysis                     99
  Purpose and structure of the guidelines            99
  Analysis of impacts of general audible noise      100
  Analysis of impacts due to special noises         103
  Analysis of impacts due to vibration              104

Appendix A




Acoustical terms and symbols used in the guidelines,
  and some mathematical formulations for them       A-1

Procedures of other Federal agencies                B-1
1.  Environmental noise measures and their purposes
                          in Federal programs
                    2.  Estimating I^n from other noise measures
                    Summary of human effects of general audible noise   C-1


Measurement of and criteria for human vibration
Example application of guideline procedures for
  general audible noise

                               List of Tables

Table                            Title                                   Page

  1              Sumnary of methods for noise  impact analysis             6

  2              Yearly day-night sound level  as  estimated by
                   population density                                     28

  3              Values of the weighting function for general
                   adverse response                                       46

  4              Average hearing loss as function of 8-hour Leg          58

  5              Criterion function for severe health effects            59

  6              Sample data presentation:   future noise  levels
                   without proposed project                               67

  7              Sample data presentation:   future noise  levels of
                   project alone                                          68

  8              Sample data presentation:   future levels from all
                   sources combined                                       69

  9              Sample data presentation:   special situations           70

 10              Values of weighting function  for high energy
                   impulse noise                                          79

 11              Conversion of LQ^ to L^n via equal annoyance           81
 12              Basic threshold acceleration values for  acceptable
                   vibration environments                                 92

                              List of Figures

Figure                           Title                                  Page

  1              Preparation of a noise impact analysis                   5

  2              Flow chart and worksheet                                 9

  3              Screening diagram                                       21

  4              Summary of annoyance data from 12 surveys showing
                    close agreement                                      36

  5              Comparison of new annoyance function with previous
                   functions                                             39

  6              Suggested descriptors for various situations            50

  7              Comparison of curvilinear function and fractional
                   impact linear function                                53

  8              Potential hearing damage risk for daily exposure to
                   8-hour equivalent sound levels              .          60

  9              Recommended relationship for predicting community
                   response to high energy inpulsive sounds              77

 10              Infrasound criteria                                     86

 11              Weighting characteristic for building vibration
                   in terms of human responses for the frequency range
                   1 to 80 Hz                                            91

 12              Vibration criteria for residential areas                93

 13              Percentage of population complaining as a function
                   of peak acceleration                                  95

 14              Types of analyses suggested                            102

                                  Chapter 1

     It is the policy of the United States Government to consider the poten-
tial adverse impact en the environment of all proposed federal actions and
projects. Many states and local governments have similar policies.  The pur-
pose of such policies is not merely to provide a catalog of the adverse
environmental impacts of a project (which has already received tacit approval),
Bather, the purpose is to provide a description of the environmental con-
sequences of a possible project, so that an understanding of those conse-
quences can be an integral part of the decision on the project. In order
for this to occur, it is necessary for the environmental effects to be
expressed in a manner which can be readily understood by the decision-maker,
and by the general public whose participation in such decisions is usually
encouraged by all levels of government.

1.1  Purpose of the Guidelines
     One of the potential environmental consequences of many proposed actions
or projects is a change in the noise and vibration environment.  The "action"
may be the building of a new refinery, development of a new mine, construc-
tion of a road, use of a new piece of machinery, etc.  It may involve the
enlargement or the reduction in size of an existing facility, or an effort to
make a given facility more quiet.  It may be the promulgation and enforcement
of a new noise abatement regulation.  It may be the temporary noisy construc-
tion phase of an inherently quiet facility. Or, with no change in the noise
environment, the action may entail a change in land-use or population density

in a neighborhood.  Any proposed change that will significantly affect
either (a) the amount of noise generated or (b) the number of people exposed
to it, will result in noise-related environmental impacts.  These guidelines
contain procedures which can be used to describe and quantify those noise
related impacts.  These procedures are primarily intended for use during
initial planning stages of projects in order that the potential environ-
mental noise effects of proposed actions can be identified and considered
early in the decision process, and so that appropriate noise mitigation
measures can be conveniently implemented.
     The users of this document are expected to be federal agencies, state
and local governmental agencies, industries, environmental groups, and
individuals.  The procedures described here are applicable to the preparation
of environmental assessments and environmental impact statements, and to any
other situation in which a description of noise environment changes would be
useful.  Although individual agencies have their own specific procedures, in
most instances the approach described here is consistent with those procedures.
These guidelines are not intended to replace existing approaches, but to
complement and extend them, by showing how to proceed from a description of
noise levels to a quantitative description of the impacts of noise on people.
It is hoped that this document will assist in achieving nationwide consistency
in dealing with noise problems, and provide an objective and uniform evalua-
tion of the noise impacts.
     The approaches described in these guidelines are not mandatory, nor are
specific numbers which appear in the guidelines intended to be construed as
standards.  The guidelines are offered as an aid to the treatment of noise
impacts in the preparation of environmental assessments, reviews, and impact

statements.  Paraphrasing a statement by the Council on Environmental Quality,
these guidelines are intended to help public officials make decisions which are
based on an understanding of environmental consequences, and to take actions that
protect, restore, and enhance the environment  [1*, p. 25233].  The purpose of
these guidelines, then, is to present procedures which can be used to express
noise impacts in terms which are easily understood by decision-makers, so that
those impacts can be fully incorporated in the comparative evaluations which
constitute the decision.

1.2  Overview of the approach
     The guidelines are based on the philosophy that the technical approach,
the descriptors of the noise environment, the measurement and prediction
methods, the evaluation criteria, and the techniques for impact assessment
should be as simple as possible consistent with reasonable accuracy.  To the
extent that they are also uniform across different projects, public understand-
ing of noise impacts will be improved.
     It appears feasible to follow these principles to arrive at an objec-
tive, and for most situations, quantitative definition of the noise impact.
In many situations, it will be possible to calculate a single number which
expresses the total noise impact of a proposed project on the population
exposed.  When this single number index can be produced, the prospects are
enhanced for a more objective and rational comparison of noise with the other
impacts associated with specific projects.  Quantitative tradeoff studies are
made possible—for example between noise impacts and societal benefits.
In some cases, this level of quantification might seem unwarranted, or

^Numbers in square brackets refer to the reference list at the end of the
main text of this report.

      overly mechanistic.   For such cases,  the guidelines suggest a tabulation, in
      5 decibel (dB)  increments,  of the land area or number of people affected by
      adverse noise levels.  In addition, a traditional,  non-quantitative descrip-
      tion of the noise impact is encouraged, either as a supplement to these
      numerical descriptions, or, in unusual cases,  as the sole analysis of the
      noise impacts.
           The preparation of a noise impact analysis proceeds through several
      distinct steps to arrive at these descriptions of the noise impact, which are
      then used in the decision-making process (Figure 1).  The methods proposed for
      use in each of these steps  (Table 1)  are based, in  part, on the work and
      the progress achieved over  the last few years  by interagency committees, on
      the recommendations  of the  National Academy of Sciences-National Research
      Council, and on other scientific findings.
           For measurement of the noise environment, use  of the day-night sound
      level (L^n), officially adopted by several  Government agencies since publica-
      tion of the Environmental Protection  Agency's  "Levels Document" [2], is recom-
      mended as the primary measure of general audible noise.   Circumstances calling
      for the use of short-term measures of general  audible noise are also discussed.
      A modification of the day-night sound level for impulse noise is based on the
      report of a National Academy of Sciences, Committee on Hearing, Bioacoustics
. o A  and Biomechanics (CHABA) working group [3], and of  an interagency task force
      [4]  on this subject.  Measures to be  used for  infrasound, ultrasound, and
      vibration are also described in these guidelines.
           The quantification methods recommended for impact assessment in these
      guidelines are further developments of the  Fractional Impact Methodology
      used by EPA for assessing the health  and welfare effects of a noise

                 Description of Project or Action
look for Noise-Related Effects of Project or Action
                •..   —"   * "   ~   /
      Does  Noise rnVlrormienilCh'ahge?	
     Does Exposed Population Change?
     Are Changes Significant Enough for Detailed Documentation?
Measurement and Documentation of Noise/Exposed Population

     a.  Definition of Existing Noise/Exposed Population
     b.  Projection of Future Noise/Exposed Population
     c.  Change in Noise/Impact of Project
Assessment of Impact

     a.  Health and Welfare Effects
     b.  Potential Loss of Hearing
     c.  Environmental Degradation
Discussion and Analysis of Results
Decision on Proposed Project
         Figure 1.  Preparation of a Noise Impact Analysis

6.1 Hz TO 20 Hi \ SOUND
20 kHl to 100 kHz f PRESSURE
Ldn = 75dB
1 m/sec2 INSIDE
0.1'HzTO5Hz:120dB '
6 Hz TO 20 Hz: 120-30 LOG-E
105 dB
1 m/ioc2 FOR MOST
0.6 m/tec* FOR SENSITIVE
0.05 m/Mc2 FOR CERTAIN
0.0036 m/KC2. OR

environment.  For the general adverse response to noise  in the 55 to 75
dB (l£n) range, the function is based on data presented  by Schultz in
a recent review paper [5].  Similar impact assessment methods are proposed
in these guidelines for quantifying the following:  the  potential for loss
of hearing at 24-hour equivalent sound levels in excess  of 70 decibels;
the general adverse response to impulse noise; and the complaints caused
by vibration.  For general audible noise in rural and wilderness areas,
and for infrasound and ultrasound, no quantification methods are sug-
gested, but more qualitafeative descriptions are recommended instead.
     The measures and methods listed in Table 1 and described in this
report are simplifications, and the recommendation for their use is not
intended to discourage more rigorous approaches.  However, to provide a common
framework for comparison of different environmental noise assessments
(conducted by different persons in different parts of the country), it
is strongly recommended that the methods of these guidelines also be used
along with any other additional approach.

1.3  Structure of the guidelines
     Three principal types of noise and vibration environments are considered:
general audible noise; special noises; and vibration.  General audible noise
is noise as commonly encountered in our everyday living  environment.  It can
be adequately described by either the equivalent A-weighted sound level
(Leq) or its variation that includes a nighttime weighting, the day-night
sound level (I^n) • for most practical cases this type of noise measure will
adequately describe the noise environment, and much of the document concerns
the evaluation of general audible noise.  Not all noises can be adequately
evaluated by average sound levels, however.  Examples of such special noises

are infrasound (frequency range of 0.1 to 20 Hz), ultrasound (frequency range
above 20 kHz), certain types of impulse noises (such as blasts and sonic
booms), and sounds that convey more information than random noise sources
with comparable average sound levels (such as voices, warning signals, or
barking dogs).  Procedures are also included for evaluating the impact of
vibration on man.  While the main reason for their inclusion here is to
account for vibration generated by airborne noise, the impact of certain
types of vibration can be assessed whether the transmission paths are air-
borne or structureborne.
     There is a separate chapter for each of the three principal types of
environment.  Each chapter covers four topics:  the appropriate physical
measurement for that type of noise; methods for determining the existing
levels and for predicting the levels for the proposed project; human noise
exposure criteria; and procedures for quantifying the impact, usually in
terms of those criteria.  All of the information necessary to deal with
one type of noise environment is thus in one place, to minimize the effort
required by a user to follow these guidelines.

1.3.1  Preliminaries
     The logic of the structure of these guidelines has been set out in a
combined flow chart and worksheet (Figure 2), to provide guidance for using
this report and for carrying out the various parts of the noise impact
analysis.  There are four principal branches in the flow chart (labeled A, B,
C and D) to be followed, depending on the nature of the proposed project and
its potential impact.  There are exit points along each of the branches, at

                 DOES THE
                 PROJECT I
                 CHANGES IN
which the analysis for that branch may stop without the need for any further
analysis, since it is clear by then that there is no significant noise impact
with respect to the concern on that branch.  At the right-hand edge of the
flow chart there are three columns that can be checked to indicate the
outcome of the analysis at each branch point.  When this flow chart is used
as a worksheet, these columns summarize the noise impact analysis for the
project, showing the stages at which exit points occurred, and calling
attention to aspects of the noise impact receiving explicit evaluation
according to the methods of Chapters 2, 3, and 4.
     Usually, there will be not just one version of the proposed project, but
a number of alternative proposals.  Each of these alternatives must be analyzed
for noise impacts.  There will thus be a flow chart worked through for each
of the alternative schemes.  Each of the worksheets, by its summary columns
of checked boxes, will indicate what aspects of the noise impact of that
alternative received explicit consideration. A comparison of these columns
will facilitate choosing the project alternative with the least noise impact
on the environment.  Flow Chart
     The following discussion of the use of the flow chart provides a brief
explanation of each of the branches.  (The section of this report contain-
ing the more detailed discussion is indicated in parentheses on the flow
     The first step is to provide a general description of the proposed pro-
ject, including those aspects that are expected to contribute to noise
impact. The expected noise impact may be either adverse, if the noise environ-
ment would be worsened by the project, or beneficial, if the environment


would be improved.  Both the short term and long term effects expected from
the project should be described.  For example, the construction of a new
airport or highway in a sparsely settled region would have as its initial
impact an increase in noise that would affect relatively few people.  However,
the new facility will attract new people and business which will increase the
nearby population density, unless proper land-use planning and implementation
occur. Thus the ultimate noise impact may be significantly greater than that
projected on the basis of the initial effect alone.  To evaluate an action
over time, it is suggested that a time interval of 20 years be used, unless
the project will be in existence less than 20 years, in which case the
project lifespan should be used.  Thus, the initial impact and the expected
impact after 20 years should both be evaluated.  To present a complete
picture, the impact after 5, 10, and 15 years might also be presented.  When
comparing the impact between projects or alternatives or when assessing
cost-effectiveness, the average impact over a 20 year period may be used.
     The first branch point in the flow chart occurs after the potential
noise impacts of the proposed project have been described.  At branch
points such as this, each of the available branches (labeled A, B, C and D)
should be taken, and followed through to the appropriate indicators in the
righthand columns.  At each of the question points following each branch, if
the project will entail no change at all, the 'NO* answer will be followed to
'EXIT*, and the analysis for that branch is complete.  In that case, check
the box at the right-hand side of the page under  'No environmental change1.
If exit points have been found for each of the branches A, B, C, and D, the
noise analysis need not proceed further:  four check marks will be in the
column labeled "No environmental change" at the right of the page and the

noise analysis is finished.  The environmental impact assessment on noise

will simply state this fact.  Otherwise, the analysis continues in those

branches in which no exit point has been found.

1.3.2  General audible noise

     If on the first branch the project was found to include a potential

change in general audible noise, chapter 2 is appropriate, describing how

to identify and quantify the noise impacts for such an environment.  The

analysis begins with the screening step, to determine if the potential

change is large enough to pursue in a detailed analysis (section 2.1).

If it is not, the 'NO1 response is again followed to 'EXIT* and the analysis

for this branch is complete.*

     If the potential change is large enough to warrant further analysis,

the next question is whether people will be exposed to the noise from

the project.  If the answer is 'YES1, there are three branches to follow,

depending on the level of noise resulting from the project.  The first

branch deals with the range bounded roughly by L$n values of 55 dB and

75 dB, in which general health and welfare criteria are the impacts of

interest (section 2.2).  The second branch concerns projects which include

noise levels above 75 dB (I^n).  Where this occurs, severe health effects

due to noise should be considered (section 2.3).  The third branch is for

projects which result in levels less than 55 dB (I^n).  Although these
*The flowchart and worksheet is designed primarily for those cases where
the noise (or vibration) impact is expected to be adverse, that  is, the
noise environment is anticipated to be worsened by the project.   If the
project entails a reduction in noise, thus improving the environment,  the
flowchart and worksheet can still be used as a guide to carrying out any
noise assessment that may be desired to ascertain the degree of  improve-
ment.  See the discussion in section 1.4.1.

are levels below which adverse health or welfare effects generally do not
occur, environmental degradation is of concern, and should be discussed
(section 2.4).  For each of these three ranges of general audible noise, the
section identified provides a discussion of the human exposure criteria, and
methods for quantifying the impact in terms of these criteria. It should be
pointed out that at this branching point, at least one of the three sound
level ranges must lead to the requirement for a noise impact analysis, and
possibly more than one branch will do so.  That is, these categories are
exhaustive (i.e., they cover all the possibilities), but they are not sin-
gularly exclusive (i.e., more than one can occur).
     Even if people do not normally live or work in the area exposed to the
new noise levels, environmental degradation is still of concern.  As such,
this is difficult to quantify, but it should be discussed, preferably in
terms of the principal uses made of the affected area (e.g., urban recreation,
wilderness recreation, wildlife).

1.3.3  Special noises
     If the project is found to involve special noises, namely impulse
noise, infrasound, ultrasound, or sounds with negative information content,
it is necessary to follow branch B further (chapter 3).  The screening step
verifies that the levels involved are high enough to warrant an analysis.*
These levels are discussed separately for each type of special noise:
impulse noise in section 3.1.1; infrasound in 3.2.1; ultrasound in 3.3.1;
and noises with information content in section 3.4.  If people are exposed
to one or more of these special noises, the second part of the appropriate
*See footnote on page  12.

section of chapter 3 is available, describing procedures to be used to
discuss and quantify the impacts.  For impulse noise, there is also the
possibility of structural damage occurring (section 3.1.3).
     This section of the flow chart is obviously somewhat simplified.  If  it
were drawn in full detail, there would be a branch such as this for each one
of the four special noises.  Thus one should repeat this branch four times.
The reason for the simplification is that in most (but not all) cases encoun-
tered, no more than one type of special noise will generally be involved.

1.3.4  Vibration
     Branch C is followed if the project involves vibration (chapter 4).
Again, there is a screening step, to allow an 'EXIT1 to 'NO ENVIRONMENTAL
CHANGE* if the levels are low enough (section 4.1.2).*  If the levels are
higher than the cut-off, there are two branches to be pursued, the first
if people are exposed (section 4.1.3), and the second if buildings or monu-
ments are exposed (section 4.2).  Thus for vibration there are both human
and structural criteria to be considered in assessing potential impact.

1.3.5  Potential changes in population
     Some projects entail exposing new populations to existing noise levels,
for example the construction of a housing development in an area adjacent  to
a major roadway.  Branch D of Figure 2 describes the procedures to be fol-
lowed in such a situation.  If the noise levels from existing sources are
presently below 55 dB (L^), and are expected to remain this low in the
future, then there is 'NO IMPACT', and no further analysis is required on
this branch (as long as there are also no special noises encountered).  It
*See footnote on page 12.

should be noted, however, that higher density development (whether residen-
tial, industrial, or commercial) usually brings with it increasing noise lev-
els, such that it is unlikely that sound levels after project completion will
be as low as they are at present.  This 'EXIT1 is unlikely to be realistic for
any major development.  If the noise levels are not below 55 dB (I^n), or
if special noises are present, the analysis follows the same steps as did
branch A, when new noises affected existing populations (sections 2.2, 2.3,
and chapter 3).

1.3.6  Examples
     The following examples are presented to illustrate which of the major
branch(es) of the flow chart to use for various projects:
(1)  A project that entails a change in land use may cause only a change
     in the existing noise in an area, so only Branch A or B would be fol-
     lowed; on the other hand, it may only involve relocation of some
     of the population, in which case only Branch D would be followed.
     If the project is expected to cause (or diminish) vibration, Branch
     C would be followed.  Most land use changes, however, will involve
     a combination of A, B, C, and D.
(2)  A project involving the installation of new equipment, or the replace-
     ment of old equipment, is likely to require analysis of only branches
     A, B, and/or C, since no population shift is likely to be involved.
(3)  A project that consists of a new regulation, or a change in an exist-
     ing regulation, might follow either A, B, or D (see discussion in
     section 1.4.1).  For example, a new regulation reducing the noise output
     of heavy trucks would change the noise along a highway, and thus Branch

     A should be followed.  On the other hand, a change in the noise policy
     of the Department of Housing and Urban Development may alter the distri-
     bution of future dwellings among neighborhoods with different levels of
     existing noise; such a regulation would change the population exposed to
     noise without affecting the noise anywhere, and hence would warrant
     analysis along Branch D.
(4)  A new airport, whose primary effect would be increased noise levels in
     the neighborhood (Branch A), might impact only presently undeveloped land
     that could be spoiled for later residential development by the airport
     noise.  The future magnitude of the impact would be quite different for
     a prospective airport where land is purchased around the proposed site
     for controlled leasing to non-noise-sensitive activities as compared to
     one where such precaution was not taken.
(5)  A project that causes a change in the interior noise of aircraft cabins
     or a change in the noise insulation of automobile bodies would be ana-
     lyzed on a path along Branch Af since it changes the noise environment
     in existing spaces with a definable existing population.

1.4  Other considerations
     The preceding summary of the structure of these guidelines, and of
the flow chart representing that structure, is written to deal with a proposed
project which will increase noise levels, or in which more people are exposed
to existing noise levels.  These are not the only types of projects for
which noise impacts should be considered.  This section describes an additional
situation in which noise impacts are a concern—projects aimed at reducing
noise levels.  The section also discusses two topics which are relevant to

all three of the chapters which follow:  shortened analysis procedures for
temporary projects; and the treatment of uncertainties encountered in the

1.4.1  Projects which reduce noise
     Two of the examples of proposed actions with noise-related impacts
described en page 1 deal with the reduction of noise.  If an action is pro-
posed in order to reduce noise-related impacts, it is immediately obvious
that an analysis of those impacts is called for.  In most cases it is
equally obvious what kinds of impacts are of interest (e.g., general health
and welfare impacts).  Hence the flowchart and worksheet are not really
needed as an aid to identifying the area of concern.
     Instead, one may simply turn to the appropriate sections of these guide-
lines for a discussion of useful procedures for quantifying noise impacts.
In other words, the procedures (described subsequently) apply equally well to
projects which reduce noise as they do to projects which increase noise.  The
summary of those procedures (described previously) is written only in terms
of projects which produce noise.

1.4.2  Temporary projects
     At this stage of a noise impact analysis, the specific types of noise
impact requiring detailed documentation will have been identified.  This
will have been accomplished either through the use of the worksheet (Figure
2), or by the fact that the action is intended to reduce noise impacts.
The next question to address is how far into the future the analysis should
go.  Earlier, it was suggested that either the project duration or twenty
years, whichever is less, should be considered (section 1.3.1).  This means

that not only noise levels, but also affected populations, need to be pre-
dicted over the time period of interest.  The impacts to consider are not
merely the immediate ones, but long-term ones as well.
     Documentation of the impact for temporary projects is simplified by
the fact that population prediction is unnecessary; existing population
or land use information is sufficient.  In this context, temporary is
taken to mean less than roughly two years duration.  Beyond that length
of time, significant population changes may take place in an area, so that
population forecasts become important.  This would be true even for longer
duration construction projects, which might be in place for 5-10 years.

1.4.3  Uncertainties in the analysis
     There will almost always be areas of uncertainty in the noise impact
analysis, usually because of the unavailability of needed factual information.
For example, the projected future traffic volume for a proposed freeway may
be uncertain; the noise of a not-yet-built device may be only approximately
known; or the population estimated to be exposed to various sound levels from
the project may be subject to error.  In all cases, a discussion of the
probable source and degree of these uncertainties should be included in the
analysis.  Perhaps the most suitable approach for this purpose is to take the
upper and lower bound for each of the uncertain quantities that enter into
the analysis, and group the "most favorable" and "least favorable" bounds of
these quantities together to arrive at two estimates of the environmental
noise impact:  the best and worst cases- that together bracket the range of
likely actual results of proceeding with the proposed project.

                                  Chapter 2
                            General Audible Noise

     This chapter describes procedures to be followed for analyzing and docu-
menting the noise-related impacts of proposed projects which affect or are
affected by general audible noise.  The first section outlines a  screening
procedure for determining whether the expected noise levels of the proposed
project are high enough to warrant detailed analysis.  As part of that discus-
sion, appropriate noise measures are identified, and methods for  estimating
and predicting them are indicated.  The second section discusses  the general
health and welfare effects of noise on people, which serve as the criteria
for evaluation of noise in most urban and- suburban settings.  It  also pro-
vides a procedure which can be used to summarize these effects with a single
number impact descriptor.  The third section discusses the severe health
effects which can be caused by higher levels of noise, and suggests a single
number impact descriptor for these.  The fourth section discusses procedures
to follow for projects which will have reasonably low noise levels, but
which are located in very quiet areas—that is, projects for which environ-
mental degradation is the primary concern.  Simplifications of these analysis
procedures which can be used for temporary projects are described in the  fifth
section.  The final section contains an example application of the procedures,
including samples of the types of tabulation which are recommended.
     The criteria used in this chapter are not to be considered all-inclusive;
additional information should be used depending on the scope and  magnitude of
the environmental change.  The EPA Criteria and Levels documents  [6,2] can be
consulted as additional reference sources as well as any other applicable

2.1  Basic screening procedures

     Some proposed projects will obviously cause a severe noise impact on

their surroundings, others may obviously be so quiet as not to change the

noise environment at all.  In the first case there is no doubt that a full

analysis of the noise impact is required; in the second case one would simply

state, with minimal documentation, that no impact is expected.  About many

projects, however, there will be a question as to whether their noise impact

is significant enough to warrant a full noise impact analysis.  This section

offers a screening test to determine how extensive a noise analysis is needed.

     Figure 3 presents a screening diagram for use in determining whether a

full impact analysis is needed. The diagram is based on a comparison of the

existing noise environment and the noise environment due to the proposed

project alone.*  This comparison should normally take place at the noise

sensitive location(s) in closest proximity to the proposed project.**  So

long as the expected yearly L^ (see section 2.1.1 for explanation of

from the project is lower than 10 dB below the existing yearly L^n, the
* The meaning of "project alone" is clear when an entirely new facility is
  to be built.  But what about the expansion of an existing facility?  In
  such cases, "project alone" should be considered to be the total expanded
  facility or project.  For example, if the project is the widening of an
  existing highway from two to six lanes, future noise levels from the
  "project alone" would be the noise from the six-lane highway, not just the
  noise from the additional four lanes.  That is, the "project alone" is
  the new six lane highway.
**There are some exceptions to this rule.  If it is known that the
  greatest noise impact will occur at a noise sensitive location farther
  away, the comparison should take place at that point.  An example would
  be a close-in area protected from the noise by natural terrain, relative
  to an unprotected point farther away.  If the latter location receives
  the greatest impact, the comparison should take place at that point.










ANAL1. -IS /
' /
^ \s \ \
40 50 60


1 . 1
70 80

\ ^
                                                EXISTING Ldn(y)

                                      FIGURE 3. SCREENING DIAGRAM

project is screened out.*  No further analysis is required, because the

change in the environment is not significant.   The rationale for the location

of the screening line in Figure 3 is that for  any project which is not

screened out, the total environmental noise level after project completion

will be increased from the existing levels. For example, if a project alone

is expected to produce the same level of noise as the existing yearly L, ,

then post-project total environmental noise will be increased by 3 dB.  If the

difference obtained by subtracting the project alone noise levels from the

existing levels is greater than 10 dB, the post-project total noise level

will increase by less than 0.5 dB, which rounds off to a zero increase.

     For any new project in which the expected I^n is greater than 55 dB,

the probability of significant noise reductions of some of the existing

sources should be considered.  If the existing levels are high because of'one

major source which is likely to be quieted in  the future, then the proposed

project should not be screened out; further analysis is needed since, in

the future, the proposed project could become  relatively more dominant than

expected on the basis of existing noise levels.  If the existing levels are

unlikely to be reduced, then the project can be screened out.  However, even

if no reductions are likely an impact analysis can still be carried out, and

in many instances is strongly encouraged, based on idealized or hoped for

future noise levels for the area.  In other words, noise impact analysis is

reconmended for noisy projects even if they are in already noisy areas.
*If the project is temporary with a duration of less than one year,
 expected yearly 1^ should not be used.  Rather,  it is more appropriate
 to use the day-night sound level averaged over the actual duration of the
 project (see Section 2.5).  In any case, existing yearly L^ is always

     The diagonal line in Figure 3 continues as the indicator of screened-out
projects to very low noise levels.  For proposed projects which are above
this line, but produce levels below 1^ = 45 dB, a modified form of noise
impact documentation is suggested.  The modification is needed in part
because of the lack of data on the effects of very low level noise.  Such low
levels are in fact extremely rare, if indeed they ever occur.  Even on the
north run of the Grand Canyon 1^ was found to be close to 44 dB, due to
bird, animal, and insect noises [7].  Hence it is not expected that situations
in the lower left portion of the diagram will be encountered often, and the
diagram has been truncated accordingly.

2.1.1  Measures for the description of general audible noise
     The primary measure for describing general audible noise is the day-night
sound level, symbolized as L, .  The unit for L,  is the decibel.  The
day-night sound level is a 24-hour equivalent sound level in which nighttime
noise levels occurring between 10:00 p.m. and 7:00 a.m are increased by 10 dB
before calculation of the 24-hour average.  Equivalent sound level is numerically
equal to the value of a steady sound level that would convey the same mean-square
A-weighted sound pressure level as does the actual time-varying sound in the
same time period.  Equivalent sound level is also called average sound
     long term environmental impact is evaluated by the yearly day-night
sound level, symbolized as L^/yx*  The yearly average is recommended on
the grounds that the noise metric used should be one which reflects any change
in the noise environment, and that this should be done consistently for differ-
ent sources.  Yearly day-night sound level is analogous to the traffic
engineering concept of annual average daily traffic.  In other words, it is


meant to represent levels measured under average conditions, or, if condi-
tions vary during the year, weighted averages of levels at the different
times of year.
     In some instances, a rough approximation to annual average conditions or
noise levels will be sufficient; in others, it will be necessary to be more
precise.  For example, noise levels for some airports are reported for an
average busy day, rather than an annual average.  If the project under
analysis is a land-use development, it is quite reasonable to use such
existing noise information, even though it is not exactly the annual average.
The error in the data is small enough that the cost of a more exact estimate
of the annual average is not warranted.  On the other hand, if the project
being analyzed involves a change in airport use (for example Sunday flights
when there were previously none) the noise level typical of an average busy
day may lead to nonsensical results.  (The busy day level would be reduced,
in the example, if the aircraft noise on Sunday was less than the average on
other busy days—even though over the year more noise was being produced
because of the added operations.)  Approximations for the annual average L,
can be very useful shortcuts, but need to be applied with careful judgment.
     Day-night sound level is the primary measure of general audible noise,
and is appropriate for noise environments that affect a community over an
entire 24-hour day.  There are two kinds of situations where such a measure
is not appropriate, however.  The first kind consists of those situations
in which it is desirable to assess the effect of a noise environment on an
activity of less than 24-hour duration.  An example is the effect of noise
on speech communication in classrooms or in offices.  In these situations it
is useful to consider the equivalent sound level, Ii  ._ , over the time

period of interest (T) , for example one hour or eight hours, (Leq(i) and
     The second kind of situation covers those in which the noise is not
present for enough of the day to greatly affect the L,  reading, but is
intrusive and disruptive when it is present.  Examples of such noise sources
may include motorcycle passbys, trains, and specific aircraft flyovers.  The
appropriate sound measure for such an event is the cumulated sound produced
by the single event, the A-weighted sound exposure level, L.  It is a   -
measure of accumulated, not average, sound energy.
     Precise mathematical descriptions of all these measures are provided in
Appendix A.  All are expressed in A-weighted decibels; the reference sound
pressure is 20 micropascals.  •

2.1.2  Determining the yearly day-night sound level
     For the screening procedure, two yearly L,  values need to be deter-
mined:  the existing levels; and the levels expected to be caused by the
proposed project.  In addition, for the impact assessment it will be necessary
to estimate the future yearly L,  values in the area if the project is not
constructed.  (The total post-project noise level can then be calculated as the
(logarithmic) sum of the project levels and the future levels in the absence
of the project.)
     Determining Isfri) by direct measurement.  To establish the existing noise
exposure accurately, field measurements are oftentimes the preferred
approach.  Unfortunately, such measurements can be expensive and time-consuming.
Nonetheless, measurements may be warranted.  For example, if the present
average sound levels are already high, so that the noise impact of a new
project will not be much greater, or may be even less than the impact from the


existing noise environment, it may behoove the applicant to conduct a
measurement program, so as to predict the noise impact more accurately.
     When an existing noise environment is to be determined by direct measure-
ment, it will be necessary to make measurements at a number of locations suffi-
cient to establish a credible baseline for estimating total impact.  The number
of measurement locations and their geographic disposition will depend on the
spatial extent of the impact expected to be produced by the project.
     Measurement periods and the time intervals between them should be
determined by the characteristics of the existing noise, in order to obtain
a reliable estimate of yearly L, .  If the existing noise is expected to be
substantially the same from day to day, measurements during a single typical
24-hour period may be adequate.  Locations where the noise is caused primar-
ily by well-established motor vehicle traffic patterns are an example.  In
other situations where strong daily, weekly, monthly, or seasonal effects
occur, it may be necessary to measure for a number of different daily periods
suitably chosen to account properly for these variations.  In some particular
situations, the variations may be large enough to make measurement practi-
cally infeasible.  A case in point might be in the vicinity of an airport
with more than one runway which has no on-going noise monitoring program.
     The most reliable temporal data are obtained by techniques that approach
continuous measurement of the sound level over the time period in question.
In some instances it may be reasonable to obtain measurements over only
fractions of the total time—e.g., several minutes per hour.  However, any
measurement method used to approximate continuous measurement of L,  must be
justified by adequate technical reasons and data to show the accuracy of the
procedure when applied to the specific noise sources affecting the noise
environment being described.

     If field measurements are undertaken, they should be conducted  in
accordance with accepted procedures  [8] .
     Determining I^n    by the use of engineering prediction models.
Several kinds of noise have been extensively studied, particularly the
noise of transportation, and procedures have been developed  for calculating
day-night sound levels based on the type of noise source and operational con-
siderations.  Procedures for estimating the noise of specific  sources such as
roadways [9,10,11] and aircraft near airports  [12,13,14], are  available and
may be easily adopted for those situations in which the existing noise envi-
ronment is dominated by a major noise source.  A partial bibliography of some
of these engineering prediction models is included at the end  of Appendix E.
Procedures for other sources, such as railroads, have been developed and will
be published shortly.
     Determining ^jn    from the population density.  Where  no dominant
source of this nature is present, the existing noise environment may be
considered to be caused primarily by local automotive  traffic noise.  For
these instances the day-night sound level may be estimated  on the  basis of
population density in accordance with the values listed  in  Table 2,  which are
based on the equation:
                        Ldn - 10 log  p+ 22 dB                     Eqn. 1        P
where p is the population density in persons per square  mile.  This  relation-
ship was derived from measurements at 130 urban locations  [15] . The equa-
tion has a standard error of 4 dB, which means that the  95% confidence
interval around the estimate is roughly +8 dB.  The reliability of the
relationship is approximated by the correlation coefficient of 0.723 between
L,  and the log of the population density over the 130 data points.  This

can be interpreted as indicating that the log of the population density

explains 52% of the variation in Ldn.  This amount of uncertainty about

the true L,  may or may not be acceptable for a given project.  If it is

not, measurements or source-based predictions are recommended.

     The levels shown in the table represent average values for residential

areas that are not in the vicinity of an especially noisy existing source

such as an airport, a freeway, a railroad, or a switching yard.  If such a

noise source exists, its contribution to the existing L,  should be

estimated separately, and then combined with the level given in Table 2.  The

values in the table are representative of space average values over areas of
the order of 1 km  (0.4 sq. mile), or larger, for typical urban conditions.

     For purposes of estimating the existing noise in relation to permanent

changes in areas with population density greater than 20,000 persons/sq. mile,

the day-night sound level should be taken as 65 dB.  Higher estimates of the

background noise by the use of equation 1 require specific justification such

as direct measurements or detailed calculations based on existing noise sources.

The reason for this suggestion is to avoid obtaining low numbers for the impact

                                   TABLE 2

       Yearly Day-Night Sound Level As Estimated by Population Density
            (To be used only for residential neighborhoods where
                  there is no well-defined source of noise)

                             Population Density         L.  in dB   Population Density
       Description           (People/Sq. Mi.)            ""           (People/Sq.km.>

Rural (undeveloped)                  20                   35                   8
Rural (partially developed)          60                   40                  23
Quiet Suburban                      200                   45                  77
Normal Suburban                     600                   50                 232
Urban                              2000                   55                 772
Noisy Urban                        6000                   60                2317
Very Noisy Urban                  20000                   65                7722

Note:  Iyn estimates for population densities lower than 1000 persons/
       sq. mi are extrapolations.


of noisy projects  in heavily populated areas.  This  is  in  line with  the
discussion of accounting for existing noisy areas when  using  the  screening
diagram.  Particularly when an area  is noisy because of high  population
density it is important to consider very carefully, any  project which will add
to the noise in the area, and therefore not to initially screen it out by
using high estimates of existing levels.
     With respect  to problems of estimation in rural areas, there simply is
not enough known about noise levels  in such areas, since measurements such as
those used to calculate equation 1 are routinely conducted only in the absence
of wind, rain, and other natural sounds.  Values obtained  using equation 1 (or
Table 2) extrapolate beyond the data base, and should be used with caution.
Whenever possible, measurement of existing levels is recommended.
     Estimation of future noise levels with and without the proposed project.
Most of these procedures which have been identified  for estimating
existing noise levels can also be used, as appropriate, to estimate  the
noise levels due to the proposed action or project.  Prediction procedures,
approved by various federal agencies, are available  for a  number  of  typical
situations, including aircraft, motor vehicles, railroads, construction equip-
ment and other noise sources.  In some instances those  procedures do not pro-
vide predictions for L, .  In those cases, the conversion  equations  provided
in Appendix B can  be used to estimate the L,  values, in order to use the
information in these guidelines to express the impact of the  noise on people.
     If prediction techniques are not available for  a particular  project, for
example a specific industrial installation, measurement of noise  levels at a
similar existing installation is appropriate, although  an  engineering descrip-
tion should be included of the reasons for anticipating such  similarities in
a new installation.  Likewise, where the introduction of a new noise source

is anticipated and neither an existing approved procedure nor a similar
installation is available, an engineering description of the procedure
employed to estimate noise emissions  should be provided in adequate detail
for technical evaluation of its acceptability.
     For predicting future noise levels  in the absence of the proposed  proj-
ect, the available methods are  (1) extrapolation of existing levels to  the
future, (2) the use of source-specific prediction techniques, or  (3) the  use
of equation 1.

2.1.3  Determining the population affected by the noise of the proposed
     For each of the alternatives that involves the introduction  of some  form
of a new noise source, the affected population is defined as that population
experiencing sound levels produced by the new noise source above  a specified
yearly L- .  Ihis will be called the  yearly L, , or base *v.  . ..  Hie
base yearly L,  may be determined by  references to existing yearly L,   con-
tours in the area of interest  (See section 2.1.2).  Consistency with the
screening diagram requires the  consideration  of impacts whenever  the* overall
post-project level is greater than the pre-project level, that is, when the
project alone L.  is greater than pre-project L,  less 10 dB.  Thus, the
base L,  will usually be 10 dB  lower  than the existing (pre-project) yearly
L. .  For example, if the existing yearly L.  is 60 dB, it is suggested
to start with a base L,. . of.  50 dB, if possible, in order to determine
the number of people affected by the  project  noise.  In some instances, however,
it will not be feasible to predict project noise levels to such low values.
(An example of such an instance would be around commercial airports, where
existing prediction techniques  are not particularly reliable below an L,  of
65 dB (NEF 30), due to lack of  information about aircraft flight  track  usage.)

In such cases, it is still imperative to consider as large a range of
levels as is feasible.  A difference of twenty decibels between the maximum
Ldn(v) for the P0^64* an^ the base Ldn(v) is a good ran9e to attemPt to
achieve, providing that it results in a base L,,. of 55 dB or less
since L^ = 55 dB has been identified as a point of significant adverse
noise effects.  If the procedure results in a higher L, . . , a base
level of 55 dB should be chosen.
     When several alternatives are compared, a common base area or base
population must be used for all alternatives.  In such cases the base area
or population for all alternatives will be the largest area or population
affected by any alternative.  In other words, the base population will
be determined by the project which has the highest yearly L,  in a given
location in a given year.  The reason for requiring a common base is that
several of the measures of relative impact, to be discussed subsequently,
will be meaningless if the total number of people over which they are cal-
culated changes from one alternative to the next.  The base population,
therefore, should include all people who are affected by the noisiest alterna-
tive.  As a consequence, for some of the less noisy alternatives the base
population will be considerably larger than the population actually affected
by those alternatives.  If the base L^. . is consistent with the screen-
ing diagram, no person exposed to project noise levels less than the base
Ldn(v) wou^ be rsgarded as impacted.
     There are cases when, over time, people will move into or out of a pro-
ject area at the same time the project is expanding and environmental noise
levels are increasing.  Such changes in population may be entirely unrelated
to the project under analysis.  In these cases it may be necessary to define

several base populations or base areas, one for each year of interest.   (See
Appendix E, section E.2 for an example of this type of analysis.)
     There are actions that do not add new noise sources, but only change  the
noise output of existing sources.  In these cases, the changed source should
be treated as a new source for purposes of determining the base population.
     There are actions that will neve people into noisy areas.  For these  cases,
the base population will be that population that is moved into an area  in  which
the existing yearly L^n is greater than 55 dB.
     There are actions which affect large segments of the population that  are
not easily related to specific areas.  Laws and regulations that directly
affect mobile noise sources are examples of such actions.  For actions  affect-
ing regulation of noise sources in general, the base population might best be
described as the total population experiencing day-night sound levels
above 55 decibels from such sources.  For actions affecting source control
for equipment operators, the base population might be only the users of the
specific noise source.  In the final analysis, the preparer of a noise
impact analysis must use his or her judgment.  In all cases, an explanation
should be included in the final report of how the base population was determined.
     Population estimates for residential areas identified in the analysis
may be taken directly from census tract data, local master plans, or by
counting residential units identified on aerial photographs of the area.
Non-residential populations may be estimated from industrial, commercial,
or public facility employee statistics; student enrollments and employee
statistics can be used to estimate school populations.  Population esti-
mates should strive to identify total populations within ±10 percent of
the true population.  One way to deal with uncertainty in predicting future

populations is to use the local community's  land use plans or  zoning designations,
whenever they exist, to estimate the most likely future  population density.

2.2  Health and welfare effects range
     This section deals with the most commonly  encountered noise problems,
the general health and welfare effects of noise due to the noise environment
encountered in most urban and suburban areas.   Those effects are the major
concern at yearly L,  values which range approximately from 55 dB  to 75 dB.
Above 75 dB, severe health effects need to be considered in addition to the
effects considered here (see section 2.3.).  The first sub-section describes
the health and welfare criteria which apply  in  this range of general audible
noise, and the second covers the procedures  to  be  used to quantify those

2.2.1  Human noise exposure criteria
     As the primary criterion for evaluating the impact  of noise on people,  the
effect on "public health and welfare11 was selected in  the Levels document [2].
Interference with speech communication, with general well-being, and with
sleep are related to the general annoyance produced by the noise environment,
and were accepted as indicators of effects on public health and welfare.  The
same criteria are proposed here as the basis for environmental impact  assessment.
     A summary of the expected effects of noise on human activities for outdoor
yearly day-night sound levels of 55, 60, 65, 70, and 75  dB, in terms of
health effects, interference with speech communication,  community  reaction,
annoyance, and attitude towards the area is  provided in  Appendix C.  Basic infor-
mation in these tables on speech intelligibility and general community reac-
tion was derived from the Levels document  [2].  The relationships  given in the

Levels document between noise and annoyance have been modified in the light of
a substantially increased set of data subsequently available  [5].  These tables
allow the preparer of a noise analysis to make an explicit statement as to the
expected impact of any day-night sound level.
     In order to achieve the simplicity which these guidelines are intended
to promote, it is desirable bo be able to summarize these several health and
welfare effects with a single indicator.  The response of interest is the
general adverse reaction of people to noise, which includes speech interfer-
ence, sleep interference, desire for a tranquil environment, and the ability
to use the telephone, radio, and television satisfactorily.  A measure of this
response is the percentage of people in a population that feels high annoy-
ance about noise of a specified level.  High annoyance is selected on both
theoretical and practical grounds.  First, it arises as a consequence of
the activity interference and interruption caused by noise [16], and there-
fore summarizes all the effects better than any one of the direct effects
would.  Second, there is available a large set of data which allows reponse,
expressed as percentage of a population highly annoyed, to be characterized
by a single functional relationship of the noise environment  [5].
     The percentage highly annoyed is used rather than the percentage at all
annoyed for a number of reasons [5, pp. 378-379].  Perhaps the most important
of these is that when people are highly annoyed by noise the effects of non-
acoustical variables are reduced, and the correlation between noise exposure
and the expressed subjective reaction is high.  This is not to say that all
individuals have the same susceptibility to noise; they do not.  Even groups
of people may vary in their response to noise, depending on previous exposure,
age, socio-economic status, political cohesiveness and other social variables.

In the aggregate, however, for residential locations, the average response of
groups of people, as measured by the percentage highly annoyed,  is quite stably
related to cumulative exposure to noise as expressed in a measure such as L,.
     For schools, offices, and similar spaces where ease of  speech communica-
tion is of primary concern, the same relationship can be used to estimate
the potential average response of people, as a group again,  ignoring  individ-
ual variations from person to person.
     Data used to relate annoyance to noise environment in the Levels document
[2] was based on two social surveys around airports in the United States and
England.  Data have now been analyzed from 19 social surveys (in 9 countries)
associated with aircraft, urban traffic, freeway traffic, and railroad noise
[5].  These data allow a much more definitive relationship to be developed
between percentage of the population highly annoyed and average  noise level.
The data support the previous assumption that the statistical relationship
between population annoyance and noise level is essentially  independent of
the type of noise source  [17].
     The results of this synthesis show quite clearly that the best fit of
response data to average sound level is provided by a curvilinear function;
usually a cubic equation was used in the regression analyses.  Further, 12
of the surveys, covering aircraft, railroads, urban traffic, and expressway
traffic as noise sources, "clustered" closely around an average  curve for the
set of data, as shown in Figure 4.  The remaining 7 surveys  showed similarly
shaped annoyance/sound level functions, but deviated in differing detail from
the 12 clustering surveys for various qualitative reasons  [5].   It is worth
noting that the average of the non-clustering surveys was essentially the
same as the average for the clustering surveys.

5  50

-~*~» 1ST HEATHROW A/C (1961)
	 FRENCH A/C (1966)

***»< SWIS.
oo««» U.S. S


» •^^rf~J^- *
HEATHROW A/C (1967)-
ICH A/C (1969)
S STREET (1969)
)ISH A/C (1972)
5 ROAD (1972)
30N STREET (1972)
> A/C (1973)
CH RR (1973)





1 /M . •


// ./


60            70
                     SOURCE: SCHULTZ[5]

     Based on these data, Schultz proposes the following equation "as the

best currently available estimate of public annoyance due to transportation

noise of all kinds" [5, p. 382], relating percent highly annoyed, (%HA),

to day-night sound level:

               %HA = 0.8553 L^ - 0.0401 L^2 + 0.00047 L^3          Bqn. 2a

This expression represents the least-squares fit of percent highly annoyed

to day-night sound level for the clustering survey data.

     A second form of this equation, based on two power law functions, is

preferred on the grounds that it suggests an explanation for the behavior

represented in equation 2a.  At lower levels, the first power function

represents increasing awareness, or arousal.  At higher levels, annoyance

increases at the sane rate as the well-known loudness function, represented

by the second power function.  The smoothed version of the function based

on the two power law functions is expressed as*:

                     (1.24 x 10~4)
            (1.43 x 10"4) (10°'08I<3n) + (0.2) (IQ0'031^")            Bqn. 2b

     In the absence of any studies relating average response to noise level

for non-transportation sources, equation 2b has been adopted in these

guidelines for use as the criterion for all noise sources.  If information
'Another very useful and simpler expression which approximates the annoyance
function is:

              %HA = —
                    i j. J10-43 - 0.132 L. )
                    1 + e                dn

This expression has the particular advantage of not allowing predicted values

to go below zero percent or above 100 percent.
                                                                                W "A

becomes available which identifies a different relationship for certain
sources, the guidelines will be revised accordingly.
     This function for the percentage highly annoyed differs from previously
suggested equations, including the one in the Levels document  [2, p. D-27],
as is illustrated in Figure 5.  The relationship shown in the Levels document
was taken from a study by an EPA Task Group under the EPA Aircraft/Airport
Noise Study 'in 1973 [4].  In this study, social survey data from the first
study around Heathrow airport in England [18], and from the Tracer study of
U.S. airports [19] were combined to develop a relationship between "percent
highly annoyed" and day-night sound level.  This function was expressed as:
                     % Highly Annoyed = 1.8 (L^ - 46)                 Eqn. 3
The Task Group also noted a similar relationship developed in an OECD  study
[20] that used the relationship:
                     % Highly Annoyed = 2 (L^ - 50)                   Eqn. 4
This equation was also based on airport noise studies.  The primary reason
for these differences is a redefinition of what is meant by highly annoyed.
In fact, the Heathrow study is included in the clustering surveys (Figure
4). As Schultz's paper makes clear [5, pp. 391-392], the annoyance scale
used in the first Heathrow study requires some interpretation:  it is  not
a direct question about degree of annoyance.  The earlier analysis (Eqn. 3)
considered the top three scale points as highly annoyed; Schultz used  only
the top two.  His discussion is persuasive, and his function has been  adopted
for these guidelines.
     It is important to point out that this redefinition of annoyance  does
not affect the conclusions reached in the Levels document, because that
document relied on speech and sleep interference indicators to identify the

   80 i-

     40 •



                              NEW ANNOYANCE FUNCTION
50       60       70       80



actual levels which were "requisite to protect the public health and welfare
with an adequate margin of safety."  That approach led to the statements that
a day-night sound level of 55 decibels in residential areas will result in
negligible impact on public health and welfare and that the degree of impact
will increase as the day-night sound level increases.  The EPA levels document
asserts that no significant effects on public health and welfare occur, for
the most sensitive portion of the population and with an adequate margin of
safety, if the prevailing day-night sound level is less than 55 decibels.
The difficulty with using annoyance for such a calculation is obvious in
Figure 5:  there are still some people affected at sound levels as low as 45
dB (La).  These guidelines, then, use as the criterion in populated areas
the function given by Eqn. 2, which shows some impact at levels as low as
45 dB, impact which is fairly low into levels in the low 60's (dB), and
impact which begins to increase fairly rapidly above 65 or 70 dB (L^J.
     For those events in which the single event measure, sound exposure
level, is used to describe the noise environment, the previous discussion
will not apply.  Information characterizing response as a function of sound
exposure level is not readily available.  Some information can be approxi-
mated for sleep interference [21,22,23,24] and speech interference [22,24],
but it is not as easily dealt with as is the information on L, .

2.2.2  Quantification of the noise impact,.
     The impact of a noise environment on people regularly experiencing the
environment is the degree to which the noise interferes with various activi-
ties such as speech, sleep, listening to radio and TV (i.e., the peaceful
pursuit of normal activities), and the degree to which it may impair health,

through, for example, the inducement of hearing loss.  Sound levels produced
by sources being considered in an environmental assessment will generally
vary with distance from the source, sometimes over a large geographic area.
As a consequence, people occupying different geographic areas will experience
different sound levels.  The total impact of a particular noise environment
is a function of both sound level and the size of the population experiencing
a particular value of sound level.
     The first step in describing the noise impact of an action is to tabulate
the number of people regularly experiencing various sound levels.  In many
cases, particularly those in which noise impacts must be compared with a vari-
ety of other costs and benefits, such a tabulation is insufficient, because it
contains too much information.  In those cases, it is desirable to derive a
single number which represents quantitatively the integrated impact of the
action on the total population experiencing the different sound levels.  This
single number quantification is defined below as the sound level weighted popu-
lation, LWP.  Sound level weighted population, together with the tabulations of
populations experiencing sound levels of a specified value, constitute the mini-
mum quantification of environmental impact of noise recommended in these guide-
lines.  This subsection describes procedures for preparing the tabulations, and
for calculating the sound level weighted population.  It also describes a useful
second descriptor of noise impact, the noise impact index, Nil, which is formed
by the ratio of sound level weighted population to the total base population.
The procedures proposed here do not rely on establishing specific criterion
noise levels for different land use categories.  (For information on criterion
levels suggested by different organizations, see Appendix B.)

      a.   Necessary tables - As  a minimum the data characterizing the noise
 impact should be tabulated in a set of summary tables.   Typical tables are
 included in the example in section 2.6 (Tables 6 through 9).   For a given year
 the areas and population are to be listed against the yearly day-night sound
 level at increments not greater than five decibels,  for the following esti-
 mated noise environments:
          (1)  without the project's existence;
          (2)  due solely to the project action;
          (3)  due to all sources including the project  action.
'All three tables may not always be necessary, especially if there are insig-
 nificant differences between any two of the tables.
      If  the tables are properly constructed, the total  population and/or land
 area for each of the three conditions will be equal  (i.e., will equal the base
 population or area defined in section 2.1.3).  The tables should include enough
 increments of yearly L,  that all  residential populations, industrial, commer-
 cial land and special situations experiencing L,  values above the base
 L^ are  included.
      The column headings might  typically include: total land area, industrial/
 commercial land area, residential  land area, industrial/commercial employees,
 residential population, and special situations.  Depending upon local condi-
 tions, different classifications of land use may be  appropriate.  Industrial/
 commercial land area is meant to include all land not considered as residential
 or associated with special functions.  This land area would include farm land,
 undeveloped land, industrial plants, and similar uses.   Depending on local
 plans, this category may be further broken down.  Residential land includes
 all land associated with a residential population.  It  may include land actually

zoned oonntercial or industrial.  For residences on farm lands,  approximately
1 acre should be considered as residential  land for each separate residence.
     Special situations are those situations which must be highlighted or
treated separately in order to represent the impact properly.   Situations
of this category can be religious facilities, outdoor auditoriums, schools,
precision laboratories, hospitals, etc.  The detail to which each special
situation should be discussed will depend on the  size of the project and the
size of the area being evaluated.  Special  situations should be combined as
necessary to keep the total number of special situations within reason (nor-
mally less than 20 or 30 items).  One useful approach to the listing of spe-
cial situations is to number each one, and  then to use this number in the
special situation column to indicate the corresponding L.  for  that
situation (see'examples, section 2.6, Tables 6 through 8).
     If there are more than a few special situations, an additional table
summarizing them will also be useful (Table 9 in  the example in section 2.6).
This should list the number of exposed people for each situation.  At some
locations the population does not remain constant from day to day, week to
week, or month to month.  Examples of such  places are churches, parks, and
stadiums.  In such situations the population entered in the special situation
table is the time-weighted average number of people present during the year.
This number should be calculated by summing the products of the number of
people using a facility, multiplied by the  number of hours these people are
present in the facility during a year, and  dividing by the total number of
hours in a year.  If a noise measure other  than yearly L,  is being used,
the average number of people can be calculated similarly for that time period.

such as the working day for office buildings.  The concept of average number
should not be used for residential areas.
     Formats other than that used in Tables 6 to 9 may be appropriate and
may be used; however, the information conveyed to the reader should be
effectively the same as or greater than is contained in these tables.
     For each alternative of a permanent project or action, a separate set
of tables as outlined above should be prepared for (1) the first year of the
commencement of the project, (2) the last year before the end of the project
(or at the 20 year point, whichever is shorter), and (3) the worst case year
if such a year is not the first or last year.  In many cases, only one of
these sets will be necessary because the conditions with respect to time
can be expected to remain reasonably constant.  By "reasonably constant," it
is meant that the change in exposed population will be small enough so any
resulting errors are consistent with the error in the overall analysis.
     In addition to the tables, it would be helpful to present a map or
drawing of the area including surrounding facilities such as airports, fac-
tories, highways, or electrical plants, with contours representing constant
values of yearly day-night sound level.  In general, the decibel increments
between contours should be consistent with the tables as discussed above.
Other contours may be presented as needed.  There should be a set of contours
for each of the alternatives studied.
     b.  Sound level weighted population - For those projects in which it
is necessary to compare or trade-off noise impacts with other costs and bene-
fits of a proposed project, a compilation of the data characterizing noise
impacts into the tables described above will usually not prove sufficient.
The tables contain too much information for easy comparison to be possible.

A single number representation , combining the extensity (number of people
exposed) and intensity (severity of the exposure) of the noise impact is
desirable.  Using the criterion function based on the percentage highly
annoyed (Eqn. 2), described in section 2.2.1., such a single-number index
can be constructed which summarizes the impact in terms of the total number
of people who respond adversely to the effects of noise.
     Several assumptions are made in this method of analysis.
         ( 1 )  The intensity of human response is a measurable consequence of
equivalent sound level, and in the noise range of interest here (namely that
generally encountered in populated areas), is appropriately measured as the
percentage of the population which is highly annoyed.
         (2)  When measured this way, it is clear that the impact of high
noise levels on a small number of people is equivalent to the impact of
lower noise levels on a larger number of people in an overall evaluation,
when both yield the same number of people responding adversely.  Thus the
properties of intensity (level of sound) and extensity (number of people
affected by the sound) can be combined mathematically.
         (3)  On the basis of these two assumptions one can ascribe differing
numerical degrees of impact to different segments of the population of con-
cern, depending on the equivalent sound level, and can sum over all of these
segments to obtain the total impact (total number responding adversely).
     On the basis of these assumptions, the following equation is obtained
for the sound level weighted population, UWP:
  r-                                                 LU/P
f Pt^Jn)  • W^)  ddfln)                 Eqn. 5   ^
where P(I^n) is the population distribution function, W(I^n) is the weighting
function described in Equation 2b, characterizing the severity of the impact

                                   TABLE 3  '

        Values of the Weighting Function for General Adverse Response*
                                    = (0.01)%HA]
             Ldn    W
as a function of day-night sound level  (Table 3), and dfl^fa)  is  the differen-
tial change in day-night sound level.  Although Table 3 contains values for
I^ln as low as 35 dB, the values below 45 dB should be used only  with great
caution, as they represent extrapolation beyond the range of  the data  [5] .
In any event, for most projects in populated areas, the future noise level even
without the project will probably be considerably higher than the 45 dB limit
of the data.
     It is usually not necessary or possible to use the integral form  to  com-
pute LWP.  Sufficient accuracy is obtained by taking average  values of the
weighting function between equal decibel increments, up to 5  decibels  in
size, and replacing the integrals by summations of successive increments  in
average sound level:
     d.  Change in level-weighted population and relative change in impact -
A primary concern in an environmental noise assessment is a comparison of ( 1 )
the effect of the action on the noise environment with (2) the environment
before the action was to take place.  Two additional descriptors of this
change due to the action are useful.  The first descriptor is simply the
numerical change in sound level weighted populations before and after the
action, the change being an increase or decrease in sound level weighted
population (or the neutral effect case, no change).  The second descriptor
is the relative change in impact (PCI), where the effect of the action is
expressed as the value of the change in the sound level weighted population
after the action, divided by the sound level weighted population before the
                               a - LWP.                               __  ft
                     RCI =  - *• - £                              Bqn- 8
where LWP  is the impact after the action or project is in place, and LWP.
         cl                                                               D
is the impact before the action is taken.
     e.  Level-weighted area - In those rare cases where it is known that an
area will be developed, but there is no information with which to predict the
future population, it may be necessary to calculate a level-weighted area, as
a proxy for the population impacts.  Such a calculation would be equivalent
to that for level-weighted population (equation 6), but would use the tabula-
tion of area within decibel intervals, rather than population, assuming, in
effect, a constant and undefined population density.
     f .  General discussion - A number of different noise impact descrip-
tors are available, based on the four single-number indexes — level-weighted
population, noise impact index, change in level-weighted population, and

relative change in impact—and the three noise characterizations—the project
alone, the environment without the project/ and the total future noise envi-
ronment obtained by combining the other two.  Die result is almost a confu-
sion of supposedly simplifying descriptors.  Two or three of these will be
most useful in each case, depending on the relationship between project levels
and expected levels without it (Figure 6).  Where existing levels are already
high, the level-weighted population or noise impact index based on the project
noise alone is suggested as the best descriptor.  The other descriptors will
minimize the impact by putting it on relative terms.  Where project levels
are much higher than existing levels, the project levels will dominate the
combined levels, so either will give the same result.  Where project levels
are similar to existing levels, it is necessary to use the combined levels to
identify the full impact.
      For projects which will move people into areas with Lfa values above
55 dB, and for projects which reduce noise, Figure 6 is not applicable.  Pro-
jects, such as housing developments, in areas with 1^ above 55 dB need to be
evaluated in terms of the level-weighted population or noise impact index
based on the non-residential noises to which they will be exposed (e.g. road
traffic or aircraft noise).  The only basis for calculating the change in
level-weighted population or relative change in impact which might be useful
in such a situation is one based on the national average Nil, which has been
calculated to be 0.35.  Projects which reduce noise, on the other hand,
should be evaluated on the basis of the change in level-weighted population,
or relative change in impact.  Since the project is proposed to reduce noise,
it is obviously the reduction or change which is of interest.

' 1
^ tf\
•»- 70
£ 60
O •
5 50
H- 40


^^ ^^ ^» m^ «^ ^^
"*"* J

°- / X
x 4 X
"J LJ^. I/ 1 1


111 w^
                 40    50

      Relationships between annoyance and average sound level have been used
previously to define a weighting function for numerical evaluation of impact
analyses.  It is useful to compare the present function (Eqn. 2 and Table 3)
to the one used earlier by EPA, which was first introduced in the fractional
inpact method developed initially for use in the analysis of highway noise
problems [25].  This method took into account the data and recommendations
both of the EPA Levels document [2], and of the earlier report on Impact
Characterization of Noise [4], which indicate that a community would not be
expected to exhibit significant reaction at noise exposures of L,  = 55 dB or
below, but would be expected to show strong, organized reaction at L^ = 75 dB
and higher.  Using these two anchor points, and the linear relationship of
Equations 3 and 4, a weighting function, "fractional impact,11 F.I., was defined   £" T"
to be zero at L,  = 55 dB, and unity at L,  = 75 dB, varying linearly with
average sound level, such that:
               P.I. = 0.05 (L^ - 55)                                  Eqn. 9
The weighting function for F.I. has been used by EPA in impact analyses of
a number of potential regulatory actions.
     Several features of equation 9 are unsatisfactory.  It is not likely
that community response is adequately described with a linear function of
average noise level over a wide range of levels.  Even though the data from
the individual social surveys are reasonably well fitted by linear regres-
sions over the limited range of levels represented in the separate surveys,
the individual survey results indicate that the rate of change of annoyance
with sound level is greater at higher sound levels than at lower sound levels.
Moreover, the choice of an arbitrary zero at L,  = 55 dB is not easily justi-
fied.  Finally, few data from noise sources other than aircraft were available

at the time the original weighting functions were developed, and a weighting
function derived only from aircraft-related social surveys may not be  satis-
factory for use in evaluating other sources of noise.
     Despite these flaws, however, this linear function is quite similar in
its relative ratings to the curvilinear function used in this document (Egn.
2).  If the two functions are placed on the same scale (Figure 7), it  can be
seen that, in the day-night sound level range of 55 to 80 decibels,  this
linear weighting function will generate relative values for level  weighted
population that differ only by the order of one percent from the curvilinear
weighting function in many applications.  The change in scales necessary to
make this comparison stems from the fact that fractional impact was  defined
to be unity at L,  = 75 dB, while the present function is based on the
number of people reporting a high degree of annoyance in a social  survey
situation.  Both are equally legitimate interpretations of available impact:
the first provides an indicator of absolute impact, while the second is more
easily understandable in comparisons with other costs and benefits of  pro-
posed projects.  Because the linear function (Eqn. 9) closely approximates
the curvilinear relationship (Eqn. 2) between the day-night sound  level range
of 55 to 80 dB, the user may wish to employ the more simple linear relation-
ship in some cases.

2.3  Severe health effects range
     In some high level noise environments people will be exposed  regularly
to 24-hour equivalent sound levels in excess of 70 decibels.  In these environ-
ments special consideration should be given to the potential for severe health
effects.  This section discusses the criteria to be used for describing severe







         UJ  _s>«.
         DC  20'
         UJ  -••

         -'  -Tdi
                         0.000143) 10°-08Ldn ^
                                          80    .90
                          Ljn ('DECIBELS)'.


health effects, and then describes a procedure for calculating a single number
index, analogous to the level-weighted population index, for statistically sum-
marizing expected severe health effects.

2.3.1  Human noise exposure criteria
     The discussion of severe health effects in an environmental analysis
is meant bo supplement the discussion of general adverse effects, not to
replace it at high noise levels.  The general adverse effects - speech inter-
ference, sleep interruption, annoyance - continue to be present at high noise
levels, and in fact increase more rapidly at the higher noise levels.  Equa-
tion 2 (or Table 3) can be used to summarize these effects for 1^ values as
high as 85 dB.  These effects, however, only include those of which people
are aware, and which have been articulated in attitudinal surveys.  In many
instances, people are not aware of the potential severe health effects which
long-term noise exposure can cause.  Hence a separate discussion of severe
health effects is necessary, which helps to emphasize the severity of the
problems caused by high noise levels.
     Noise-induced hearing loss can begin bo occur at high noise levels.  Other
noise-induced physiological effects and/or changes may occur.  However, a firm
causal link between community noise and extra-auditory disease has not been
established at this time.  Therefore, this document proceeds on the assumption
that protection against noise-induced hearing loss is sufficient to protect
against severe extra-auditory health effects.*  However, one has to keep in
mind that as the noise level increases above the threshold for severe health
 This is not to say that non-auditory physiological effects do not occur at
 levels below those sufficient bo protect against hearing damage.  In any event,
 rigorous causal links between noise and extra-auditory health effects have not
 yet been firmly established, but await further study.

effects so does the probability that other health effects in addition to noise-
induced hearing loss might become important.  The adverse effect of noise on
hearing rapidly accelerates as the noise exposure increases and it is reason-
able to use expected noise induced hearing loss as a basis for assessment of
severe health effects.
     A problem arises in specifying the noise measure to be used when quanti-
fying severe health effects.  If hearing loss is used as the indicator, the
noise measure needs to reflect at-ear measurement to be valid.  Further, hear-
ing loss is properly expressed as a function of L  , rather than of L, , but
it will not usually seem warranted to calculate and draw noise contours for
more than one noise measure.  The data to be discussed below predict noise
induced permanent threshold shift (NIFTS) for 8-hour equivalent sound levels
(at-ear) starting at values of 75 dB.  If the remaining 16 hours of the day
are spent in a noise environment of 70 dB L   or lower, the at-ear 8-hour
equivalent sound level of 75 dB results in a 24-hour long L   of 70 dB, at
the ear.  For many proposed projects (particularly actions where assessment of
hearing damage risk is of primary concern) it will be appropriate to use one of
these two measures.
     It is also important to be able to identify the 1^ values at which it
is appropriate to look for severe health effects.  Those persons who have the
greatest outdoor activity, including young children, retired persons living
in warm climates, and people in certain outdoor occupations, are clearly the
people of major concern when outdoor 1^ is considered.  For outdoor exposure,
daytime levels are the important ones for establishing at-ear values.  The
values of 1^ corresponding to an A-weighted equivalent sound level of 75 dB
during daytime hours range between 73 and 81 dB.  The lower value corresponds

to a situation where the equivalent sound level during the night is 20 dB or more
lower than that occurring during the day, whereas the higher value corresponds
to the situation when the equivalent sound level during the night equals that
occurring during the day.  The most probable difference between the daytime
and nighttime values of I^g is 4 dBf as shown for the noise levels of interest
in Fig. A-7 of the Levels document [2].  For this day-night difference, L^n
is three decibels above the daytime value of I^q, that is, L^ = 78 dB.  This
value of 78 dB is considered to be the most probable value of L^n to be found
in real environments that have a daytime I^g of 75 dB.  (This estimation is
based on that in reference 4, pp. B-8 - B-9.)  However, due to the wide range of
possible values, it is recommended that an outdoor 1^ of 75 dB be used as the
threshold above which severe health effects are investigated.  This has the
advantage of being an Lfa value for which contours will already be mapped, and
is therefore information readily available.
     Consequently, for areas with L^n of 75 dB or above, it is important to
look for potential severe health effects.  The way to do this is to estimate
the size of the population spending time outdoors, the length of time they
are outdoors, and the actual levels while they are outdoors.  The last two of
these numbers can then be used to estimate the at-ear 24-hour Lgq for these
people (using the equation for I^g in Appendix A).  As long as the outdoor
noise exposure exceeds 3 hours per day, the contribution of the indoor noise
environment may be neglected in computing the 24-hour I^g.  This conclusion
does not depend greatly on the actual noise attenuation provided by the house
so long as the attenuation is greater than 10 dB [4, p. 8-9].
     There have been numerous studies conducted for the purpose of determining
the long term effect of noise on the hearing ability of an exposed population.

In particular, three studies  [26,27,28] have provided reasonable predictive
models of the relationship between noise exposure and changes in the statistical
distribution of hearing levels of the exposed population.  These changes are
called Noise Induced Permanent Threshold Shifts  (NIPTS).  The results of
these three studies were combined [29] and used  in the  EPA Levels document
[2, Table C-1], to provide a summary of the expected NIPTS that would occur
from a 40 year exposure beginning at an age of 20 years.
     Inspection of Table C-1 in the EPA Levels Document [2] shows that as the
average sound level of the exposure increases, there is a widening of the fre-
quencies affected by the exposure.  As would be  expected, the average of 500
Hz, 1000 Hz, 2000 Hz and 4000 Hz does not show a uniform constant increase in
loss with a rising exposure level, but instead increases at an accelerated
pace with increasing average sound level.  While use of the most sensitive
frequency is proper for the determination of an  absolutely safe daily equiva-
lent sound level, assessment of the relative impact of  exposure to higher equi-
valent sound levels requires that all audiometric frequencies be considered.
Therefore the average of 0.5 kHz, 1 kHz, 2 kHz,  and 4 kHz is the recommended
measure.  Since each of the four frequencies describe the center of the
preferred octave bands, there is no overlapping  in octave bands as would be
the case if 3000 Hz was included.
     Having selected a method to handle the question of frequency, the next
problem is time.  One way to consider time is to select a point in time at
which the relative impact will be described.  Selection of such a point is
somewhat arbitrary and not entirely meaningful.  For instance one could argue
that it is more important to describe the effects of noise when a person is
middle-age, and not when a person is 60 years old.  An  alternative approach

is to use the average NIPFS of the population during or over a normal working

lifetime.  Averaging NIPTS with respect to time avoids arbitrarily selecting

any one point in time and provides a realistic assessment of the overall

effect of noise on hearing on a large population, recognizing that many indi-

viduals, because of differences in sensitivities and ages or lengths of expo-

surer may incur either more or less hearing loss than would be assessed using

this procedure.

     A grand averaging of the NIPFS with respect to frequency (0.5 kHz, 1 kHz,

2 kHz, 4 kHz) and time (0 to 40 years of exposure) and percentiles (0.1 to 0.9

percent iles) from references 2 and 29 is listed in Table 4.  These NIPTS data

can be very well described by the formula:

                 Ave NIPTS = (Leq(8) - 75)2/40 = (I^q(24) -70)2/40       Bqn. 10

The slight differences shown in Table 4 between equation 10 and the NIPFS

data should be considered insignificant, especially in view of the fact that

the original data were rounded to the nearest whole integer in any case.

     Equation 10, then, is the criterion for estimating the potential severe

health effects due to a proposed project.  For applications, it can be

                                   TABLE 4

                Average Hearing Loss as Function of 8-hour
     8)                    Ave. Hearing Loss              (LeafS) ~ 75)2/40
                                    dB*                          dB
   75                                0                            0.0
   80                                1                            0.625
   85                                3                            2.5
   90                                6                            5.625
   95                               10                           10.0
*Source: Table C-1 of Levels Document [2], and Johnson [29]

calculated directly, read from Table  5, or read from Figure  8.  The outdoor

day-night sound level, L^n, should be used only to  identify  potential problem

areas.  Within those areas, an effort should be made to estimate  the actual

exposure of groups of people.  In any application,  it  should be remembered

that since this equation was developed from averaging  the effects of noise

over frequency, time, and percentiles, it cannot estimate the effect on an

                                   Table 5

                 Criterion Function for Severe Health  Effects
             or 1^-jn        I%q(24)                  °B  loss
          (dB)                (oB)                   	

           75                 70                      0
           76                 71                      0.025
           77                 72                      0.100
           78                 73                      0.225
           79                 74                      0.400
           80                 75                      0.625
           81                 76                      0.900
           82                 77                      1.225
           83                 78                      1.600
           84                 79                      2.025
           85                 80                      2.500
           90                 85                      5.625
           95                 90                      10.0

individual at one audiometric frequency at one point of time.  This equation

should be used only to assess the average relative  impact of exposure  to

different equivalent sound levels.

     It is also useful to look at individual susceptibility to noise induced

hearing loss.  Therefore, a user may wish to consider the NIPTS  for the most

sensitive ten percent of the  population after 40 years  of exposure.  This  infor-

mation can be read from the 'Max. NIPTS 90th Percentile1 curve of Figure 8.

eg  10



                                 HEARING LOSS


2.3.2  Quantification of the impact
     The first step in quantifying the impact is to construct the tables indi
cating the number of people within decibel intervals.  In many  instances the
same tables setting out the extent of the general audible noise impact can
serve for this noise range also, but if there are very many people exposed to
high levels , smaller contour intervals are recommended for tabulating the
severe health effects.
    As with the general adverse effects, it is desirable to quantify the
exposure of individuals to different levels by a single number.  A term simi-
lar to the level-weighted population may be calculated using the hearing loss
function (eqn. 10) identified in the previous section.  This would result in
a hearing loss-weighted population, HWP, measured in terms of hearing loss,
expressed as person-decibels:

          HWP = f P(W8)) - H(Ioq(8)) . d(I«j(8))             Eqn. 11

where P(Ieq(8)) is the population distribution as a function of 8-hour Leq,
H(Leq(8)) is the weighting function given in equation 10 (and Figure 8 and
Table 5), and d(Leq(8)) is the differential change in 8-hour average sound
level.  Replacing the integrals by summations of successive increments in
average sound level we have:

          HWP =    P(I)   •  H(I)                     Bn- 12
where i indexes the successive increments  in average sound level.   If the
Leq (24) measure is preferred for a particular application, summation would
start at 70 dB.

            The disadvantage of the hearing-weighted population is that it is not
       easily understood:  the product of persons and decibels of hearing loss is
       not an intuitively obvious concept.  A more understandable indicator of
       severe health effects is the average potential hearing loss (PHL) which is
       analogous to the noise impact index for general audible noise:

       where the terms are as defined for equation 12, and Ptotal ^ equal to the
       base population, which is normally the population exposed to levels above 75
       dB.  Care must be taken in defining the base population, however.  If it is
       to be used to compare alternatives, the same base population must be used for
       all.  Otherwise, the average hearing loss could be lowered by a project which
       affected more people, and the indicator would not be a reliable measure of
       impact.  The simplest approach is to use as the base population the largest
       total population subjected to severe health effects by any of the alternatives.
       If this is done, PHL indicates the average hearing loss, in decibels, for
       those people subjected to severe health effects due to noise.
            Again, the above equations may be replaced by a summation over successive
       increments of day-night sound level.  It is recommended that increments of
       day-night sound level less than five decibels (i.e., preferably one or two
       decibels) be used in calculating values of PHL.
            Further, analogous to the assessment of general audible noise, the change
       in hearing loss-weighted population is a useful descriptor for many assessment
       purposes, as is the relative change in impact defined in Equation 8 with HWP
       substituted for IMP.

2.4  Environmental degradation

     Even in areas where no people are presently living, a significant increase

in noise over the existing conditions will constitute a noise impact.  The envi-

ronment may be degraded either because the increased noise affects wildlife

or monuments, or because it destroys the tranquility of a wilderness area to

which urban dwellers wish to go for an escape from city noise, or because it

makes the area unsuitable for future residential or other noise sensitive

development.  In each casef some of the quiet which is one of our national

natural resources is lost; the quality of the environment is lowered.

     Unfortunately, there are no data available which express this reduction

in environmental quality as a function of noise level, or of the change in

noise level.  Consequently, it is not possible to identify a special criterion

function for these areas, such as those identified in sections 2.2.1 and 2.3.1.

Instead, quantification of environmental degradation normally proceeds only

as far as the tabulation of the extent of the impact,  The only modifications

necessary for the standard tabulation (such as the example Tables 6 through

9) is the likely deletion of the columns on residential and employee popula-

tions, and a revision in the use of the special situations column.  Animal

species which are particularly vulnerable and recreational uses of the areas

will be the principal kinds of special situations to be listed.  As a supple-

ment to this numeric quantification, a word description of the environmental

impact should be provided in terms of the expected change from the present

conditions, paying particular attention to the special situations.  In some

circumstances, it may be useful to reduce this tabulation to a single number,

for example for comparisons or trade-offs with other planning criteria.  In

those cases, a 'level-weighted area1 can be calculated by using the population

weighting function of Table 3, which is the best available indicator of rela-
tive impact.
     Rural areas can be treated by the methods of either this section or sec-
tion 2.2.  That is, the analysis can stop with the tabulation of impacts, or
it can proceed to calculate the level-weighted population.  The equation used
(equation 2b) shows some adverse response to general audible noise at levels
as low as 45 dB (I^n).  However, because the percentage responding adversely
is so snail - less than 0.5% below 48 dB - and the number of people in most
rural areas is so low, the magnitude of the level-weighted population will
usually be so small as to be of little help in environmental assessment.
Although the single-number index can be used in such areas, it is not recom-
mended as strongly for them as it is for urban areas.

2.5  Treatment of temporary projects
     The major simplification in the analysis for temporary projects has
already been mentioned (section 1.4.2):  the fact that prediction of future
population in the affected area is unnecessary.  For temporary projects
lasting more than one year, that is the only modification necessary.
     For temporary projects, the same as for permanent noise environments,
the yearly day-night sound level should be used in computation of impact
indices.  Impact assessment is done in the same manner as for permanent noise
environments by the use of tabulations and calculation of the sound level-
weighted population and noise impact index.
     For temporary projects lasting less than a year, it is useful to compute
the level-weighted population for two situations:
     (1)  for the temporary noise environment as if it were permanent, but
          also stating its actual duration; and


     (2)  for the temporary noise environment in terms of its contribution
          to the yearly day-night sound level.
For example, consider a population of 1000 experiencing a temporary day-night
sound level of 70 decibels for nine months due to a construction project,
after which the day-night sound level drops to 60 decibels on a long-term
basis.  The following two situations would be described. .
     1.  During the nine-month construction period itself, the level-weighted
population is (0.245) (1000) = 245 persons responding adversely to the noise.
     2.  To calculate the effect of the construction activity on annual average
impact requires calculation of the yearly day-night average sound level:
                       1          70           60
     lOn(y) = 10 logio 12"  (9 x 10^) + (3 x 107"5")  = 68.9 decibels      Eqn. 14
On the basis of this I
2.6  Example application
     Sample tables to demonstrate the approach discussed in this chapter have
been drawn up for a simple example problem.  The example is based on the pro-
posed expansion of a highway which runs through a suburban area, and is simpli-
fied to facilitate understanding of the suggested procedures.  Details of the
example are contained in Appendix E, which also contains an additional example
application.  This section is intended primarily to provide samples of the
appropriate tables.  It does not cover all possible types of problems for
which these guidelines are appropriate.
     As discussed in section 2.2.2, a number of tables are usually helpful.
The first table documents affected areas (i.e., the base population and base
area) for future noise levels without the proposed project (Table 6); the
second deals with project noise alone (Table 7); and the third tabulates
effects for the project noise together with all other sources (Table 8).
The final table provides details of various special situations, which may
be particularly affected by noise (Table 9).  At the bottom, of Tables
6 to 8, several of the single number indexes are stated.  By comparing the
single-number indexes presented in Tables 6 and 8, we see that the anticipated
change in impact is an increase of 111 more people responding strongly to the
adverse effects noise has on them.  Likewise, the expected increase in
potentially severe health effects (Ldn 2. 75 dB) is in the range of 13 person-

                                           Table 6
                                  Sample Data Presentation:
                         Future Noise Levels Without  Proposed  Project
Yearly I85
Land Area
(sq km)
Land Area
(sq km)
Land Area
(sq km)
(See Table)
Level Weighted Population  (LWP) = 501 people
Noise Impact Index (Nil) =0.10
Hearing-loss Weighted Population (HWP) = 0

                                                   Table 7
                                          Sample Data Presentation:
                                    Future Noise Levels of Project Alone
Yearly L^n
Land Area
(sq km)
Land Area
(sq km)
0.. 161 25
Land Area
(sq km)
(See Table)
       Level Weighted Population (LWP)  = 362 people
       Noise Impact Index (Nil)  = 0.07
       Hearing-Loss Weighted Population (HWP) = 13 person-decibels
       Average  Potential  Hearing Loss (PHL)  = 0.16 dB per person for 83 people

                                                   Table 8
                                          Sample Data Presentation:
                                    Future Levels from all Sources Combined
Yearly L^n
Land Area
(sq km)
Land Area
(sq km)
Land Area
(sq km)
(See Table)

        Level Weighted Population (LWP)  =612 people
        Noise Impact Index (Nil)  = 0.12
        Hearing-Loss Weighted Population (HWP) = 13 person-decibels
        Average Potential Hearing Loss (PHL)  = 0.16 dB per person for 83 people

                                  Table 9
                         Sample Data Presentation:
                            Special Situations
                   Average Population
                    Day         Night
            Area (sq km)
1.  School          300

2.  Playground       40

3.  Park             30

4.  Church           63

5.  Nursing Home    200

6.  School         1000

7.  Library          25

8.  School          500





Night Classes

                                  Chapter 3
                                Special Noises

     Not all noises can be adequately evaluated by average A-weighted sound
levels.  Examples of the special noises which require other measurement sys-
tems are the following:  (1) infrasound, in the frequency range of 0.1 to 20
Hz; (2) ultrasound, frequency range above 20 kHz; (3) certain types of
impulse noises such as sonic booms and blasts; and (4) sounds that convey
more information than random noise sources with comparable average sound
levelsF such as voices, warning signals, or barking dogs.  This chapter
contains a section discussing each of these four special noises.  For the
first three, the section discusses measurement, screening levels, and human
effects.  For the fourth, the section merely provides a brief description of
the nature of the problem and of how it might be treated in a noise impact

3.1  High energy impulse noise
     The assessment of impulse noise presents unusual problems.  In many cases
the appropriate techniques and measures are applicable only to particular
situations.  (For example, with respect to blast noise, damage to certain
types of buildings can be predicted in terms of non-acoustic parameters, such
as effective distance and the amount of explosive charge.)  Moreover, the sig-
nificance of the noise impact cannot always be quantified for the same effects
suggested for general audible noises.  Whereas low-level impulse noise is
accounted for as part of normal general audible noise, high energy impulses
require additional measurements for impact assessment.  In many situations an
individual interpretation of the criteria is required.


     At present, high energy impulse noise comes primarily from sonic booms,
blasting operations, or artillery fire.  Some limited community response data
for sonic booms and artillery fire are available [30, 31].  Noise measurement
instrumentation at the time of the sonic boom study (1965) was not as sophis-
ticated as it is now, so the physical measures from that study (peak overpres-
sure in pounds per square foot) need to be converted to more recently developed
measures.  Tine conversion factors appear to be different for artillery and
blast noise as opposed to sonic booms.  Consequently, the methods presented
in this section need to be verified with more data, some of which is being
collected at the time of this writing.  The methods presented here are based
on the only available data, and should be applied with some caution.

3.1.1  Description of high energy impulse sounds
     Day-night sound level is the primary descriptor for environmental noise.
High-energy impulse sounds, such as those produced by sonic booms, quarry
blasts, or artillery fire, in addition to the high-level audible sound,
can excite noticeable vibration of buildings and other structures.  These
induced vibrations — caused by airborne sound or transmitted through the ground
or structures — may generate additional annoyance beyond that due to simple
audibility of the impulse, because of "house rattling11 and "startle," as well
as because of additional contributions to interference with speech or sleep.
The annoyance data which are used in the next section to summarize community
response to impulse noise are based on the annoyance caused by house rattle.
     It has been general practice in the past to describe such high-energy
impulse sounds in terms of the peak sound pressure in a wide frequency band.
While the peak pressure may be satisfactory for assessment of impulses in a
restricted range of peak pressures and durations, it is not sufficient as


a general descriptor for use in measurement or prediction of the combined
environmental effects of impulses having substantially different pressure-time
characteristics.  Use of the peak pressure is also unwieldly or misleading
when a succession of impulses, sometimes overlapping, must be evaluated.
     Two noise measures, similar to the measure used for general audible noise,
are recommended in these guidelines for assessing the environmental impact of
high-energy impulse noise.  The choice between these two measures is based
primarily on the number of impulses occurring during a day.  If only isolated
high-energy impulse noises occur, the C-weighted sound exposure level, symbol-
ized as LSO should be used.  If there are more than perhaps eight occurrences
per day, it will be helpful to compute a C-weighted day-night sound level,
Icon, derived solely from the individual Lgc values.
     There are two reasons for using a C-weighting.  First, it does not dis-
count the low frequency components which are a major part of impulse noise and
of vibration, as the A-weighting network does.  Second, subjective estimates
of blast noise magnitude conform with magnitude estimates of other noises
when the blast is measured by C-weighting and the other noises are measured by
A-weighting [32].  In general, C-weighting has been found to closely relate to
average human response to high-energy impulse noise [33].
     The use of sound exposure level is recommended to facilitate combination
of data when more than one impulse noise event occurs per day, as is usually
the case.  Further, it is consistent with subjective evaluations of sonic
booms where duration of the signal influences subjective response [34].
     The assessment procedures suggested in this section should be used for
impulse sounds that have daytime C-weighted sound exposure levels greater
than about 80 dB.  This corresponds to unweighted peak pressures for sonic

booms greater than about 106 dB, which appear to be the threshold of adverse

community response on the basis of the data on sonic booms.  This in turn

corresponds to unweighted peak pressures for quarry blasts and artillery fire

of about 100 dB.  At night, the threshold of response should be reduced to a

Oweighted sound exposure level of 70 dB (corresponding to unweighted peak

pressure levels of 96 dB and 90 dB for sonic booms and blasts, respectively),

because of the decreased acceptability of nighttime impulsive exposures  [30,

p. 150].*  Impulse events with smaller peak pressure than described above are

assumed to elicit normal auditory responses and are assumed for most situa-

tions to be described adequately by 1^.  For impulses with peak pressures

greater than 140 dB, assessment criteria based on actual physiological or

structural damage should also be applied.  In addition, the effects of

groundborne vibration should be assessed (Chapter 4).
     In most cases where impulse noise needs to be considered, the task for

the noise impact analysis is to predict or estimate in advance what the levels

of impulse noise will be.  With rare exceptions (e.g., reference 35), there

are no reliable predictive techniques other than using a measurement of a

similar event occurring elsewhere.  When the only data available are

expressed as peak pressure, useful approximations can be based on indications

that Lgc is roughly 26 dB lower than the peak pressure for sonic booms  [36],

and 20 dB lower for blast noise [37, Fig. 29].  In those cases where it is

possible to conduct measurements of a similar event elsewhere, it is important
*For situations which are characterized by less than about eight impulse
 occurrences per day, and where LJJC is computed rather than Icdm threshold
 values 10 dB higher (90 dB daytime and 80 dB nighttime) may be used.   Bow-
 ever, the more stringent (80 and 70 dB) threshold values are recommended  for
 situations where there are many (greater than about eight) high-energy impulse
 events per day, and where Icdn is calculated, in order to ensure that
 relatively low level L^dn cases are not disregarded.

to be able to distinguish impulse noise  (such as sonic boom) from other high-
energy noise events  (such as jet aircraft flyovers).  A useful rule of thumb
to aid in making such a distinction is that for an impulse noise the maximum
C-weighted sound exposure level in any 2-second time period is 10 dB greater
than the C-weighted  sound exposure level in any contiguous 2-second period of
the event.

3.1.2  Human noise exposure effects of high-energy impulse noise
     The Oklahoma City sonic boom study  [30] is the artillery fire study  [31]
form the primary bases for the procedure proposed for assessment of the effects
of high-energy impulse sounds.  In the sonic boom study  [30], eight supersonic
overflights were performed on a daily basis for six months.  Altitudes and
airspeeds were selected to obtain three  different nominal overpressures, on an
increasing basis, during the tests.  Personal interviews of respondents were
made during three time periods that corresponded to the three different nomi-
nal overpressures.   Interviews were conducted at three different distances
from the ground projection of the flight path to obtain different exposures
for each of the three boom levels.
     The questionnaire structure and response scaling used in the sonic boom
social survey are such that direct comparison with other surveys is difficult.
The responses to a question on the degree of annoyance due to "house rattles"
caused by the booms  is used here as the primary measure to quantify community
response.  The category "serious" annoyance is considered to be most compar-
able to the highly annoyed response used (in section 2.2.1) to summarize the
adverse effects of general audible noise.  It should be noted that the percent
of respondents reportings serious annoyance at different boom levels was not
a percentage of the  total population sample, but only of that fraction of the


sample that believed it appropriate to complain about governmental actions.
This fraction is of the order of 60 percent.  To compare these responses to
the total populations used in other surveys, an adjustment for the total pop-
ulation was made in the current analysis.
     The noise measurements in the sonic boom study were collected in terms
of nominal peak overpressures.  Conversion of nominal overpressures to C-
weighted sound exposure levels were made using the average difference of 26
decibels between peak overpressure, in decibels, and C-weighted sound expo-
sure level.  The resulting values were then used to compute L^dn for tne eight
daytime sonic booms, using the approximation:
          lojn = Isc + 10 log (Ifc + 10 Nn) - 49.4                 Bqn. 16
where N^ and Nn represent the number of impulse events during the day and
night, respectively.  Thus for eight sonic booms per day, equation 16 reduces
                             ICdn = tGC - 40-5                    Ban- 17
The resulting data for the percent serious annoyance at the computed C-weighted
day-night sound level values are plotted as filled-in squares in Figure 9.
     In the artillery fire study [31], groups of residents were interviewed  at
nine sites in the vicinity of an Army base where extensive artillery firing
training takes place.  Six of the sites that were off-base were considered
here.  Noise monitoring on a 24-hour basis took place at 17 locations for an
average of approximately 25 days per site.  These measured average sound
levels in conjunction with computer based predictive models, were used to ob-
tain annual average C-weighted day-night average sound levels for blast noise
associated with the environments in which the survey respondents lived.  The
social survey used scales similar to other recent surveys.  The percent of


                Sonic Boom
                                                                             " ^n 'For ' Eventil
       i 90
                                              . 105
                                                                          .Peak Preouib
      110          115          120          125


 1    (Blan&
135   Artillery)

respondents reporting high annoyance are plotted as filled-in circles on Fig-
ure 9.
     Using annoyance data from both the Oklahoma City sonic boom study and
the artillery fire study, a function is plotted in Figure 9 which shows a
reasonably good fit of projected high annoyance against C-weighted day-night
sound levels.  For use in situations where there are less than approximatley
eight booms or blasts per day, a separate abscissa for C-weighted sound expo-
sure level is also included in Figure 9.  Further, because sonic boom and blast
noise data are oftentimes collected in terms of peak pressure, additional ab-
scissas are also included in Figure 9.
     Over the range of data available, the function illustrated in Figure 9
provides a reasonably good prediction of the percentage of the population who
can be expected to be highly annoyed at given exposures to high-energy impulse
noise.  Consequently, it is proposed to use the function shown in Figure 9
and presented in equation 18 below* for the assessment of high-energy impulse
noise, despite the fact that such applications may need to extrapolate beyond
existing data.
                      %HA	10°
                                         - 0.153 I)              E°n-  18
     Quantification of adverse human response anticipated from high energy
impulse noise is performed in the same manner as for general audible noise
(Section 2.2.2).  The appropriate weighting function describing the population
exposed to high energy impulses who are highly annoyed with the noise may be
computed from equation 18, or read from Figure 9 or Table 10.  Level weighted
*Note that the format of equation 18 is similar to that footnoted on page 37
in section 2.2.1.

                         Table  10

Values of Weighting Function for High Energy Impulse Noise

                           = (0.01)%  HA]
                45               0.014
                46               0.016
                47               0.018
                48               0.021
                49               0.025
                50               0.029
                51               0.033
                52               0.039
                53               0.045
                54               0.052
                55               0.060
                56               0.069
                57               0.080
                58               0.091
                59               0.105
                60               0.120
                61               0.137
                62               0.157
                63               0.178
                64               0.201
                65               0.227
                66               0.255
                67               0.285
                68               0.317
                69               0.351
                70               0.387
                71               0.424
                72               0.462
                73               0.500
                74               0.538
                75               0.576

population may then be computed from equation 6.  Likewise, Noise Impact

Index and Relative Change in Population may be calculated from equations 7

and 8, respectively.

     In many situations, both iitpulse noise (measured in loan) and general

audible noise (measured in LA^) will be of concern, and it will be necessary

somehow to combine the results to obtain a total number of people affected

by all aspects of the noise.  An assessment of the overall noise environment,

combining the effects of high energy impulse sounds and of general audible

noise, can be made by equating the degree of annoyance expected from the two

types of noise sources.  Using Figures 4 and 9, then, it is possible to

identify a specific C-weighted L^ which causes as much annoyance as an

A-weighted day-night sound level.  For example, an LQfo of 65 && is expected

to result in 23% of the exposed population being highly annoyed by the noise

(Figure 9).  This same level of annoyance is reached at an A-weighted day-night

sound level of 69 dB (Figure 4).  Thus the Lean may be converted to L^n

via equal annoyance (Table 11).  This converted L^n is added, logarithmical-

ly, to the general audible noise already measured in terms of L^, and the

resulting composite noise level is used for assessment of the overall noise

environment, using Figure 4 and Table 3 as necessary.  This procedure is

performed in order to avoid the double-counting of affected people which

could result if they were tallied separately for impulse noise and for

general audible noise.

3.1.3  Structural damage criteria for impulse noise

     It is normally considered that the most sensitive parts of a structure

to airborne noise or overpressure are the structure's windows, although in

some cases it may be plastered walls or ceilings.  Such noise or large pressure


                  Table 11
Conversion of I<^n to L^n Via Equal Annoyance
            * Highly Annoyed
. . 	 52_ . . .....
' 59
an _ . 	

waves also introduce building vibration in addition to that due to ground
notion.  Thus the effects of airborne sound on structures may need to  be eval-
uated in terms of vibration criteria as well as in terms of criteria based on
peak overpressure.  For most airborne sound, however, evaluation of the peak
overpressure is sufficient to determine the threshold of possible damage.  On
the other hand, for some types of underground blasting and when the building
is close to the blast site, the vibration is transmitted essentially through
the ground.  In this case the vibration inside the house must be predicted
and evaluated according to the vibration criteria (Chap. 4).  This subsection
describes structural damage criteria for three kinds of impulse noise:  blast
noise; sonic boom; and artillery fire.  A brief paragraph is appended  relating
to structural damage from continuous sounds.
     For blast noises, the probability of broken windowpanes should be esti-
mated.  Empirical formulas given below allow an estimate of "safe" distances
from the blast, beyond which window damage is negligible.  These formulas
include sufficient safety factors to account for the negative influence of
such variables as wind direction, atmospheric temperature gradients, and win-
dowpane shape and size.  These formulas are newly proposed and are somewhat
tentative  [38].  They are suggested here essentially as screening tools:  if
these equations suggest there will be no structural problems for a particular
project, the impact analysis needs to proceed no further.  If these formulas
suggest a potential impact of blast noise on structures, then the analyst (or
blasting engineer) should undertake a more detailed analysis which involves
explicit consideration of the variables covered by a safety factor in  these
formulas.  In addition, a monitoring program might be recommended to identify
any damage, or lack of it, actually caused by an explosion.

     For surface explosions, window breakage in residential type structures is
expected to be negligible (less than 50% probability of even one broken pane)
if the charge per delay equivalent weight  of high explosive (WHE) in kilograms
is less than that specified by the more appropriate of the following two
     (1)  If /-the population is non-uniformly distributed, but is clustered,
          then each population cluster, including the nearest residence,
          should be checked.  The amount of WHE for any cluster should be
          less than 328 R /M where R is the distance in kilometers from the
          explosion to the center of a cluster of residences and N is the
          number of people residing in that cluster with the provison that N
          must always be at least 4.
     (2)  If the population is reasonably uniformly distributed, then the
          amount of WHE should be less than 40 R , where R is the distance
          in kilometers to the nearest residence.
The use of these formulas requires some judgment as to what constitutes a pop-
ulation cluster and what constitutes a reasonably uniform distribution.  In
some cases, both formulas might be checked and the one that predicts the least
allowable amount of WHE used.
     For explosives buried deeper than 1.4 meter per (Kg)   , the peak ampli-
tude will be attenuated by at least a factor of 5.  For such underground ex-
plosions the preceding formulas need to be adjusted as follows:
     (1)  Population clusters - the amount of WHE should be less than 26430 R /N.
     (2)  Uniformly distributed population - the amount of WHE should be less
          than 3200 R3.
"Weight per detonation where each detonation is delayed to go off in a
predetermined sequence (usually within a fraction of one second) for each
event.  The duration of the total event is normally less than one second.

            For explosive charges greater than those determined by the above formulas,
       the peak overpressure should be predicted and the number of broken windows
       estimated.  The statistical estimator (Q) for the number of "average typical"
       panes broken is:
                 Q = 1.56 x 10~10 N(PK*)2*78                                   Eqn. 17
..i     where N = number of people exposed (assuming 19 panes per person) and PK* is
           peak-to-peak amplitude of the pressure variation (in pascals) at ground
level.  For convenience of measurement, the peak-to-peak pressure amplitude
reflected at ground level (PK*) may be used.  The conversion between the peak
free air pressure (Ap) and PK* is given by the relation:
          PK* = 2.7 AP                                                 Eqn. 18
However, the peak pressure may be amplified by a factor of 5 as the result of
refraction, ducting, and focusing; therefore, in the "worst case" condition
                                                                           2 73
the number of broken panes, Q, may be multiplied by a factor as high as (5) *
or 88 to obtain Qmax-  Atmospheric meteorological effects can increase this
factor further.  In addition, for peak pressures (AP) above 140 dB (200 Pa),
structural damage other than window damage may occur.  Measurement or predic-
tion of vibration should be accomplished.
     For sonic boom and artillery fire, the amount of window damage can be esti-
mated by calculating Q and CL..  for the expected peak pressure, as discussed
for blasts.  These formulas, however, should be used only for peak pressure
levels above 130 dB.  Above 140 dB, structural damage should also be assessed
by prediction or measurement of vibration levels in the exposed structures.
     For continuous sounds above sound pressure levels of 130 dB, there is the
possibility of structural damage due to excitation of structural resonances
for infrasound, as well as low and medium frequency sound.  While certain

frequencies  (such as 30 Hz for window breakage) might be of more concern than
other frequencies, one may conservatively consider all sound lasting more than
1 sec above a sound pressure level of 130 dB  (1 Hz to 1000 Hz) as potentially
damaging to structures.

3.2  Infrasound
3.2.1  Description of infrasound
     Infrasound is defined as sound  in the frequency range from 0.1 to 20 Hz.
The measurement of infrasound should be made with instrumentation having a flat
frequency response (+3 dB) from 0.1  Hz to 1000 Hz.  The reason for the extended
measurement range is that in evaluating a noise that is composed of both infra-
sound and higher frequency sound, the higher frequency sound must also be mea-
sured for proper assessment of the infrasound, because sounds above 20 Hz can
mask the infrasonic sounds.
     Although blasting operations cause infrasound as well as impulse noise
and vibration, it is not intended that all of these analyses be conducted.
Among other considerations, the necessary instrumentation is different for
each of these special noises.  For blasting, an impulse noise evaluation is
adequate, covering both human and structural effects.  Because infrasound can
be related to vibration, the vibration analysis (Chap. 4) also helps reduce
any need for a special infrasound analysis.

3.2.2  Human noise exposure effects  of infrasound
     On the basis of a summary of infrasound effects (Figure 10), compiled from
the Levels document [2] and more recent work  [39, 40], it is suggested that for
exposures of less than 1 minute the  maximum sound pressure level should be below
the following values:

                                            SOUND PRESSURE LEVEL BELOW WHICH
                                            INFRASOUND IS NOT EXPECTED TO
                                            PRODUCE ANY ADVERSE PHYSIOLOGICAL
                 CURVE B1
                               CURVE B2
                               LOUDNESS OF
                               45 PHON
                      RANGE OF HEARING THRESHOLD
          0.1 HZ   0.2
0.5    1HZ     2        5     10HZ   20HZ



          0.1 Hz to 5 Hz ... 120 dB
          5 Hz to 20 Hz . . . 120 dB - 30 log |                     Eqn. 19
where £ is the dominant frequency.  For exposures longer than 1 minute and less   >
than 100 minutes, the levels should be reduced by (10 log t) dB where t is       £,
time of exposure in minutes.  Exposure longer than 100 minutes should use the
100 minute limits.  In other words, exposures 20 dB less than the one minute
criterion should be regarded as having no impact, regardless of exposure time.
The 100 minute criterion basically insures that the infrasound is inaudible.
These levels serve essentially as screening levels.  As long as they are not
exceeded, infrasound does not need to be included in the noise analysis.
     For evaluating the impact, if this screening criterion is exceeded, a
single-number index is not suitable.  Instead, the impact should be qualita-
tively described; the effects that might occur at different sound levels are
given in Figure 10.  Any assessment of the effects beyond those in Figure 10
is not contained in these guidelines and will require further research and

3.3  Ultrasound
3.3.1  Description of ultrasound
     Ultrasound is defined as sound at frequencies between 20 kHz and 100 kHz.
Seldom is ultrasound an environmental problem and, unless the level is expected
to exceed 105 dB, it can be ignored in an environmental noise analysis.
     Measurement of ultrasound should be accomplished by instrumentation with
flat response (+_ 3 dB) from 10 kHz to 100 kHz.
3.3.2  Human effects of ultrasound
     Ultrasound noise levels below 105 dB (frequencies above 20 kHz) are con-
sidered to have no significant impact on people.  Noise levels above 105


decibels should be reported in the analysis and individually evaluated based
on specific research studies.  In particular, studies of effects on animals
may be important.  No further quantification of the environmental impact of
ultrasound is recommended.  Barely is ultrasound (except for some occupational
situations, e.g., ultrasonic cleaners) an environmental problem of practical
interest.  Evaluation of ultrasound exposure above 105 dB requires additional
investigation and research.

3.4  Noises with information content
     Some general audible noises are also more annoying than their level alone
would indicate, due in part to their information content or clear detectability.
Examples include barking dogs and back-up alarms, but the primary problem is
voice communication (live, amplified or recorded) that crosses residential
boundaries at high levels.  There is no formal method for assessing the
impact of such sounds; each case must be assessed on its particular merits.
It is recommended, however, that the analyst mentions how, as a result of the
proposed action, the intrusion of understandable voices into some area might
cause loss of privacy and consequent undesirable effects.  The actual content
of the typical messages or words might be stated along with the number of
people that are impacted.

                                  Chapter 4

     This chapter contains procedures for evaluating the impact of vibration
on nan.  While the main reason for their inclusion here is to account for
vibration generated by airbone noise, the impact of certain types of vibra-
tion can be assessed whether the transmission paths are airborne or structure-
borne.  The two sections of this chapter deal with the human effects of
vibration and the structural effects of vibration.
     The material in the first section is based on an approved ISO standard
and its proposed amendments [41 ], and its United States Counterpart [42].
These are summarized in Appendix D, to provide the necessary background to
follow the recommendations in section 4.1. The recommendations in section 4.2
are based on consideration of that material and data contained in Bureau of
Mines Bulletin 656 [43].

4.1  Human effects of vibration
     Vibration of structures may be due to airborne acoustical waves or solid-
borne vibration.  Most problems caused by airborne impulse noise, when build-
ing vibrations are caused as a side effect of the primary auditory stimulus,
should be accounted for by the procedures of section 3.1.  Nevertheless, at
certain times it may be necessary to assess separately the vibration caused by
such sources.  Groundborne vibration which is quite likely to accompany some
mining, construction, and other industrial activities usually requires special
evaluation.  A method to evaluate human response to vibration inside buildings
is presented which should be used to evaluate the impact of such activities.
The method applies to the frequency range between 1 Hz and 80 Hz.

4.1.1  Description of building vibration
     For continuous vibration environments, rms acceleration should be measured
along three orthogonal axes, one axis of which is normal to the surface being
measured.  The acceleration will be weighted to account for the dependence
of human reaction on frequency by use of a low pass filter with a corner
frequency of 5.6 Hz (Figure 11).  This accounts for the fact that human sensi-
tivity to acceleration decreases over the frequency range under considera-
tion; above 10 Hz this decrease is approximately proportional to frequency.
The assessment of the impact should be against greatest acceleration on any
of the three axes used.
     For building measurements to be appropriate for the criteria of the next
subsection, the measurements should be taken on the floor at a point that has
the maximum amplitude of all the reasonable points of entry of the vibration
to the human occupants.  Normally this point may be assumed to be at the
mid—span or center of a room.
     For impulsive shock the measurement should be the same as for the
continuous vibration measurement, except that the peak acceleration, not
the rms value, should be used.  The duration for impulsive shock excitation
will be determined by either the time the acceleration of an event exceeds
0.01 m/sec2 or by the time the acceleration is within one-tenth of the peak
value.  ftoichever gives the snorter duration should be used.

4.1.2  Human vibration exposure criteria
     Threshold levels are presented in Table 12 for most types of structures.
Not all types of buildings are classified, but common sense should suggest
the most appropriate classification.
     The overall vibration that will not cause an adverse impact for any con-
dition and time period corresponds to rms acceleration values below

           ATTENUATION (dB) = 20 log >/1 + (f/5.6}2
                                          15  20
30  40  50 60  80
                              FREQUENCY (Hz)


                   FOR THE FREQUENCY RANGE 1 TO 80 Hz.
                   Note: Electrical network for low frequency cutoff below
                        1 Hz and high frequency cutoff above 80 Hz not yet

3.6 x 10~3m/s2, evaluated by means of the weighting described in Figure
11.  Ebr hospital operating areas and other such critical areas, no higher
levels should be permitted without analysis and justification of the
acceptability of such levels.
     Ebr residential and other similar areas,  continuous acceleration of
greater values are normally expected to cause  virtually no complaints (less
than 1%).  Even greater acceleration values could be permitted for shorter
times during the daytime (0700 to 2200 hours), as indicated by Table 12 and by
Figure 12.  These also indicate that the maximum value of the impulsive shock
excitation that is expected to cause virtually no complaints can be raised,
dependent on the number of such impulses during the daytime.  Ebr residential
areas or other areas where people sleep, the nighttime peak acceleration should
be less than 0.01 m/sec2 at any time and the continuous rms acceleration should
be below 0.005 m/sec2 if no complaints are to  occur.  No differentiation is
made as to the types of residential areas, i.e., city center, urban or rural.

                                  Table 12
  Basic Threshold Acceleration Values* for Acceptable Vibration Environments
                                  Continuous or          Impulsive Shock
                        Time      Intermittent rms       Excitation Peak
Type of Place	of Day    Acceleration (m/sec2)  Acceleration (m/sec2)
Hospital Operating
Booms and Other Such
Critical Areas
Factory and Workshop
*Weighted as shown in Figure 11.
t = duration seconds of vibration, for durations greater than 100 sec, use
    t as 100 sec.
N s is the number of discrete shock excitations that are one sec or less in
    duration.  Ebr more than 100 excitations, use N = 100.
Daytime is 7 am to 10 pm.  Nighttime is 10 pm to 7 am

|_ 1 1 1 2. B § b g
~ 1 Threshold of risk of damage to normal dwelling-houses with plastered
.X \ ceiling and walls
*^x**x Threshold of risk
- ^~<>
of damage to sensitive structures \
^ , ,
^ x Daytime peak impulses (complaints <20%)>^
\ Daytime rms (complaints <20%K^
** Daytime peak impulses ; (complaints <1%) *-*^
^"V^ Daytime rms (complaints o%) -^^
Nighttime rms (complaints <1%) • — ^
I I 1

                                                                                      0.01 m/sec2 „
                                                                                    J   SEC OR NR. OF
                                NUMBER OF IMPULSES PER DAY OR EXPOSURE TIME

     For office type spaces, the threshold at which no adverse effects occur
is twice the daytime residential rms or peak value.  No distinction is made
between daytime and nighttime exposure.
     For factory and similar type spaces, the threshold at which no effects
occur is 4 times the daytime residential values.  No distinction is made
between daytime and nighttime exposure.
     Offices and workplaces may in many cases require vibration levels as
low as residential areas if any adverse reactions are to be avoided.  In
certain critical areas, such as operating rooms and laboratories and possibly
research laboratories, standards rooms, tool rooms and the like, even lower
vibration exposures levels may be required than indicated by Table 12.
     The acceleration values that are specified to cause less than 1% com-
plaints are near or at the perception threshold level of vibration during
normal activity and should serve as a realistic threshold of any adverse
reaction to the vibration.  The percentage of complaints likely to occur for
higher levels of vibration are shown in Figure 13, which summarizes the
complaint history from the Salmon Nuclear Event [43].  For a single event
the number of complaints for residential areas varies roughly as 10 log K
(for peak acceleration range of 0.1 m/sec2 to 1.0 m/sec2), where K is
the ratio of the observed acceleration to 0.1 m/sec2.

4.1.3  Quantification of the impact
     There is a lack of data related to the assessment of the severity of
the impact that results if the vibration guidelines proposed in this
section are exceeded.  It is recoranended that the number of people exposed
to vibration levels above the "no complaint" value (Table 12) be estimated.

                           5           10           15
                                  PERCENTAGE COMPLAINTS
Figure-13. Percentage of Population Complaining as a Function of Peak Acceleration (Source: Reference 43)


For a specific action, therefore, contours of the appropriate "no complaint"

acceleration values as determined by Table 12 should be predicted or measured.

For example, if an action causes a steady vibration that lasts a total of

25 seconds a day (during daytime hours), the contour of 0.014 m/sec^ should

be evaluated (0.072/ 25 = 0.014).

     In addition to the mapping and tabulation of the impact, which cover

sensitive non-residential as well as residential buildings, single-number

indexes can be calculated which are similar to those suggested for general

audible noise (the level-weighted population and hearing-weighted population).

These indexes are based on the relationship for the percent complaining,

documented in Figure 13.  It is suggested that this concept be tentatively

broadened to apply the vibration exposure to more than one impulse or to

intermittent/continuous exposures by using the ratio (k) of the actual

acceleration to the recommended "no complaint" acceleration value.  A term

for the impact of vibration on residential areas can then be defined by using

a vibration weighting function.  This function is described by:

                 V(k) = 20 log k                                     Eqn. 20

where k is the ratio of the actual acceleration to the recommended "no

complaint" acceleration values listed in Table 12 for a specified time

period, and where k is limited to values from 1 to 20.

     This function can be used to calculate a descriptor of the total

vibration impact of a project, by multiplying the number of people

exposed to each vibration condition by the vibration weighting function

for that condition, and then finding the sum of these products.  This

Vibration-Weighted Population (VWP) is defined as:

                VWP =  /P(k) V(k) dk                            Eqn. 21


where V(k) is the vibration weighting function described above, P(k) is the
population distribution function, and dk is the differential change in k.  An
index, similar to the Noise Impact Index, but applied to vibration, is called
the Vibration Impact Index (VII) and is calculated as:                         \/ll
                      /* P(k) V(k) dk
                VII = -!	                            Eqn. 22
                         / P(k) dk
where the denominator is based on the alternative affecting the largest
number of people.  In other words, the base population for calculating the
vibration impact index needs to be constant across alternatives for the
number to be meaningful.  Given that restriction, then changes in VWP and
VII can both be used to evaluate various alternatives and actions with
respect to vibration.  The change can also be discussed by listing the
expected effects at the nearest residence.

4.2  Structural effects of vibration
     A structural vibration velocity of 2 in/sec has commonly been used as
the safe limit, and certainly vibrations above this value will have a
very adverse environmental impact.  Note that, except for frequencies below
3 Hz, if the acceleration measured with the weighting network of Figure 11
is less than 1 m/sec2, then the velocity will be 2 in/sec or less.  For
frequencies from 10 Hz to 80 Hz a weighted acceleration of 1 m/sec2 is
essentially equivalent to a velocity of 1 in/sec.  In most practical cases,
in which the acceleration is made up of several frequency components, an
acceleration of less than 1 m/sec2 will also mean that the resultant velocity
will be less than 2 in/sec, and possibly less than 1 in/sec, regardless of

frequency.  Therefore, it is recommended that 1 m/sec2 be used as the normally
safe acceleration with respect to structural damage.  Vibrations above this
should be avoided, or special arrangements should be made with the owners
of the exposed structures.  Since some minor damage has occasionally been
reported at vibration as low as 1 in/sec, (0.5 m/sec2 to 1 m/sec2), exposures
in the range between 0.5 m/sec2 and 1 m/sec2 should also be regarded as a
potentially adverse exposure with respect to structural damage.  Finally, the
safe peak acceleration for ancient monuments or ruins should be considered
as 0.05 m/sec2.  Higher exposure values for such ancient structures should
not be considered safe without a detailed structural analysis.
     No single-number index is suggested for summarizing the structural
effects.  Quantification of the impact will consist of a contour map and
tabulation, showing the number of structures above the potentially damaging
accelerations of 1 m/sec2 and 0.5 m/sec2.  A description of the expected
damage and the likelihood of such damage occurring should be provided for
each type of structure.  The information in Appendix D will be of some help
in making this assessment, but sufficient data will not often be available to
make this assessment fully.  In such cases, a program for monitoring the
actual damage, or lack of it, may be necessary.

                                  Chapter 5
                       Sumnary Of Noise Impact Analysis

     This chapter provides an overview of the analysis that might be expected
to characterize noise impact fully, by summarizing each of the four preceding
chapters.  In addition, Figure 1 and Table 1 provide useful overviews of the
kinds of analyses suggested.
5.1  Purpose and structure of the guidelines
     These guidelines contain procedures which can be used to describe and
quantify the noise related impacts of proposed projects.  The resulting
description of noise impacts is intended to be easily understood by those
making decisions, so that consideration of these impacts can be an integral
part of the decision.  The approach described here is applicable to any situa-
tion calling for the evaluation of noise-related impacts, such as EIS or envi-
ronmental assessment preparation for the NEPA process, and is consistent with
noise evaluation procedures used by EAA, EHWA, and HUD, among others.  The
approach is not mandatory, but is meant to complement these other procedures
by showing how to proceed to a quantitative description of impacts on people
(which all procedures desire) from information on noise levels (which those
procedures require).
     These guidelines provide procedures for arriving at verbal, tabular, and
single number descriptions of noise environments.  The quantitative approaches
rely on tables detailing the affected area or population, and on a modifica-
tion of the earlier fractional impact method [44] to reduce the tabulated
information to a single number index.  These descriptions should be applied
to future as well as to immediate impacts.

     Three principal types of noise environments are covered:  general audible
noise; special noises; and vibration.  There is a separate chapter for each,
which covers (1) the appropriate physical measurement, (2) methods for predict-
ing that measure for the proposed project and for determining the existing
levels, (3) human noise exposure criteria, and (4) procedures for quantifying
impact.  Within the chapter on general audible noise, three subsections are
provided, entailing different approaches for human exposure criteria and
quantification procedures in different noise ranges:  urban and suburban
settings (L,  usually 55 to 75 dB); projects producing greater than 75 dB
(L, ); and rural and wilderness areas (L,  usually less than 55 dB).
     Additional types of proposed actions for which these guidelines will be
useful are projects which entail new populations to be introduced into noisy
areas, and actions which are intended to reduce noise.  The impact of temporary
projects may be evaluated using a more simplified analysis.  For all impact
analysis, the necessary estimation and prediction entail uncertainty.  When
possible, the degree of uncertainty should be specified.  In some circumstances,
optimistic and pessimistic forecasts can be used to bracket the estimate.
5.2  Analysis of impacts of general audible noise
     General audible noise is noise as commonly encountered in the environ-
ment.  Therefore, the material in this chapter should cover the great major-
ity of situations in which an evaluation of noise impacts is desired.  The
primary measure of general audible noise is L, , and whenever possible,
an approximation to the annual average value should be used.  In some
instances this measure is inappropriate, and shorter term measures such as
1-hour L   or the sound exposure level should be used.  The screening diagram
(Figure 3) shows that whenever the noise level after the project will be greater

than the existing level a noise analysis  is necessary  (i.e., when the existing
level is less than 10 dB greater than the project noise level).  The diagram
applies to both permanent and temporary projects.
     Depending on the approximate range of L^ values, different types of
noise effects are of concern, and therefore different analyses are needed
(Figure 14).  At levels generally encountered in populated areas (approx. L,
values of 45 dB to 80 dB), the general health and welfare effects of noise
are the primary concern.  At levels above 75 dB (8-hour L   at-ear) severe
health effects become important.  The threshold level at which these should
be investigated is an L,  of 75 dB.  In rural or wilderness areas, with
very low residential populations, environmental degradation is as much a
concern as the effect of noise on residents.  In such areas, judgment will
have to be used in deciding between a health and welfare analysis and an
environmental degradation analysis, depending more on characteristics of the
area than on the existing or project noise level.
     Regardless of which noise effects are the focus, two elements are always
recommended for describing the impact.  The first is a table (or set of tables)
setting out the number of people and total area affected as a function of
different noise levels.  Five decibels is usually an appropriate interval to
use for those tables.  The second element is a verbal, qualitative description
of the principal components of the impacts identified in the tables.
     For general health and welfare effects and for severe health effects
the quantitative analysis can proceed further, to calculate a single-number
index which summarizes all the impacts.   The human noise effects information
discussed in the Levels Document apply in the general health and welfare
effects range:  speech interference; sleep interruption; annoyance; and

                  SEVERE HEALTH EFFECTS (SEC. 2.3) VjS&X
                                                HEALTH AND
                                             WELFARE EFFECTS
          EFFECTS (SEC. 2.2)
   (SEC. 2.4)
                        J5Q       70
                       EXISTING Ldn(y)

possible health effects.  Given the existence of Schultz's synthesis  [5], it
is appropriate to use the percent of people who report being highly annoyed
as the indicator of general adverse response, and to use his equation to
summarize the total impact of noise on residential areas in terns of the
number of people responding adversely to the noise.  In the severe health
effects range, the human noise exposure effects may include cardiovascular
effects and other stress related health problems.  It is not known yet at
what levels these begin to occur, but it is known that at an 8-hour L   of
75 dB hearing damage (NIPTS) begins to occur.  The curve for average NIPES
versus L  .g. is used as the function to reduce tabulated data for
these extreme levels to a single-number indicator, because it is the only
direct health effect for which such a function has been established.
5.3  Analysis of impacts due to special noises
     The special noises discussed in this document are impulse noise, infra-
sound, ultrasound, and noises with information content.  Effects on humans,
structures, and animals all need to be considered.
     For any special noise, the main task is to describe the noise environ-
ment for the population.  As with general audible noise, tables such as those
in Chapter 2 may be needed.  Except for large impulse sounds, only a word
description of the effects of the special noise is recommended.  The criteria
of Chapter 3 should be referenced, but in many cases additional reference
material may be required.  A discussion of previous experience with such
noises should be made, if available.  For high energy impulse noise, the
analysis can be carried further and the expected percent highly annoyed, and
changes in this quantity, can be estimated.
     For structures exposed to impulse noise, the noise environment should
be described for each building or set of buildings in terms of maximum

sound pressure levels.  Either a worst case or a statistical estimate of the
distribution of maximum levels should be provided.   A discussion of the pos-
sible damaging effects of noise on structures or monuments is required.  The
chance that such effects could occur should be estimated.  Finally, the sig-
nificance of such damage, in monetary and/or non-monetary terms, should be
5.4  Analysis of impacts due to vibration
     If people are exposed, the analysis should include documentation of the
vibration environment such that the expected vibration acceleration values
due to the action are provided for all residential and other sensitive areas
in which the weighted acceleration exceeds the "no complaint" level (Table 12).
The change in the vibration environment can be discussed both by using the
average Vibration Impact Index for the exposed population and by listing the
expected effects at the nearest residence.  A discussion of the effects of
the vibration environment on sensitive non-residential buildings is also
     When structures are exposed to potentially damaging vibration, a descrip-
tion of the expected damage and the likelihood of such damage occurring should
be provided for each type of structure.  The information in Appendix C will
be of some help in making this assessment, but often enough data will not be
available to fully make this assessment.  In such cases, a program for moni-
toring the actual damage, or lack of it, may be necessary.


1.  Council on Environmental Quality.   National Environmental Policy Act-
    Regulations, Federal Register,  Vol. 43, No. 112,  June 9,  1978,
    pp. 25230 - 25247.

2.  "Information on Levels of Environmental Noise .Requisite TO Protect Public
    Health and Welfare with an Adequate Margin of-Safety," Report 550/9-74-004,
    U.S. Environmental Protection Agency,  March 1974.

3.  "Human Response to Impulse Noise," Report.of Working Group 84, Committee
    on Hearing, Bioacoustics and Biomechanics, The National Research Council,
    Washington, D.C.

4.  Von Gierke, H.G. (Task Group Chairman).  "Impact  Characterization  of Noise
    Including Implications of Identifying  and Achieving Levels of Cumulative
    Noise Exposure," Report NTID 73.4, U.S. Environmental Protection Agency,
    Aircraft/Airport Noise Study Report, July 1973.

5.  Schultz, T.J., "Synthesis of social surveys on noise annoyance," J. Acoust.
    Soc. Am., 64 (2), 1978, pp. 377-405.

6.  "Public Health and Welfare Criteria for.Noise," Report 550/9-73-002,
    U.S. Environmental Protection Agency,  July 1973.

7.  "Community Noise," Report NTID  300.3,  U.S. Environmental  Protection Agency,
    December 31, 1971.

8.  "Code of current practices for  enforcement of noise ordinances," forth-
    coming report, U.S. Environmental  Protection Agency.

9.  Kurze, U.J., Levison, W.H., and Serben, S., "User's Manual for the
    Prediction of Road Traffic Noise Computer Programs," Report DOT-TSC-315-1,
    U.S. Department of Transportation, May 1972.

10.  Gordon, C.G., Galloway, W.J.,  Kugler, B.A., and  Nelson,  D.L., "Highway Noise-
     A Design Guide for Engineers," Report 117, National Cooperative Highway
     Research Program, 1971.

11.  Kugler, B.A., and Pierson, A.G.,  "Highway Noise  - A Field Evaluation of
     Traffic Noise Reduction Measures," Report 144, National  Cooperative Highway
     Research Program, 1973.
                                   - . '                v.
12.  "Calculation of Day-Night.Levels- (L,) Resulting from Civil Aircraft
     Operations", Report 550/9-77;-450, .DVS. Environmental Protection  Agency, 1977.

13.  "FAA Integrated Noise Model Version 1:.  User's 'Guide".  Report FAA-EQ-78-01.
     Federal Aviation Administration,  1978.

14.  "Planning in the Noise Environment,"  draft joint services noise manual under
     cover of the Environmental Planning Bulletin Series No.  12, dated Dec 1976.


15.  Galloway, W.J., Eldred, K. McK., and Simpson, M.A.,  "Population Distri-
     bution of the United States  as  a Function of Outdoor Noise Level,"
     Report 550-9-74-009, U.S. Environmental Protection Agency, June 1974.

16.  Borsky, P.M., "The use of social surveys for measuring community responses
     to noise environments", in Transportation Noises, J.D. Chalupnik (ed).
     University of Washington Press, Seattle, 1970, pp. 219-227.

17.  Schultz, T.J., "Social surveys  on noise annoyance  -  further considerations."
     Paper presented at the International Congress on the Biological Effects of
     Noise, Freiburg, Germany, October 1978.

18.  "Noise-Final Report," H.M.S.O., Cmnd. 2056, London,  July 1963.

19.  Connor, W.K. and Patterson,  H.P., "Comnunity Reaction to Aircraft Noise
     Around Smaller City Airports", NASA CR-2104, August  1972.

20.  "Social and Economic Impact  of Aircraft Noise," Sector Group on the Urban
     Environment, OECD, April 1973.

21.  Lukas, J.S., "Measures of Noise Level:  Their Relative Accuracy in Predicting
     Objective and Subjective Responses  to Noise during  Sleep," Report
     600/1-77-010.  U.S. Environmental Protection Agency, February 1977.

22.  Goldstein, J., "Assessing the impact of transportation noise:  human  response
     measures," in Proceedings of the 1977 National Conference on Noise Control
     Engineering, G.C. Maling (ed).  Noise Control Foundation, Poughkeepsie,
     New York, 1977, pp. 79-98.

23.  Goldstein, J. and Lukas, J.S.,  "Noise and Sleep:  Information Needs  for
     Noise Control," in Proceedings of the Third International Congress on
     Noise as a Public Health Problem, J.V. Tobias (ed).  American Speech
     and Hearing Association, Washington, D.C., 1979.

24.  "Noise Emission Standards for Surface Transportation Equipment:  Regulatory
     Analysis of the Noise Emission Regulations for Truck-Mounted Solid Waste
     Compactors," Report 550/9-79-257, U.S. Environmental Protection Agency, August

25.  Galloway, W.J., "Evaluating  the Impact on the Public Health and Welfare of  a
     Change in Environmental Noise Exposure," Appendix D  of "Design Guide  for
     Highway Noise Prediction and Control," Final Report  Project 3-7/3, Submitted
     to Transportation Research Board, NCHRP, National Academy of Sciences,
     November 1974.

26.  Baughn, W.L., "Relation Between Daily Noise Exposure and Hearing Loss Based
     on the Evaluation of 6,835 Industrial Noise Exposure Cases," Joint EPA/USAF
     Study, prepared for 6570th Aerospace Medical Research Laboratory, Wright-
     Patterson AFB, Ohio.  AMRL-TR-73-53, 1973.

27.  Passchier-Vermeer, W.,  "Hearing Loss Due to Exposure to Steady-State Broad-
     band Noise," Instituut Voor Gezondheidstechniek,  Sound & Light Division,
     Report 35, April 1968 (with supplement,  1969).

28.  Robinson, D.W., "Estimating the risk of  hearing loss due to exposure to
     continuous noise," Occupational Hearing  Loss, British Acoustical Society
     Special Volume No. 1, London and New York,  Academic Press,  1971, pp. 43-62.

29.  Johnson, D.L., "Prediction of NIPTS Due  to Continuous Noise Exposure"
     Report 550/9-73-001-B (or AMRL-TR-73-91), U.S.  Environmental Protection
     Agency, 1973.

30.  Borsky, P.N., "Community Reactions to Sonic Booms in the Oklahoma City
     Area," Report AMRL-TR-65-37, National Opinion Research Center, 1965.

31.  Schemer, P.O., "Community Reaction to Impulse Noise:  First Army Survey,"
     Draft CERL Technical Report, U.S. Army Construction Engineering Research
     Laboratory, 1980.

32.  Schemer, P.O., "Evaluation of C-weighted L,  for assessment of impulse
     noise," J. Acoust. Soc. Am., 62 (2), 1977,^p.  396-399.

33.  Schomer, P.D., "Human and Community Response to Impulse Noise: A
     Literature Review," Illinois Institute for Environmental Quality
     Document No. 78/07, March 1978.

34.  Johnson, D.R. and Robinson, D.W., "The subjective evaluation of sonic
     bangs," Acustica, 18, 1967, pp. 241-258.

35.  Pawlowska, V. and Little, L., "The Blast Noise  Prediction Program:  User
     Reference Manual," Interim Report N-75,  Construction Engineering Research
     Laboratory," August 1979.

36.  Schomer, P.D., "Growth functions for human response to large-amplitude
     impulse noise," J. Acoust. Soc. Am., 64  (6), 1978, pp. 1627-1632.

37.  Schomer, P.O., Goff, R.J., and Little, L.M., "The Statistics of Amplitude
     and Spectrum of Blasts Propagated in the Atmosphere," Vol.  I., Report
     CERL-TR-N-13, U.S. Army Construction Engineering Research Laboratory,

38.  "Standard for Single Point Explosions in Air" draft report  dated Aug 20,
     1976 of Working Group S2-54, Committee on Mechanical Vibration and
     Shock, American National Standards Institute.

39.  Johnson, L., "Auditory and physiological effects of infrasound," Inter-
     noise 75 Proceedings.  Sendai, Japan, 1975, pp. 475-482.

40.  Von Gierke, H.E. and Parker, D.E., "Infrasound", in Handbook of Sensory
     Physiology, Vol. 5, Auditory System, Part 3, Springer-Verlag, Berlin-
     Heidelberg-New York, 1976, pp. 585-624.

41.  "Guide for the Evaluation of Human Exposure to Whole-body Vibration,"
     ISO 2631-1978, International Organization for Standardization, and
     Addendum 2, "Vibration and  Shock Limits for Occupants  in Buildings."

42.  "Guide for the Evaluation of Human Exposure to Whole-body Vibration,"
     ANSI S3.18-1979,  American National Standards Institute.

43.  Nicholls, H.R., Johnson,  C.F.,  and Ouvall, W.I.,  "Blasting Vibrations
     and their Effects on Structures," Bulletin 656, U.S. Department of  the
     Interior, Bureau  of Mines,  1971.

44.  "Guidelines for Preparing Environmental Impact Statements on Noise,"
     Report of Working Group 69, Committee on Hearing, Bioacoustics and
     Biomechanics,  The National  Research Council, Washington, D.C., 1977.

                                 Appendix A

             Acoustical Terms and Symbols Used in the Guidelines,
                 and Sane Mathematical Formulations for Them

A.I.    Acoustical terms

      Some acoustical terms are defined or described here, which have been
used in the main body of this report.  They are arranged alphabetically, to
facilitate finding them as needed.  Three key terms — sound level, equivalent
sound level, and sound exposure level — receive non-technical as well as
technical descriptions.

A. 1.1   C-weighted sound exposure level.  In decibels, the level of the time
integral of C-weighted squared sound pressure, with reference to the square
of 20 micropascals and to one second.

A. 1.2   day-night sound level.  The 24-hour equivalent sound level, in deci-
bels, obtained after addition of 10 decibels to sound levels in the night
from midnight up to 7 a.m. and from 10 p.m. to midnight (0000 up to 0700 and
2200 up to 2400 hours).

A. 1.3   day-night sound level contour.  A curved line connecting places on a
map where the day-night sound level is the same.  If only one kind of contour
is shown on the map the fact may be made known by a single legend, "Contours
of day-night sound level in decibels."  In this case only the number of deci-
bels need be marked on a coutour.

A. 1.4   day sound level.  Equivalent sound level over the 15-hour time period
from 7 a.m. up to 10 p.m. (0700 up to 2200 hours).

A. 1.5   decibel.  A unit measure of sound level and other kinds of levels.  It
is~a logarithmic measure.  For sound level specifically it is equal to 10 log
(P VP ref) or 20 log (p/Pref).
A. 1.6   8-hour equivalent C-weighted sound level.  Equivalent sound level, in
decibels, over a given 8-hour time period, measured with the C-frequency

A. 1.7   8-hour equivalent sound level.  Equivalent sound level, in decibels,
over an 8-hour period.  The A-frequency weighting is understood.

A. 1.8   equivalent sound level.  A sound level typical of the sound levels at
a certain place in  stated time period.  Technically, equivalent sound level
in decibels is the level of the mean-square A-weighted sound pressure during
the stated time period, with reference to the square of the standard refer-
ence sound pressure of 20 micropascals.  Equivalent sound level differs from
sound level in that for equivalent sound level, equal emphasis is given to
all sounds within the stated averaging period, whereas for sound level an
exponential time weighting puts much more emphasis on sounds that have just
occurred than those which occurred earlier.

A. 1.9   fast sound level.  In decibels, the exponential-tine-average  sound
level measured with the squared-pressure tine constant of 125 ms.

A. 1.10  hourly equivalent sound level.  Equivalent sound level,  in decibels,
over a one-hour time period, usually reckoned between integral hours.   It may
be identified by the beginning and ending tines, or by the ending tine  only.

A. 1.11  impulse sound level.  In decibels, the exponential-tine-average sound
level obtained with a squared-pressure time constant of 35 milliseconds.
(But note that there is as yet no U.S. standard for this.)

A.1.12  instantaneous sound pressure, overpressure.  Pressure at a place and
instant considered, minus the static pressure there.

A.1.13  maximum sound pressure level.  Sane as peak sound pressure level,
provided that the tine interval considered is not less than a complete
period of a periodic wave.

A. 1.14  night sound level.  Equivalent sound level, in decibels, over the nine-
hour period from midnight up to 7 a.m. and from 10 p.m. to midnight  (0000 up
to 0700 and 2200 up to 2400 hours).

A.1.15  noise level.  Sane as sound level, for sound in air.  Some people use
"noise" only for sound that is undesirable.  A sound level meter does not,
however, measure people's desires.  Hence there is less likelihood of mis-
understanding, if what is measured by a sound level meter is called sound
level, rather than noise level.

A. 1.16  peak sound pressure.  Greatest absolute instantaneous sound pressure
in a stated frequency band, during a given tine interval.   (Also called
peak pressure.)

A. 1.17  peak sound pressure level.  In decibels, twenty tines the cannon
logarithm of the ratio of a greatest absolute instantaneous sound pressure
to the reference sound pressure of twenty micropascals.

A. 1.18  slow C-weighted sound level.  In decibels, the exponential time aver-
age sound level measured with the squared-pressure time constant of one second
and the C-frequency weighting of the sound level meter.

A.1.19  slow sound level.  In decibels, the exponential-tine-average  sound
level measured with the squared-pressure tine constant of one second.

A. 1.20  sound exposure.  Tine integral of squared, A-frequency-weighted sound
pressure over a stated tine interval or event.  The exponent of  sound pres-
sure and the frequency weighting may be otherwise if clearly so  specified.

A.1.21  sound exposure level.  The level of sound accumulated over a  given
tine period or event.  It is particularly appropriate for a discrete  event
such as the passage of an airplane, a railroad train, or a truck.  Sound
exposure level is not an average, but a kind of sum.  In contrast to  equi-
valent sound level which may tend to stay relatively constant even though


the sound fluctuates, sound exposure level increases continuously with the
passing of time.  Technically, sound exposure level in decibels is the level
of the time integral of A-weighted squared sound pressure over a stated time
interval or event, with reference to the square of the standard reference
pressure of 20 micropascals and reference duration of one second.

A.1.22  sound level.  The weighted sound pressure level, which reduces to a
single number the full information about sound pressure levels across the
frequency range 20 Hz to 20 kHz.  It can be measured by a sound level meter
which meets the requirements of American National Standard Specification for
Sound Level Meters SI.4-1971.  In these guidelines, fast tine-averaging and
A-frequency weighting are understood, unless others are specified.  The
sound level meter with the A-weighting is progressively less sensitive to
sounds of frequency below 1000 hertz (cycles per second), somewhat as is the
ear.  With fast time averaging the sound level meter responds particularly
to recent sounds almost as quickly as does the ear in judging the loudness
of a sound.

A. 1.23  sound pressure.  Rcot-mean-square of instantaneous sound pressures
over a given time interval.  The frequency bandwidth must be identified.

A. 1.24  sound pressure level.  In decibels, twenty times the common logarithm
of the ratio of a sound pressure to the reference sound pressure of twenty
micropascals (0.0002 microbar).  The frequency bandwidth must be identified.

A.I.25  (vibratory) acceleration.  The rate of change of velocity of a vibra-
tion, in a specified direction.  The frequency bandwidth must be identified.

A. 1.26  (vibratory) acceleration level.  In decibels, tarenty times the common
logarithm of the ratio of a vibratory acceleration to the reference accel-
eration of ten micrometers per second squared (nearly one-millionth of the
standard acceleration of free fall).  The frequency bandwidth must be

A.1.27  yearly day-night sound level.  The day-night sound level, in decibels,
averaged over an entire calendar year.

A.2.  Symbols used in the guidelines

      A. 1.1   C-weighted sound exposure level
      A.1.2   day-night sound level
      A. 1.5   decibel
      A.I.7   8-hour equivalent sound level
      A.1.8   equivalent sound level                       Lea
      A. 1.10  hourly equivalent sound level                Leam
      A.I.21  sound exposure level                         LT*
      A. 1.23  sound pressure                               p
      A.1.25  vibratory acceleration                       a
      A.I.27  yearly day-night sound level                 Ldn(v)

A.3.  Mathematical Formulations for the Descriptors used in the guidelines

A.3.1   Equivalent sound level

              Leq = 10
                                                      Eqn A-1
      where:  T is the length of the tine interval during which the
              average is  taken, and L.(t) is the time varying
              value of the A-weightea sound level during the time
              interval T.

      Note:   Equivalent  sound level may be calculated from the sound
              exposure levels of individual events occurring within
              the time interval T:
                                          Eqn A-2
      where:  Lg.  is the sound exposure level of the i-th event, out
              of a total of n events in time interval T.  Lu is
              defined in A.2.3.4.                          b
A.3.2  Day-night Sound Level
                    •   f
2200                    2400
     10 A       dt +
      Time t is in seconds, so the limits shown in hours and minutes
      are actually interpreted in seconds.  It is often convenient to
      compute day-night sound level from hourly equivalent sound
      levels obtained .during successive hours:
                                          Eqn A-4
      where L^i is the hourly equivalent sound level for the i-th hour
      of the day and LHJ  is the hourly equivalent sound level for the
      j-th hour of the night.

A. 3. 3  Yearly Day-night Sound Level

                                 t  365    L,  ./10
              Ldn(y) = 1     10 355
      where:  l^ni *-s tne day-night average sound level for the
              i-th day out of one year.
A. 3. 4  Sound Exposure Level
                                                                     Eqn A-6
      where:  LA(t) is the time-varying A-weighted sound level in
              some time interval t. to t_.

      The length of the time interval may be arbitrary, or it may
      simply be large enough to encompass all the significant sound
      of an event.
      Note:   The value of the above integral is usually approximated
         *    with sufficient accuracy by integrating LA(t) over the
              time interval during which LA(t) is between 10 decibels
              less than its maximum value and the maximum value, before
              and after the maximum occurs.
A.3.5  C-weighted Sound Exposure Level
              1^- 10 Iog10\  y  10 *-      dt/                     EqnA-7
                             \ti              /
      where:  Lg(t) is the time-varying C-weighted sound level in
              some time interval t-j to t2.

      Note:   In practice the  integral is often approximated by inte-
              gration within the time during which the sound level of
              the event exceeds some threshold value such as 20 dB
              less than the maximum sound pressure level.

A.3.6  C-weighted Day Night Sound Level
      Analogous to the A-weighted L^n, with a nighttime penalty of 10 dBf the
      C-weighted day-night average sound level is:

             - 10
15 x 10
                     9 x 10
                                                         + 10
                                                  Eqn A-8
    is the average C-weighted sound level over  the daytime  period of
0*700 to 2200 hours, L   is the C-weighted average level over the night-
time period of 2200 tff\)700 hours.

     The C-weighted average level is most easily calculated from the
C-weighted sound exposure levels during  the time of  interest as  follows:
                     15 x 3600
                                                       Eqn A-9
10 log
                                       Eqn A-10
where L^ is the C-weighted sound exposure  level of  the i-th

             Appendix B
Procedures of Other Federal Agencies

                                                       bl.  Knvironucntal Noise Measures and Their
                                                              Purposes in Federal Programs

        Ty|je of Program
        or Policy
        Key Documents
               1. FIIUA

              Highway Noise
              FIlPM 7-773
                   ±. EPA

                   Health & Welfare
                   EPA "Levels"
                   Document (1974)
3. HUD

HUD Noise
hUD Circular
1390.2 (1971)
4. DOD

Air Installation
Compatible Use
Zones (AICUZ)

DOD Instruction
4165.57 (1977)
                     5. FAA

                     Airport 6 Land
                     Use compatibility
                     Advisory Circular

                     150/5050-b (1977)
                     b. VA

                     VA Noise
                     Section vm
                     Appraisal of
                     near Airports
Title of
             furpose of
             source to
              wh ich

Design Noise
              These levels are
              used in deter-
              mining where
              noise mitigation
              on a particular
              highway project
              is warranted.
              They reflect cost
              and feasibility
              They are not
              appropriate land
              use criteria.
              Highway only
Levels which are
required to protect
the public health and
welfare with an ade-
quate margin of safety

These levels identify
in scientific terms
the threshold of
effect. While the
levels have relevance
for planning, they
do not in themselves
form the sole basis
for appropriate land
use actions because
they do not consider
cost, feasibility or
the development needs
of the community.  The
user should make such

All sources
                  or L|0
              for design hour
Level determin-
ing whether pro-
jected sites arc
eligible for HUD

See above. Levels
can be used as
general planning
levels. Considers
balance between
cost, feasibility
effect, community
developuent needs
and availability
of land for deve-
lopment .  Com-
munity wide con-
                                               All sources
                                               airport,  highway
                                               and rail

                                               (accepts  Ldn)
Levels used as
guidance to
communities in

Guidance to com-
munities for
planning. Considers
balance between
cost, feasibility
effect, community
development needs,
and availability
of land for
Community wide
Levels used as
"starting points"
in determining
appropriate land/
use relationships

Guidance to com-
munities for plan-
ning. Considers
balance between
cost, feasibility
effect, community
development needs,
and availability
of land for
                   Civil Airports
                   (including NEF
                     and Ldn)
                                          Levels determi-
                                          ning whether
                                          projected sites
                                          are eligible fur
                                          VA assistance

                                          See above. Consi
                                          balance between
                                          cost, feasibilit
                                          effect, communit
                                          development need
                                          and availability
                                          of land for deve
                                          ment needs and
                                          availability of
                                          land for develop
                   Airports only
                   (including Ldn)

B2.  Estimating L,  from other Noise Measures

     The equations listed here are approximations only, and are provided for
use in those situations in which measurement or prediction of the other noise
measure is already available.  If no such information is available, it is
strongly recommended that L,  be measured or predicted directly, instead
of using these equations.

     NEP:                         L^ » NEF + 35                      Eqn B-1

     CNR:                         L,  = CNR - 35                      Eqn B-2

     CNEL:                        L^ = CNEL                          Eqn B-3

     24-hour L^:3                L^ . L       +4                   Eqn B-4
     Peak  (traffic) hour I^q:     *dn =

     Peak  (traffic) hour L-josb    I<3n a L10 "3                        &$\ B-€

aSource:   [4], Parts II B, F, and Addendum A, approximated.

 Source:   Department of Housing and Urban Development.  Notice of proposed
 rulemaking.  Environmental criteria and standards, Federal Register, Vol 43,
 No. 249,  December 29, 1978, p. 60399.  "The day-night average sound level may
 be estimated from the design hour L-JQ or Lgq values by  [these] relationships,
 provided  heavy trucks do not exceed  10 percent of the total traffic flow in
 vehicles  per 24 hours and the traffic flow between 10 p.m. and 7 a.m. does not
 exceed  15 percent of the average daily traffic flow in vehicles per 24 hours."

                                  APPENDIX C

               Summary of Human Effects of General Audible Noise

TABLE Cl  Summary of Human Effects for Outdoor Day-Night Sound Level of
          75 Decibels

  Type of Effect                        Magnitude of Effect

Hearing Loss           May begin  to occur in  sensitive  individuals, depending
                       on actual  noise levels received  at-ear.

Risk of non-auditory   *
  health effects
Speech  - Indoors     Some disturbance of normal conversation.  Sentence
                       intelligibility (average) approximately 98%

        - Outdoors     Very significant disturbance of  normal voice or relaxed
                       conversation with:  100% sentence intelligibility not
                       possible at any distance

                         or, 99%  sentence intelligibility  (average) at 0.15 meter

                         or, 95%  sentence intelligibility  (average) at 0.5 meter

High Annoyance        Depending  on attitude  and other  non-acoustical factors,
                       approximately 37% of the population will be highly annoyed.

Average Ccnnunity     Very severe; 13 dB above level of significant "complaints
  Reaction             and threats of legal action" and at least 3 dB above
                       "vigorous  action" (attitudes and other non-acoustical fac-
                       tors may modify this effect).

Attitudes Towards     Noise is likely to be  the most important of all adverse
  Area                 aspects of the community environment.

^Research inplicates noise as one of several  factors producing stress-related health
 effects such  as  heart disease, high-blood pressure and stroke, ulcers and
 other digestive  disorders.  The  relationships between  noise and these
 effects have  not yet been quantified, however.
  The speech effects data in these tables are drawn from the Levels
 Document, as  follows.  Indoor effects are based on Table 3, and on Fig. D-l,
 with 15 dB added to the indoor level to obtain the outdoor reading.  Outdoor
 effects come  from  Fig. D-2, using L, (as determined with Fig. A-7).  Both
 Figures D-l and  D-2 are based on steady noise, not on  L   .  Table D-3
 shows that for a fluctuating noise, the average percent interference can be
 higher or lower  than for steady  noise with the same L  .  The values
 given in this report are the best estimates  of the interference.


TABLE C2  Summary of Human Effects for Outdoor Day-Night Sound Level of
          70 Decibels
     Type of Effect

Hearing Loss

Risk of non-auditory health
  effects (stress)

Speech - Indoors
       - Outdoors
High Annoyance
Average Community Reaction
Attitudes Towards Area
          Magnitude of Effect

Will not likely occur

See Table C1
Slight disturbance of normal conversation
approximately 99% sentence intelligibility

Significant disturbance of normal voice or
relaxed conversation with 100% sentence
intelligibility (average) possible only at
distances less than 0.1 meter
                                      99% sentence intelligibility (average) at
                                      0.3 meter
95% sentence intelligibility  (average) at
0.9 meter

Depending on attitude and other non-
acoustical factors, approximately 25
percent of the population will be highly

Severe; 8 dB above level of significant
"complaints and threats of legal action,"
but at least 2 dB below "vigorous action"
(attitudes and other non-acoustical
factors may modify this effect)

Noise is one of the most important adverse
aspects of the community environment

TABLE C3  Summary of Hunan Effects for Outdoor Day-Night Sound Level of
          65 Decibels
     Type of Effect

Hearing Loss

Risk of non-auditory health
  effects (stress)

Speech - Indoors
       - Outdoors
High Annoyance
Average Community Reaction
Attitudes Towards Area
          Magnitude of Effect

Will not occur

See Table C1
Slight disturbance of normal conversation
99% sentence intelligibility (average)
with a 4 dB margin of safety

Significant disturbance of normal voice or
relaxed conversation with 100% sentence
intelligibility  (average) at 0.15 meter
                                      99% sentence intelligibility (average) at
                                      0.5 meter
95% sentence intelligibility  (average) at
1.5 meters

Depending on attitude and other non-
acoustical factors, approximately 15
percent of the population will be highly

Significant; 3 dB above level of significant
"complaints and threats of legal action,"
but at least 7 dB below "vigorous action"
(attitudes and other non-acoustical
factors may modify this effect)

Noise is one of the important adverse
aspects of the community environment

TABLE C4  Suitmary of Human Effects for Outdoor Day-Night Sound Level of
          60 Decibels
     Type of Effect

Hearing Loss

Risk of non-auditory health
  effects (stress)

Speech - Indoors
       - Outdoors
          Magnitude of Effect

Will not occur

See Table C1
No disturbance of normal conversation
100% sentence intelligibility  (average)
with no margin of safety

Moderate disturbance of normal voice or
relaxed conversation with  100% sentence
intelligibility (average)  at 0.2 meter
                                      99% sentence intelligibility (average) at
                                      0.6 meter
High Annoyance
Average Community Reaction
Attitudes Towards Area
95% sentence intelligibility  (average) at
2 meters

Depending on attitude and other non-
acoustical factors, approximately 9
percent of the population will be highly

Slight to moderate; 2 dB below level of
significant "complaints and threats of
legal act ion," but at least 11 dB below
"vigorous action" (attitudes  and other
non-acoustical factors may modify this

Noise may be considered an adverse aspect
of the connunity environment

TABLE C5  Summary of Human Effects for Outdoor Day-Night Sound Level of
          55 Decibels
     Type of Effect

Hearing Loss

Risk of non-auditory health
  effects (stress)

Speech - Indoors
       - Outdoors
High Annoyance
Average Community Reaction
Attitudes Towards Area
          Magnitude of Effect

Will not occur

See Table C1
No disturbance of normal conversation
100% sentence intelligibility (average)
with a 5 dB margin of safety

Slight disturbance of normal voice or
relaxed conversation with:  100% sentence
intelligibility (average) at 0.35 meter
99% sentence intelligibility (average) at
1.0 meter


95% sentence intelligibility (average) at
3.5 meters

Depending on attitude and other non-
acoustical factors, approximately 4
percent of the population will be highly

None expected; 7 dB below level of signi-
ficant "complaints and threats of legal
action/1 but at least 16 dB below "vigorous
action" (attitudes and other non-acoustical
factors may modify this effect)

Noise considered no more important than
various other environmental factors

                                  APPENDIX D

          Measurement of and Criteria for Human Vibration Exposure

D.I.   Introduction

     The criteria for vibration exposure in this appendix will address 3 types
of effects.  These three types of effects are:  (1) whole body vibration of
humans, (2) annoyance and interference caused by building vibration, and (3)
structural damage from building vibration.

     The existing state of knowledge is not complete in any of the above
three areas; however, there are existing I.S.O. standards that have been
approved or proposed.  Summaries of these standards, along with other data,
provide the content of this appendix.  Some simplification of the proposed
standards on building vibration and structural damage have been made in order
to provide a simple, unified and reasonable method for assessing the effects
of vibration.

D.2.   Whole Body Vibration Criteria (Summary of Approved ISO Standard 2631-1978)

D.2.1  The Three Criteria for Evaluation of Whole Body Vibration

     Experimental data show that there are various rather complex factors
that determine the human response to vibration.  Evaluation of all these
factors is difficult at this time because of the paucity of quantitative
data, concerning man's perception of vibration and his response to it.  Never-
theless, there is an international standard which does provide provisional
guidance as to what is acceptable human exposure to vibration for some types
of vibration.

     In general, there are four physical factors of primary importance in
determining the human response to vibration.  These are intensity, frequency,
direction, and exposure time of the vibration.  The current International
Standard for vibration addresses three main human criteria.  These are:

     1.  Preservation of working efficiency

     2.  Preservation of health or safety

     3.  The preservation of comfort

     For environmental problems, the preservation of comfort is considered
the best criteria for evaluation of whether or not vibration significantly
changes the environment.

D.2.2  Types of Vibration Transmissions.

     The standard lists basically three kinds of human response to vibration,

     (a)  Vibrations transmitted simultaneously to the whole body surface
or substantial parts of it.  This occurs when the body is immersed in a
vibration medium.  There are circumstances in which this is of practical
concern; for example, when high intensity sound in air or water excites
vibrations of the body.

     (b)  Vibration transmitted to the body as a whole through the supporting
surface, namely/ the feet of a standing man, the buttocks of a seated man or
the supporting area of a reclining man.  This kind of vibration is usual in
vehicles, in vibrating buildings and in the vicinity of working machinery.

     (c)  Vibrations applied to particular parts of the body such as the
head or limbs; for example, by vibrating handles, pedals, or head-rests,
or by the wide variety of powered tools and appliances held in the hand.

     It is also possible to recognize the condition in which an indirect
vibration nuisance is caused by the vibration of external objects in the
visual field (for example, an instrument panel).

     The International Standard 2631, however, applies chiefly to the common
condition (b) above; and, in particular, where the vibration is applied
through the principal supporting surface to the body of a standing or seated
man.  In the case of vibrations applied directly to a reclining or recumbent
man, insufficient data are available to make a firm recommendation; this is
particularly true of vibration transmitted directly to the head, when toler-
ability is generally reduced.  Tolerance may also be reduced when conditions
(b) and (c) exist together.  Provisionally, however, the limits for the
standing or seated man may also be used for the reclining or recumbent man.
It must be appreciated that some circumstances will arise in which the rigor-
ous application of these limits would be inappropriate.

D.2.3  Direction of Vibration

     Rectilinear vibrations transmitted to man should be measured in the
appropriate directions of an orthogonal coordinate system centered at the
heart.  The standard specifies separate criteria according to whether the
vibration is in the longitudinal (H- az) direction or transverse (± ax or ay)
plane.  Accelerations in the foot (or buttocks) - to head (or longitudinal)
axis are designated + az:  acceleration in the fore-and-aft (anteposterior
or chest-to-back) axis, + ax; and in the lateral (right-to-left side) axis,
+ ay.  These axes are illustrated in Figure D-1.

D.2.4  Acceptable Whole Body Vibration

     The ISO standard identifies the 24-hr comfort level for rms pure
(sinusoidal single) frequency or rms value in third octave band for random
vibration as given in Table 0-1.  As long as the vibration levels are below
the 24-hr levels, vibration should be considered to have no direct impact
on an individual, regardless of the duration of the exposure.  The standard
does allow for increased exposure levels for shorter exposure times.  Such a
tradeoff is given by Table D-1 for 8-hr and 1 min exposures.  For other

                ax, 3y, az = acceleration in the directions of the x, y, z axes
                x axis     = back to chest
                y axis     = right to left side
                z axis     = foot for buttock(s)-to-head

FIGURE D—1. Directions of co-ordinate system for mechanical vibrations influencing humans

TABLE D-1 - Numerical values of "comfort boundary" for vibration acceleration
            in the longitudinal, a_, direction (foot (or buttocks)-to-head
            direction) (see FigureT>-1 and in the transverse, a  or a . direc-
            tion (back-to-chest or side-to-side)                     y

Values define the boundary in terms of rms value of pure (sinusoidal) single
frequency vibration; or rms value in third-octave band for distributed vibration.
                             ACCELERATION m/sec
    Frequency (Hz)
    (Center Frequency
    of 1/3 Octave Band)




















24 hr
1 min
24 hr

exposure times and for the concept of a vibration dose, the basic standard
should be consulted.  For occupational and recreational situations, the values
of Table D-l can be raised by a factor of 3.15 (10 dB) to predict the boundary
at which working efficiency may start to decrease.  Increasing the accelera-
tion listed in Table D-l by a factor of 6.3 (16 dB) will give the boundary
necessary for the preservation of health and safety.  Thus the 1 min values of
Table D-l as multiplied by a factor of 6.3 provides the maximum recommended
continuous acceleration to which an individual should be subjected.  However,
assessment of acceleration above the comfort levels listed in Table D-l should
be made only by direct reference to the ISO standard.  In the ISO standard
there are many considerations and limitations with respect to human exposure
to acceleration that can cause reduced efficiency or health and safety

D.3.   Vibration Criteria for Occupants in Buildings.  (Summary of 1980 draft
     addendum^ to ISO Standard 2631-1978)

D.3.1  Scope

     The proposed standard takes into account the following factors:

     1.  Type of Excitation -. for example transient (shock) and/or
         steady vibration;

     2.  Usage of the Occupied Space in Buildings - for example, hospital
         operating theatres, residential, offices and factories;

     3.  Time of Day;

     4.  Limits of Acceptability - in a proposal of this type there is no
         hard and fast line of acceptability, but guidance is given as to
         the level of complaint to be achieved at different levels of
         vibration.  In cases where sensitive equipment or delicate opera-
         tions impose more stringent limits than human comfort criteria,
         then the more stringent criteria should be applied.

D.3.2  Characteristics of Building Vibration

D.3.2.1  Direction of vibration

     Because a building may be used for many different activities, standing,
sitting and lying may all occur, hence vertical vibration of the building may
enter the body as either Z axis, X axis or Y axis vibration, as shown in Figure
D-l.  The Standard is written for all three axes of vibration, however, in cases
where it is not clear which direction to apply, it is often more convenient to
consider the combined Standard detailed in Sections D.3.3.4 below.

D.3.2.2  Random or multi-frequency vibration

     Random or multi-frequency vibration represents a particular problem which
fortunately does not often occur in buildings.  There is evidence from research
concerning the building environment to suggest that there are interaction


effects between different frequencies of vibration.  Under these circumstances
and for random vibration, the proposed standard recommends an overall weighting
method such as that in section D.3.3.4.

D.3.2.3  The characterization of impulsive shock and intermittent vibration

     Continuous vibration of a repetitive nature is easy to identify and
classify.  The borderline between impulsive shock and intermittent vibration
is difficult to define.  Impulsive shock is characterized by a rapid build-up
to a peak followed by decay, and is typically excited in buildings by blasting,
forging presses or pile driving using an impact device.  Intermittent vibration
may only last a few seconds, but is characterized by a build-up to a level
which is maintained for a considerable number of cycles.  Examples of this in
buildings would be traffic excited vibration and vibration generated inside  a
building by machinery starting up or on intermittent service.  Pile driving  by
modern methods using vibrating columns would also be classified as continuous
or intermittent vibration and not as impulsive shock.

     The proposed standard recommends that impulsive shock created by forging
presses or conventional pile drivers should be treated in a similar manner to
continuous and intermittent vibration.  Research has shown that vibration
which only occurs at a specific instance, for example domestic building vibra-
tion by a passing bus, causes the same level of annoyance as continuous

     Blasting which occurs only up to three tines per day is a special case.
The proposed standard recommends that building operations of this nature
should never take place at night due to the disturbance and that during the
daytime they should be limited to a small number of occurrences.  The levels
of vibration generated due to blasting are on an order of magnitude greater
than traffic and general building vibrations, and can only be accepted on the
basis of very limited exposure.

D.3.2.4  Classification of buildings and building areas

     The criteria of classification are in the standard derived from the
human reaction to vibration.  In the home the highest standards are required,
and this is characterized by an absence of detectable vibration.  Under other
conditions, such as offices and factories, there is some tolerance to vibra-
tion disturbance.

     In the proposed Standard no differentiation has been made between differ-
ent types of residential areas, i.e. city centre, urban or rural.  It is con-
sidered that similar standards should be met for all occupants of residential
property.  Some types of areas have not been classified, i.e. restaurants
or places of entertainment, but common sense suggests the most appropriate
classification—for example standards in a restaurant should be similar to
those in residential property.  It should be noted that certain entertainment
areas in long span buildings present particular problems from self-generated
vibration, such as that from dancing.

     Hospitals have not been given more restrictive levels  in general because
there is sane evidence that patients prefer to be in touch  to some extent with
the outside world, but operating theatres and laboratories  should be considered
as critical areas.

D.3.2.5  Measurement of vibration

     The use of "root mean square11 acceleration is reconmended as the standard
unit of measurement.  If possible building vibration should be measured in
acceleration terms, but in some cases it may be found necessary to measure  in
velocity or displacement due to equipment limitations.  For these situations
the vibration should be treated as sinusoidal and the appropriate correction
factors, which are a function of frequency, used to transform either the mea-
surement or the standard into compatible units.

     In the case of impulsive vibration or shock the instantaneous peak value
of velocity or acceleration is the preferred unit of measurement.  A trace  of
the vibration should be obtained upon a suitable instrument and the peak level
estimated.  The notion should then be considered sinusoidal and the correction
factors applied for the difference between peak and rms, and the frequency
dependent factors used to transform either measurement or standard into com-
patible units.

     If frequency analysis of the vibration is required, third octave filters
are recommended.  In certain circumstances it may be useful to analyze the
vibration in terms of narrow fixed band width filters.

     Measurement of vibration should be taken on the floor  at the point of
greatest amplitude, comnonly found at mid-span.  This should be close to the
point of entry of vibration to the human subject.  Measurement should be
taken along the three orthogonal axes, and reference made to the appropriate
human axis standard to determine whether limits have been exceeded.  Alter-
natively the weighting network or combination curves (see Section D.3.3.4)
could be considered in relation to the worse case found.

     In the case of impulsive shock caused by blasting, measurement may be
made at the foundations to check for structural damage.  It is also necessary
to measure according to the technique given above in the areas of human

D.3.3  Characterization of Building Vibration and Acceptable Limits

D.3.3.1  Acceptable Limits

     All the following proposals are related to the recommendations for
general vibration on humans given in Section D.2.  The presentation of infor-
mation is in the form of a basic rating which is given for  the irost stringent
conditions.  From this basic rating a multiplication factor is then applied
according to the tables for other more permissive situations.

     The lowest basic rating has been defined in the area of the threshold  of
human perception.  It is based upon research work completed up to the end of


     Experience has shown in many countries that complaints of building
vibrations in residential situations are likely to arise from occupants if
the vibration levels are only slightly in excess of perception levels.  In
general, the limits are related to the acceptance by the occupants and are
not determined by any other factors such as short-term health and work
efficiency.  Indeed the levels are such that there is no possibility of
fatigue or other vibration induced syndromes.

D.3.3.2  Head to Foot ("Z" Axis) Vibration Limits

     For Z axis the recommended vibration values proposed by the standard
is shown in Figure D-2.  For frequencies between 4 Hz and 8 Hz the maximum
acceleration (rms) is 5 x 10~3 m/s2.  At frequencies below 4 Hz the limit
changes at 3 dB/octave.  For frequencies greater than 8 Hz the limit increases
by 6 dB/octave.  For conditions other than the base curve a series of weight-
ing factors apply and these are given in Table D-2.  For example, for residen-
tial property the weighting factor is two, hence at 4 to 8 Hz the maximum
recommended rms acceleration for residential property by day would be 10~2 m/s  Side to side or front to back (X or Y axis) vibration limits

     For X and Y axis human vibration a different base curve applies which
is shown in Figure D-2.  For frequencies from 1 - 2 Hz a maximum acceleration
level of 3.6 x 10~3 m/s2 will apply.  At frequencies higher than 2 Hz the
acceptable acceleration level will increase at 6 dB/octave.  This means that
for frequencies greater than 2 Hz a maximum rms velocity limit applies.

     It will be noted that the standard for X or Y axis vibration is more
severe than the Z axis case at low frequencies.  This is due to the sensi-
tivity of the human body towards sway at these low frequencies.

     The table of weighting factors given in Table D-2 also applies to
X or Y axis vibration.

D.3.3.4  Combined standard - recommended limits for undefined axis of human
       vibration exposure

D.  Worst case combination curve

     In many situations the same building area may be used in both the lying
and standing positions at different times of the day.  If this is the case,
then a combined Standard using the worst case combination of both the Z axis
and X and Y axis conditions may be applied.  This combination curve is shown
in Figure Dr2 and the same weighting factors given in Table D-2 still apply.

D.  Proposed weighting network

     The proposed standard also recommends a weighting network that closely
approximates the combination curve.  For routine measurement and evaluation
of environmental vibration, this frequency weighting is recommended.  The
weighting function proposed for combined or random vibrations is given by:


















                                  I   I  1  i i
                                 OMBINED  WORST CASE X,Y OR  Z  AXIS  .
                           i    i   i   I  I  i I
                                                           i    i   i   I  I  i i
                     2     3456789          20

                                  FREQUENCY   (Hz }
                                                              40  60  80
                                  I I * ^ *• >i«ta. I V %^ I   \ t t fr i

       FIGURE D—2.  Building vibration criteria for occupants in buildings. All curves are for

                   hospital and critical working areas. See Table D—2 for proper scale factors.

   TABLE D-2  Weighting Factors for Acceptable Building Vibration


Hospital operating
theatre & critical
working areas









Continuous or
Vibration &
Impulsive Shock





Impulsive Shock
Excitation with
not more than 3
Occurrences per day






Weighting Factors above basic level of Curve shown in Figure D-2

          G (JW) -  1+  J«-
                        11.2  -----                      -        Eqn D-1

where G (Jtu) is the transmissibility of the filter, J represents the square
root of -1, at represents the exciting frequency.

This mathematical expression defines the electronic weighting filter of the
low pass type.  At low frequencies the transmissibility is zero, and at high
frequencies attenuation is at 6 dB/octave.  The corner frequency is 5.6 Hz.

Accuracy - ^ 0.2 dB

Although the proposed standard recommends this function for preliminary
investigations, for practical evaluations of the overall environmental
impact of vibration on a community, the weighting function is a necessary
and useful simplification, especially with respect to residential areas,
that is not expected to introduce any significant errors.

D.4.  Structural Damage from Building Vibration.  (Summary of 1976 draft
    standard ISO/TC 108/SC 2/WG3

D.4.1  General considerations

     The proposed standard discusses the following general considerations:

Vibration in buildings (dwellings, offices, public buildings and factories)
is of increasing general importance, especially since the distances between
industrial areas with vibration exciting machines, blasts or other vibration
sources and residential areas are decreasing.  Traffic on roads and railroads
also causes vibration troubles in nearby buildings.

Various methods of rating the severity of vibration in buildings and defining
limits based on laboratory or field data have been developed in the past.
However, none of these methods can be considered applicable in all situations
and consequently none have been universally accepted.

In view of the complex factors required to determine the response of a build-
ing due to vibrations and in view of the paucity of quantitative data, this
proposed Standard was prepared, first to facilitate the evaluation and com-
parison of data gained from continuing research in this field; and, second,
to give provisional guidance as to the acceptable values in order to avoid the
risk of damage.  The limits proposed are a compromise of available data.  They
satisfy the need for recommendations which are simple and suitable for
general application.  These limits are defined explicitly in numerical
terms to avoid ambiguity and to encourage precise measurement in practice.

If the characteristics of the excitation vibration are known in relation
to the severity, position and direction of the building response—this may
be the case if the source of the vibration is within the building—and if
the parts of the buildings or the whole building influenced by the vibra-
tions can be idealized by a model, then it may be possible to estimate the
severity of the dynamic stresses by calculation.


If vibrations are transmitted via the ground and the foundation into a building,
it may be possible to estimate dynamic stresses based on vibration measurements.

In addition to simple vibration there may be other factors which influence
vibration response (foundation conditions, dilatation due to temperature etc.)
and which result in damage to buildings.  No general method exists at present
to take account into all such factors.

D.4.2  Categories of Damage

     The proposed standard provides several phases of damage which can occur,

Category 1:

   Threshold damage consists of visible cracks in non-structural members
such as partitions, facings, plasterwalls (e.g. loosening of mortar between
pantiles etc.).  As a guideline visible cracks may be taken as those of a
width of 0.02 mm.

Category 2:

   Minor damage consists of visible cracks in structural members such as
masonry walls, beams, columns, slabs and no serious reduction in load
carrying capacity.

Category 3:

   Major damage consists of large permanent cracks in non-structural and
structural members; settlement and displacements of foundations which may
result in reduction of load carrying capacity.

The proposed standard applies chiefly to damage as described in categories
1 and 2.  The limits of vibration specified in the standard were selected
to avoid the exceeding of the threshold of damage, but does include data
for estimating damage levels.

D.4.3  Measurement

D.4.3.1  Frequencies

     The proposed standard recommends the following frequency ranges:

     1.  In the case of vibration caused by shock and quarry blasting and
         the steady vibration of whole buildings:  from about 1 Hz to about
         100 Hz.

     2.  In the case of steady vibration of parts of a building, especially
         floor and wall vibrations:  from about 10 Hz to about 100 Hz.

-------  Measurement points

     The standard recommends that for vibration caused by shock, especially
quarry blasting, should be measured on the foundation structure parallel to
its stiff-axes below ground level.

     In only special cases are measurements of the floor vibration in vertical
direction and the horizontal vibration of the whole building recommended.
Such floor vibration measurements should be made in a manner similar to that
of section D.3.

     In the case of steady vibration (e.g. floor vibration), the vibration
peak velocity, VMSX, at the place of highest amplitude shall be determined.
In floor vibration it is often the midspan, for whole building vibration it
is often the upper floor in horizontal direction.

D.4.3.3  Measurement quantity

     Vibration can be measured by displacement, velocity or acceleration.
It is desirable to measure the quantity that is most simply and generally
related to damage as described below.  While for steady vibration the
proposed standard provides curves related to velocity from 10 Hz to 80 Hz
(Figure D-3), it can be seen that for the frequency range of 10 to 80 Hz,
acceleration as weighted by the function in Chapter 3 is for all practical
purposes a measure of velocity.  Plotting the weighted acceleration against
actual blast damage data, see Figure D-4, the weighted acceleration provides
a very reasonable fit to the data for frequencies below 10 Hz.  For these
reasons, the use of the weighted acceleration is proposed in the main sections
of these guidelines for assessment of impact due to annoyance of building
occupants and building damage.

     For shock the proposed standard recommends using the vector sum of
the maximum velocity along a set of orthogonal axis.  The maximum velocity
along an axis is that measured at any time during an event.  Such an approach
will be slightly more conservative than only using the maximum weighted
acceleration along the worst case axis.  However, the differences between
the two approaches is not expected to be great (at the maximum they can
only differ by a factor of the square root of 3.

D.4.4  Vibration boundaries with respect to damage categories

D.4.4.1  Vibration caused by shock

     In determining criteria for the onset of vibration damage to buildings,
the proposed standard indicates a number of factors which can affect the
results which are recorded.

These include:

- nature of the soil, clay, or rock, etc.
- stiffness of the building structure
- nature of the vibration, i.e. transient, intermittent, continuous,
  vertical, horizontal, etc.


                                  WEIGHTED  ACCELERATION BASED ON
                                  ACCELERATION  AT  80Hz  OF 0.5m/s2
                                  WEIGHTED  ACCELERATION  BASED ON
                                  ACCELERATION  AT 3Hi  OF  0.5m/s2
        I  I  I  1   I  I   I  I   I  I  L I  L  I   I  I   I  I  I   I
         1.6   2.5    4   6.3    10   16   25   6O   63   100
       .25   2   32   5    8    12.5  20   32  50  9O
                          FREQUENCY (Hz)
FIGURE D—3. Rough evaluation of vibrations of stationary floor vibrations by measurement
          •  of vibration displacements amplitude and frequency
           Note: Amplitude is defined here as the maximum absolute value of the displacement
           of the floor undergoing harmonic motion.


I  0.04
1   0.02





                        I    I
                                          I    T
                                 0      Weighted Accelerations of 1 meter/sec2
                                             •Weighted Accelerations of 0.5 meter sec/
                                              Major damage  v = 7.6 in/sec.

                                              Minor damage v = 5.4 in/sec.

                     •   ^XX       •  VI •V^'
                     ••<.  •••
    Safe blasting criterion

    v = 2.0 in/sec
— .O Bureau of Mines        ^

   c Langefors               Major damage data\

   A Edwards and Northwood                   \
                          *                    •

—  • Bureau of Mines        )                     *\(

   • Langefors               Minor damage data

   A Edwards and Northwood I
                I	I
                           J	I
4    6
40   60    100     200     400 600
                                     FREQUENCY, cps

  FIGURE D—4. Displacement versus frequency, combined data with recommended safe blasting criterion

With these uncertainties in mind, the proposed standard provides recommenda-
tions as to the maximum velocity to prevent damage for each of the three
categories.  These velocities are listed in Table D-3.
                                 TABLE D-3

Limiting values of the vector sum of the maximum velocities  (in three
orthogonal axis) caused by quarry-blasting-vibration in dwellings and
offices in good physical conditions

          Category of Damage                range VR, onset of
          (See Section D.4.2)               damage, in mm/s

                  1                             3 ... 5

                  2                             5 ... 30

                  3                                    100

          These values are based on measured foundation vibration in the
          frequency range from about 3 Hz to about 100 Hz.

The standard cautions that:

     (1)  In the range between 30 mm/s and 100 mm/s the available data  is
          not sufficient to define the nature of the damage without regard
          to the condition, type of structure and foundations.

     (2)  The limits apply only where differential settlement of the
          structure has not been excessive.

     (3)  Special consideration shall be given where buildings are situated
          on a slope or on soils which may be compacted or liquified by

     (4)  When large dynamic displacements are found to exist in the whole
          building or part of it then in addition to the reconmended measuring
          points at the foundation additional measuring points located  in the
          structure shall be used for the evaluation of potential building

The standard recommends that the limits specified in Table D-3 be used  for
the evaluation of vibration effects caused by pile drivers and foregoing
hammers when the time interval between two succeeding blows  is so large that
the vibration of the building due to one blow dissipates before the effects
of the succeeding blow are observed.  Dissipation is regarded as effective
when peak particle velocities have decayed 1/5 from their maximum.

The standard proposed that the values specified in Table D-3 may also be
used to evaluate the effects of vibration in buildings caused by traffic;
however, when shakers and vibration pile drivers are the source of building
vibration, the values given in Table D-3 should not be applied.

Finally, the standard recommends that for the evaluation of transient
response of floors and walls, the vibration limits given for steady state
vibrations may be used in a modified form.  Vfaen there is no danger of
fatigue the limits and values given in Figure D-3 may be increased by a
factor of 2.

D.4.4.2  Steady vibration of buildings
     For steady building vibration, Figure D-3 summarizes the peak velocity
boundaries between the different categories of damage.

D.4.5  Comparison of the recommendation of the proposed standard to the
       recommendations of these guidelines

     The proposed standard recommends that 6 mm/s (5 to 30 mm for shock) be
considered as the upper limit of the threshold of damage.  These velocities
are considerably lower than the 2 in/sec (50.8 mm/sec) that has commonly been
used in this country.  Based on studies such as those shown in Figure D-4,
reducing the threshold from 50 mm/sec to 5 mm/sec does not appear warranted,
however, reduction of the threshold by a factor of 2 does seem reasonable.
All of the data points of Figure D-4 will be covered by use of a velocity of
1 in/sec and it is this velocity that is recommended in the main text of the
guidelines.  Use of a weighted acceleration of 0.5 m/sec2 is consistent with
this velocity and is recommended.

                                 Appendix E

             Example Application of the Guideline Procedures for
                            General Audible Noise
E.1  Proposed highway expansion

     This example (presented briefly in section 2.6) concerns a section of
highway which runs for several kilometers through a suburban area  (Figure
E-1).  The present two lane roadway is operating at close to capacity, and
the proposal is to expand it to six lanes.  Although many factors must be
considered before undertaking such an expansion, only the noise impacts of
the project will be discussed as an illustration of the use of these Guide-
lines.  This example is divided into five sections as follows:

     (1)  Statement of the problem

     (2)  Using the screening diagram

     (3)  Determining the necessary number of figures and tables

     (4)  Completing the figures and tables

     (5)  Conclusions of the noise analysis

E.1.1  Statement of the problem

     From Figure 2 in section 1.3.1, it has been assumed for the purposes of
this example that the only concern is general audible noise that may cause an
adverse impact.  That is, special noises, vibration, and changes in population
location are not anticipated to be problems.  Tables 6 to 9 in section 2.6
document the project impact over the total area (Figure E-1).  However, to
illustrate in some detail the use of these Guidelines, this example focuses
on a small residential section only, as shown in Figure E-2.  Each of the
residential buildings consists of two semi-detached townhouse units, with an
average population of five persons in each unit, or 50 persons in each row of
housing.  Additionally, there are four special situations to be considered:
(1) a school with a student-teacher population of 2,500 in attendance 50 weeks
a year from 8:00 a.m. to 4:00 p.m.; (2) a playground where 400 children play
six hours each day; (3) a park where 160 people relax for one hour each day;
and (4) a church where 185 people meet for two hours each day, and 115 people
meet for one hour each evening.  For a larger project area (such as in Figure
E-1), this amount of detail normally would not be obtained.  Noise contours
would still be plotted, but populations could be estimated from average popu-
lation densities, census counts, or other such sources as discussed in section
2.1.3.  The example is intended to provide an easy-to-follow description of
the Guideline procedures.

covered in

                                                                          |     | School
                             FIGURE E— 1. Highway Expansion Example: Overview
                                                   Area for the
                                                   specific example

           2 Lanes
0.5km l
                                     1    1
                                 Total Population: 550
           Figure E-2. Sample Data Presentation for the Highway Example: Future Levels Without the Proposed Project

E.1.2  Using the screening diagram

     Is an environmental noise analysis necessary, and if so,  what procedures
should be followed?  Begin by examining the Screening Diagram (Figure E-3, and
discussed in section 2.1).  This diagram is helpful for determining not only
whether a noise impact analysis is necessary, but also what type of analysis
should be conducted.*

E. 1.2.1  How to use the screening diagram.   The values for the "existing I^jn(y)"
and the "expected L$n of [the]  project alone" should be obtained at the loca-
tion of the noise sensitive land use nearest the project, or the point where
the impact of the project is likely to be the greatest.  In this example, that
point would be the row of duplexes closest to the highway.  The existing I^n(y)
at the closest noise sensitive point may be obtained either by direct measure-
ment or by use of a suitable traffic noise prediction model** as discussed in
section 2.1.2.  It is assumed that future I^i was obtained through the use of
a prediction model.  Once the values of "existing I^n(y)" and "future L^" are
obtained, they are plotted as coordinates on the screening diagram (Figure E-3).
Their point of intersection (for this example, 60 dB and 65 dB)*** determines
both the necessity for and type of noise analysis that should be conducted.  In
this case, the coordinates fall into the cell that calls for a full noise envi-
ronment documentation.  Therefore, a full noise analysis should be conducted.

E. 1.2.2  Other factors to consider before beginning the noise analysis.  As
discussed in Chapter 2, there are a few other basic issues which should be
considered before beginning the noise analysis:

     o  How many projects or alternatives are we considering?

        We are considering only one, an expansion from two to six lanes.

     o  Will the population of the residential area change in the future?

        No, the population will remain the same (by assumption in this
  *For this example, it is not absolutely necessary to use the screening
   diagram since it is more or less obvious that some increase in noise will
   result, and at levels high enough for a full noise environment documenta-
   tion.  However, if in doubt, the screening diagram is a useful aid.

 **There are in use several models for calculating day-night sound levels
   based on the type of noise source and operational considerations.  These
   models are available from many sources, some of which are listed at the
   end of this appendix.
***At the row of houses closest to the highway,  the existing I^an(y) of 60 dB
   is from Figure E-2, and the predicted I^n(y)  of 65 dB from the project
   alone (the 6-lane highway) is from Figure E-4.

                           JL  •'


                         X160. 65)
                                                DEPENDS ON
                                                FUTURE OF
                                                NOISE SOURCES
                                                            ALL PROJECTS
                                                            SCREENED OUT
                 50      60     70     80      90
                     EXISTING Ldn{y)

         Figure E-3. Screening Diagram: Sample Application

     o  Is this a temporary project,  as defined in section 2.5?

        No, this is a long-term project.

     o  Will the noise of the project change with time (after the
        completion of the project)?

        No, the noise of the project will remain the same.  The
        immediate demand for the added lanes will be sufficient
        to fill them to capacity (by assumption in this example).

     o  Are existing noise levels in the area low enough that
        "environmental degradation" is the only concern (as defined
        in section 2.4)?

        No, most of the area will be exposed to project Ldn(y) greater
        than 55 dB.  This is confirmed by the screening diagram.

E.1.3  Determining the necessary number of figures and tables.  From the
discussion in section 2.2.2.a, it is clear that only three sets of contours
and corresponding tables are required.*

     o  Future levels in the area without the project, i.e., future
        levels from the existing highway.

     o  Future levels due to the proposed project alone, i.e., the six-
        lane highway alone.**

     o  Future levels resulting when the levels from the six-lane highway
        are combined with the levels generated by other noise sources,
        i.e., in this case by residential area activities.
 *As noted in the text, the population density within the residential area
  is not expected to change, nor will the noise from the highway change in
  years subsequent to the proposed expansion.  Because these conditions with
  respect to time are expected to remain constant, additional sets of tables
  and figures are not necessary.  However, if these conditions were to change
  over time, separate sets of tables should be prepared for (a) the first year
  of the commencement of the project, (b) twenty years after the expansion
  (or the latest year for which noise predictions can be reliably made), and
  (c) the worst case year (if different from the preceding two).

**Note that when the proposed action is an expansion of an existing noise
  source, the proposed project alone is interpreted to be the expanded
  project (that is, the six-lane highway), not the amount of expansion (in
  this case, the additional four lanes).

E.1.4  Completing the figures and tables

     The purpose of the analysis is to compare the future noise environment
with and without the proposed project.  This comparison can be divided into
five steps:  (1) drawing the noise contours, (2) determining the base area
and the base population, (3) transferring the data to the tables, (4) calcu-
lating the single-number comparison indices; and (5) noting special popula-

E. 1.4.1  Drawing the noise contours.  As discussed in section E.1.3, three
sets of contours and tables are required.  For purposes of this example, it is
assumed that contours describing the future noise environment in the area
without the highway expansion have been obtained by measurement, since in this
example future levels are the same as existing levels.  It is also assumed
that the noise levels from the future six-lane highway alone have been
obtained from a suitable highway noise prediction model.  These results are
illustrated in Figures E-2 and E-4, respectively.

     To draw contours reflecting the combined future noise environment from
the project levels and levels generated from residential activity requires
additional information.  That is, knowledge of residential area levels in
the absence of any highway noise.  This information can be obtained in two
ways.  An estimate can be made on the basis of population density using
equation 1 in section 2.1.1.  Or, measurements can be taken at a large dis-
tance from the road (for example, where noise from the roadway is no longer
clearly noticeable), as long as the nature of the area is not expected to
change in the future.  The background residential noise levels derived above
(assumed to be about 50 dB in this example) are then combined on a point-by-
point basis with the project alone levels presented in Figure E-4 to derive
contours depicting the total future noise environment (project plus back-
ground) as shown in Figure E-5.*

E.1.4.2  Determining the base area and base population.  As noted in section
2.1.3, the base population is defined as the number of people living in areas
with outdoor noise levels produced by the project alone above a specified
I^n value.  This is called the base 1^.  The base Lfa is determined by
reference to the existing yearly 1^ contours in the residential area
(Figure E-2).  The lowest 1^ in the residential area is about 50 dB near
*For a step-by-step explanation of the combination process, see the discussion
 at the end of this appendix.  Note also that the combination process may
 result in contour lines in other than the desirable five decibel intervals.
 Interpolation may be necessary to plot the information in the five decibel

                                   6 Lanes


0.5km I—L







*   EH
      Fiqure E-4. Sample Data Presentfition for the Hinhwav Examnle: Fnturn Nni«sR I PU«I« fmm tho Prn

                           1    1
                                       6 Lanes
                                i    1
                                       czi   ^   a
          Figure E-5. Sample Data Presentation for the Highway Example: Future Levels from All Noise Sources Combined

the bade row of houses.  Therefore, from section 2.1.3, the optimum base
to use, if possible, in order to define the base population is 40 dB  (that is,
10 dB below the existing I^n(y)).  Next, we examine Figure E-4 which shows
the noise contours from the project alone.  Applying the base I^n =
40 dB, we can derive the base area.  In this example, none of the residents
are living in areas where the outdoor yearly day-night sound levels are below
45 dB (i.e, no people live within the 40-45 dB interval).  Thus, the next best
thing is to effectively define the base area as the area exposed above
Itin(y) = 45 dB.  In this case, there is only one proposed project, and the
base population is 550 people, in an area of 2.37 sq. km. (Fig. E-4).

E.1.4.3  Transferring the data to the tables.  Tabulations of population and
area exposure information are provided in Tables E-1 , E-2, and E-3 for
Figures E-2, E-4, and E-5, respectively.  The values in the tables are derived
by summing the number of people and land area within each five decibel band.  If
a contour line bisected a row of duplexes, the residents were divided between
the noise bands.  For example, from Figure E-5:

               Noise band (I^n)                    Number of People
                    65 - 70 dB                               50
                    60 - 65 dB                               75
                    55 - 60 dB                              150
                    50 - 55 dB                              275

E.1.4.4  Calculating the single-number indices.  For this comparison, three
measures of impact should be considered:  (1) the sound level weighted
population (IMP); (2) the noise impact index (Nil); and (3) the relative
change in impact (FCI).*  The indices LWP and Nil should be computed
for each of the tables (Tables E-1 thru E-3).  For purposes of illustration,
detailed calculations will only be shown for Table E-1.

     Calculation of the level weighted population was based on the values of
the weighting function of equation 2b, shown in Table 3 of the main text.
Thus, using equation 6:

          LWP =  [P(65-70) x W(67.5)] + [P(60-65) x W(62.5)] +
                 [P(55-60) x W(57.5)] + [P(50-55) x W(52.5)] +
                 [P( 45-50) x W(47.5)]

              =  [(0) x (0.194)] + [(25) x (0.116)] +
                 [(150) x (0.064)] + [(375) x (0.032)] +
                 [(0) x (0.015)]

              = 24. 5 ~ 24 people
*In this example, since there are no outdoor exposures greater than L^n = 75 dB,  it
 is extremely unlikely that there will be any at-ear I^q(24) exposure greater than
 70 dB.  Therefore, we need not consider the single-number indices for severe health
 effects (hearing-loss weighted population and average potential hearing loss) as
 discussed in section 2.3.

                                            Table E-l

      Sample Data Presentation for the Highway Example:  Future Levels Without Proposed Project
Yearly Ldn (dB)
Land Area
(Sq km)
Land Area
(Sq km)
Total Land
Area (Sq km)
(See Table E-4)
Level Weighted Population (LWP) =24

Noise Impact Index (Nil) = 0.044

Hearing-Loss Weighted Population (HWP) = 0

Average Potential Hearing Loss (PHL) = 0
                             Corresponds to Fig.  E-2

                             Includes:  o  Self-generated
                                          neighborhood noise.

                                       o  Levels  of noise  from
                                          the existing two-lane

                                           Table E-2

      Sample Data Presentation for the Highway Example:   Future Noise Levels of Project Alone


Yearly L70
Land Area
(Sq km)
Land Area Total Land
(Sq km) Area (Sq km)
1 .3320
(See Table E-4)
Level Weighted Population (LWP) - 25

Noise Impact Index (Nil) = 0.045

Hearing-loss Weighted Population (HWP) = 0

Average Potential Hearing Loss (PHL) = 0
                              Corresponds to Pig. E-4

                               Includes:  o Levels of noise
                                           from the proposed
                                           six-lane highway.

                                           Table E-3

  Sample Data Presentation for the Highway Example:  Future Levels from All Noise Sources Combined


Yearly l#n (dB)
Land Area
(Sq km)
Land Area
(Sq km)
Total Land Situations
Area (Sq km) (See Table E-4)
0.1232 1
0.2465 1,2
0.6716 3
1.3830 4
Level Weighted Population (LWP) = 37

Noise Impact Index (Nil) = 0.067

Hearing-Loss Weighted Population (HWP) = 0

Average Potential Hearing Loss (PHL) = 0
                            Corresponds to Fig. E-5
                            Includes:  o  Self-generated
                                          neighborhood noise.
                                       o  Levels of noise
                                          from the six-lane

     The Noise Impact Index, according to equation 7 in section 2.2.2.C, is
simply LWP divided by the total (base) population.  Thus:

             PTotal=   550  =0.044

     From equation 8, the Relative Change in Impact between the case without
the proposed expansion (Table E-1 ) and the case with the expansion  (Table E-3)
is computed as:
E.1.4.5  Noting special populations.  Special populations do not affect the
calculation of the single-number indices.  However, they are noted on the
figures and tables to give the reader additional information about the
affected area.  As discussed in Section 2. 2. 2. a., the time weighted average
number of people present at these special locations during the year may be
computed as:

     (number of people) x (time the people are present during the year)

                      (number of hours in a year)

For the school in our example where 2,500 students and faculty use the school eight
hours a day, five days a week, fifty weeks a year:

     (2,500 student and teachers) x (8 hours) x (5 days) x (50 weeks)
                          8,760 hours in a year

For the playground where 400 children play six hours each day:

     (400 children) x (6 hours) x (7 days) x (52 weeks)   inn ____,
                          8,760 hours ^ a year         = 1UU P60?16'

For the park where 160 people relax for one hour per person each day:

     (160 people) x (1 hour) x (7 days) x (52 weeks) _ 7
                          8,760 hours in a year      ~

For the church where 185 people meet for two hours each day, and 115 people meet  for
one hour each night, consider the day and night population separately:

     (185 people) x (2 hours) x (7 days) x (52 weeks)   ,_ ___._,
                          8r760 hours in a year          D people.

     (115 people) x (1 hour) x (7 days) x (52 weeks)
                          8,760                      =

The results for special populations are depicted in Table E-4

                                                   Table E-4
                                   Sample Data Presentation:  Special situations

                                  Average population          Range of Ldn/v\
                                  Day          Night    Current      Future^M-th Project
         1.  Elementary school    571

         2.  School playground    100

         3.  Park                   7

         4.  Church                15
0      55-60 dB

0      55-60 dB

0      50-55 dB

5      50-55 dB
60-70 dB

60-65 dB

55-60 dB

50-55 dB

Good acoustic
Evening meetings

E.2  Proposed airport runway addition

     This example concents the addition of a runway at an angle to an existing
runway.  The additon will be completed in 1985.  After 1985, airport operation
and noise will increase until maximum levels are attained in 2001.  Bordering
the airport are a high and a low population density neighborhood.  Both neighbor-
hoods will encroach on the airport between 1985 and 2001, as is shown in Fig. E-6.
The rest of the land near the airport is zoned commercial/industrial.  Although
many factors must be considered before undertaking an airport expansion, only
the noise impacts will be discussed in order to illustrate the use of the Guide-

     This example is divided into five sections:

          (1)  Statement of the problem

          (2)  Using the screening diagram

          (3)  Determining the necessary number of figures and tables

          (4)  Completing the figures and tables

          (5)  Conclusions of the noise analysis

E.2.1  Statement of the problem

    'From Figure 2 in section 1.3.1, it has been assumed for the purposes of
this example that the only concern is general audible noise that may cause an
adverse or potentially severe impact.  That is, special noises and vibration
are not anticipated to be problems.  To illustrate in some detail the use of
these Guidelines, this example focuses on the potential noise impact upon each
of the residential areas in proximity to the airport as shown in Figure E-6.
It is assumed that the number of people living within each of these neighbor-
hoods is computed or estimated from census counts, average population
densities or other methods discussed in section 2.1.3.

E.2.2  Using the screening diagram

     Is an environmental noise analysis necessary, and if so, what analytical
procedures should be followed?  Begin by examining the Screening Diagram
(Figure E-7 and explained in section 2.1).  This diagram is helpful for
determining not only whether a noise impact analysis is necessary, but also
what type of analysis should be conducted.

E.2.2.1  How to use the screening diagram.  The values for the "existing
        and the "expected L,  of [the] Project Alone" should be
         at the location ofnihe noise sensitive land used nearest the
*For this example,Tt is not absolutely necessary to use the Screening
 Diagram since it is more or less obvious that some increase in noise will
 result, and at levels high enough for a full noise environment documenta-
 tion.  However, if in doubt, the screening diagram is a useful aid.


                         POPULATION: 31,689
                          LOW DENSITY
                     ORIGINAL RUNWAY
                                                 POPULATION: 31,689
                                          CQVPLETIONDATE: 1985
   POPULATION: 68,291
                                    Figure-E-6. Sample Data Presentation for the Airport Example: Schematic of the Existing Situation, 1985. and 2001.
                                              (Note  the Addition of the Project Runway and tho Encroachment of the Neighborhoods..)







X  40
                               X (58, 811








                                           ALL PROJECTS

                                           SCREENED OUT
                             GO       70

                           EXISTING Ldn(y,
                     Figure E-7. Screening Diagram

project, or the point where the impact of the project is likely to be the
greatest.  Relevant noise level data is contained within Figures E-8 and E-9 ,
respectively.  Because, as previously noted, airport noise will be greatest in
the year 2001, the point of greatest impact is taken from Figure E-9.  In this
example, the point where the impact of the project is likely to be the
greatest is in the high population density neighborhood, and is designated on
Figure E-9 by the mark "X".  The Itfn(y) at that location from  the airport
alone will be about 81 dB.*  Note that, in this example, the point of greatest
impact is not the closest noise sensitive location.  The noise level at the
corresponding point on Figure E-8 is approximately 58 dB.**
     Once the values of the "Existing ^3n(y)n and the "Expected 1^ of
the Project Alone" are obtained, they are plotted as coordinates on the
screening diagram (Fig. E-7).  Their point of intersection (for this example,
58 dB and 81 dB) determines both the necessity for, and type of, noise analysis
that should be conducted.  In this case, the coordinates fall into the cell
that calls for a full noise environment documentation.  Therefore, a full
noise analysis should be conducted.

E.2.2.2  Other factors to consider before beginning the noise analysis.  As
discussed in Chapter 2, there are a few other basic issues which should
be considered before beginning the noise analysis:

          o  How many projects are we considering?

             We are considering only one option, the addition of a runway
             with an increase in total airport operations.

          o  Will the population of the residential areas remain the same?

             No, between the years 1985 and 2001 both residential areas
             will increase in population and size.  In the existing situation,
             the self -generated (ambient) noise levels in the low density
             neighborhood (780 people per square kilometer) is estimated
             to be 55 dB.  By assumption, the number of people in the low
             density area will remain the same until 1985.  By the year
             2001 , even though the population will increase, the ambient
 *The "Future Noise Levels of Project Alone; 2001"  (Fig. E-9) was determined
  by using a computer model, as discussed in section 2.1.2.  There are
  several models for calculating day-night sound levels based on the type
  of noise source and operational considerations.  These models are
  available from many sources, some of which are listed at the end of this
  appendix.  These models were also used to predict all of the other future
  ttin noise values in this example (except where otherwise noted).
**For purposes of this example, the "Existing I^nfy)"  (Fi9* E~8)
  was assumed to have been determined by direct measurement as
  discussed in section 2.1.2.




             Figure E-8. Sample Data Presentation for the Airport Example: Existing

                      AND PROJECT RUNWAY

                     CORRESPONDS TO TABLE E-6
Fi9ure F,9. Sample Data Presentation for the Airport Example:  Future Noise Levels of the Project Alone; 2001

             noise levels* will remain the sane because the population
             density (people per square kilometer) will remain constant.
             In the high density neighborhood, the existing and future ambient
             is estimated to be 65 dB (7,700 people per square kilometer).

          o  Is this a temporary project, as defined in section 2.5?

             No, this is a long-term project.

          o  Will the noise of the project change with time (after the
             completion of the project in 1985)?

             Yes, as previously stated, the noise levels will be increasing
             between 1985 and 2001, reaching a maximum in the year 2001.
             The increase in noise is caused by an increase in airport

          o  Are existing noise levels in the area low enough that
             "environmental degradation" is the only concern (as defined
             in section 2.4)?

             No, surrounding the airport is a' commercial/industrial area with
             a background ^jn(y) of 55 dB in many locations.  Such a high
             background noise level when plotted on the screening diagram  is
             outside of the cell concerning low noise areas.  (This cell is
             entitled "Possible Noise Degradation Analysis.") '

          o  Are there any special situations (as explained in section 2.2.2),
             such as religious facilities, outdoor auditoriums, schools,
             precision laboratories or hospitals?

             By assumption in this example, there are no special situations.

E.2.3  Determining the necessary number of figures and tables

     From the discussion in section 2.2.2.a,  it is clear that a number of
sets of contours and corresponding tables are required.

          o  Existing noise levels

          o  Future levels without proposed project; 1985

          o  Future levels with the proposed project alone; 1985

          o  Future levels from all noise sources combined; 1985

          o  Future levels without proposed project; 2001
*Yearly day-night sound levels as estimated from population densities
 are discussed in Section 2.1.2.

          o  Future levels with the proposed project alone; 2001

          o  Future levels from all noise sources combined; 2001

E.2.3.1. Completing the figures and tables.  The purpose of the analysis  is  to
compare the future noise environment with and without  the proposed project.
This comparison can be divided into four steps:  (1) drawing the noise
contours; (2) determining the base area and the base population; (3) transferring
the data to the tables; and  (4) calculating the single number comparison  indices.

E.2.3.1  Drawing the noise contours.  As discussed  in  section E.2.3, a number
of sets of contours and tables are required.  Ebr purposes of this example/  it
is assumed that the future noise levels of the project have been obtained from
a suitable airport noise prediction model taking into  account the new generation
of quieter aircraft.  To draw contours reflecting the  combined future noise
environment from the project levels and levels generated from residential
activity requires additional information.  A knowledge of residential area
levels in absence of any other noise (such as highway  or airport noise) is
necessary.  This information can be obtained in two ways.  An estimation  can be
made on the basis of population density using equation 1 from section 2.1.2, or
measurements can be taken at a large distance from  the other noise sources.  In
this example, the former method was used, as discussed in section E.2.2.2.

     The required illustrations are discussed below:

          o  Figure E-8:   Existing levels and future  noise levels without
                           the proposed project; 1985

                           The noise levels in 1985 without the project are
                           the same as the noise levels  in the existing situa-
                           tion.  This lack of change  is by assumption in this

          o  Figure E-10:  Future noise levels of the  project alone; 1985

          o  Figure E-11:  Future levels from all noise  sources combined; 1985

                           The data contained in Figures E-8 and E-10 are
                           combined to create contours of total noise
 *Note that when the proposed action  is an expansion of an existing  facility,
  the proposed project alone is  interpreted to be  the  expanded  project (that
  is, both the old and new runways),  not the expansion alone (in this  case,
  the new runway alone).

**Iogarithmic combinations are discussed at the  end  of this appendix.

          o  Figure E-12:  Future levels without the proposed project; 2001

                           This represents the noise intrinsic to the neighbor-
                           hood which is expanded, the highway, and levels of
                           noise from the increased usage of the existing
                           single runway.

          o  Figure E-9:   Future noise levels of the project alone; 2001

          o  Figure E-13:  Future levels from all noise sources combined; 2001

                           The data contained in Figures E-9 and E-12 are
                           combined to create contours of total noise
**Logarithmic combinations are discussed at the end of this appendix,

Figure E-10. Sample Data Presentation for Airport Example: Future Noise Levels of the Project Alone; 1985


    55      INCLUDES:
                   • NOISE FROM THE HIGHWAY

                   • SELF-GENERATED NOISE FROM
                      BOTH NEIGHBORHOODS

                   • NOISE FROM BOTH THE ORIGINAL
                      AND THE PROJECT RUNWAYS
              CORRESPONDSTb TABLE E-8
Figure E-11. Sample Data Presentation for the Airport Example: Future Levels from All Noise Sources Combined: 1935



             INDUSTRIAL ZONE


Figure E-12. Sample Data Presentation for the Airport Example: Future  Noise Levels without the Proposed Project; 2001




Figure E-13. Sample Data Presentation for the Airport Example: Future Levels of Noise
          from All Noise Sources Combined; 2001
                                             E  28

E.2.4.2  Determining the base area and base population.  As noted in
section 2.1.3, the base population is defined as the number of people living
in areas with outdoor levels of noise produced by the project alone above a
specified L^ value.  This is called the base 1^.  The base L^n is
determined by reference to the existing yearly LJn contours in the residen-
tial area (Figure E-8).  From Figure E-8, the lowest L^ in the residential
area is about 60 dB.  Therefore, from section 2.1.3, the optimum base I^n to
use, if possible, in order to define the base population is 50 dB (that is,
10 dB below the existing I^n(y)).  Next, we examine Figure E-9 which shows
the noise contours from the proposed project alone in 2001 (the year with the
greatest impact).  Applying the base L^ = 50 dB, the base area should be
the area within the 50 dB contour.  However, since the level of ambient noise
is assumed to be at least 55 dB, such a base area would extend far beyond the
boundaries of Figure E-9.  In fact, the land within the 55 dB contour also
extends well beyond the boundaries of the figure.  Since there are by assump-
tion no other residential areas in the vicinity, it is not necessary to choose
such a large base area.  Instead, it would be more logical to consider as the
base population the people residing within the residential areas of interest.

E.2.4.3  Transferring the data to the tables.  Tabulations of population and
the area exposed are provided in Tables E-5, -6, -7, -8, -9, and -10, correspond-
ing respectively to Figures E-8, -9, -10, -11, -12, and -13.  The values in the
tables are derived by summing the number of people and the area of land within
each five decibel band.  For example, in Figure E-13:

           Noise Level (L^n)                  Population
                      80+                             1,233
                    75-80                            30,799
                    70-75                            30,346
                    65-70                             3,942
                    60-65                             1,971
                    55-60                                 0
                    50-55                                 0

E.2.4.4  Calculating the single number  indices.  For this comparison, five
single number  indices should be considered:  (1) the sound level-weighted
population  (LWP);  (2) the noise impact  index  (Nil); (3) the relative change
of impact (RCI);  (4) hearing loss-weighted population (HWP); and  (5) the
average potential  hearing loss (PHL).   The first three indices  (explained
in section  2.2.2.b) concern the range of general adverse noise  effects; the
latter two  (explained in section 2.3.2) concerns noise levels which possibly
may cause severe health effects.  The indices EWP, Nil, HWP and PHL should
be computed for each of the tables.

                                                   Table E-5

                                   Sample Data Presentation for Airport Example:

                                   Future Levels Without Proposed Project;  1985

Yearly I#n
n 60-65
° 55-60

31 ,689
Commercial Total Land
Employees Area (sq. km.)
0 0
0 0
1,550 8
7,264 42
10,822 19.2
6,814 135.8
41,678 95.0
68,128 300.0
Residential Land Commercial Land
Area (sq. km.) Area (sq. km.)
0 0
0 0
0 8.0
3.14 38.86
4.48 14.72
0 135.8
0 95.0
7.62 292.38
Level Weighted Population  (LWP) = 5,787
Noise Impact Index (Nil) = 0.183
Hearing Loss-Weighted Population (HWP) = 0
Average Potential Hearing Loss  (PHL) = 0
*Large number of people in the 65-70 dB band
 is attributable to high ambient noise levels
 (non-aircraft related) in the high density
 residential area (See section E.2.2.2).
Corresponds to Figure E-8
Includes:  o  Highway noise
           o  Self-generated noise from both neighborhoods
           o  Noise from the original runway
Note:  Single number indices are not computed on the
       basis of industrial/commercial employees.

                                                   Table E-6

                                   Sample Data Presentation for Airport Example:

                                Future Noise Levels of Project Alone; 2001
Yearly L^n

Total Land
Area (sq. km.)
Residential Land
Area (sq. km.)
Commercial Land
Area (sq. km.)
Level Weighted Population (LWP) = 24,498
Noise Impact Index  (Nil)- 0.359
Hearing Loss-Weighted Population (HWP) - 408
Average Potential Hearing Loss (PHL) =0.09
Corresponds to Figure E-9
Includes:  o  Noise from both the original and the
              additional runways.

                                                   Table  E-7

                                   Sample Data Presentation for Airport  Example:

                                Future Noise Levels of Project Alone;  1985
Yearly l$n
ra 60-65

Total Land
Area (sq. km.)
Residential Land
Area (sq. km.)
Commercial Land
Area (sq. km.)
Level Weighted Population  (LWP) = 4,094
Noise Impact Index   (NII)= 0.129
Hearing Loss-Weighted Population (HWP) = 0
Average Potential Hearing Loss  (PHL) = 0
Corresponds to Figure E-10
Includes:  o  Noise from both the original and the
              additional runways.

                                                        Table E-8

                                        Sample Data Presentation for Airport Example:

                                      Future Levels From All Noise Sources Combined;  1985

Yearly L^n

Total Land
Area (sq. km.)
Residential Land
Area (sq. km.)
Commercial Land
Area (sq. km.)
     Level Weighted Population (LWP) = 5,964
     Noise Impact Index = 0.188
     Hearing Loss-Weighted Population (HWP) = 0
     Average Potential Hearing Loss (PHL) = 0
Corresponds to Figure E-11
Includes:  o  Highway noise
           o  Self-generated noise from both neighborhoods
           o  Noise from both the original and the project
Note:  Single number indices are not computed on the
       basis of industrial/commercial employees.

                                                   Table  E-9

                                   Sample Data Presentation for Airport Example:

                                   Future Levels Without  Proposed  Project;  2001
Yearly L
                                                   Table  E-10

                                   Sample Data Presentation  for Airport Example:

                                Future Levels From All Noise Sources Combined;  2001
Yearly Ldn
IJJ 60-65
K 55-60

Total Land
Area (sq. km.)
Residential Land Commercial Land
Area (sq. km.) Area (sq. km.)
0.16 11.84
4 16.7
4.85 44.3
2.56 181.74
10.88 22.97
0 0
0 0
22.45 277.55
Level Weighted Population (LWP) = 24,597
Noise Impact Index  (NII)= 0.360
Hearing Loss-Weighted Population (HWP) = 408
Average Potential Hearing Loss (PHL) = 0.09
Corresponds to Figure E-13
Includes:  o  Highway noise
           o  Self-generated noise from both expanded
           o  Noise from both the original and the project
Note:  Single number indices are not computed on the
       basis of industrial/commercial employees.

     A summary of the assessment results is presented below.  Calculation
of the level weighted population was based on the weighting function of
equation 2b, shown in Table 3 of the main text.

                                                          From Data In Example
	Year	LWP	Table Number

     Present (no project)                   5,787                  E-5
     1985 with the project                  5,787                  E-5
     1985 with the project                  5,964                  E-8
     2001 without the project              12,647                  E-9
     2001 with the project                 24,597                  E-10

     The Noise Impact Index, according to equation 7 in section, is
simply the LWP divided by the total population.

                                                          From Data In Example
     Year                    %>tal    LWP       Nil         Table Number
Present (no project)        31,689      5,787     0.183            E-5
1985 without the project    31,689      5,787     0.183            E-5
1985 with the project       31,689      5,964     0.188            E-8
2001 without the project    68,291     12,647     0.185            E-9
2001 with the project       68,291     24,597     0.360            E-10

     The Relative Change in Impact is calculated from equation 8.

                    From Data on               From Data in
                       Airport                    Example
     Year   LWPa  Table Number     LWPb   Table Number         RCI
     1985   5,964       E-6          5,787        E-5              0.0306
     2001  24,597       E-8         12,647        E-7              0.9449

     As previously noted, the indices representing potential hearing damage
risk are not similar to the other three indices.  In order to emphasize the
severity of the health problems caused by high noise levels, a separate
severe health effects single number index is used.  As discussed in Section
2.3.1, for areas with an L^ of 75 dB* or above, the following information
should be estimated if possible:  the population spending time out-of-doors;
the length of time they are out-of-doors; the actual noise levels while they
are out-of-doors.  For this example, only the populations within the residen-
tial areas are being considered.  Those people in other impacted areas of the
metropolitan area are assumed to remain indoors (because the metropolitan area
is entirely commercial/industrial, so they are not subjected to the noise.

     The area for study of possible severe noise impact shown on Figure E-13
is expanded and shown as Figure E-14.
*As pointed out in Section 2.3.1, the L^ should be used only to identify
 potential problem areas.



  Figure E-14. Sampla Data Presentation for the Airport Example: Severe Health Analysis Area


              The additional information required is now assumed to have been collected
         through additional estimates or survey work.   It is listed on the next page as
         Table E-ll.

              The next step is to obtain the at-ear outdoor L   values of the expo-
         sure instead of the L,_ values.  The best way is to tike additional noise
         measurements.  A much less preferable way, as explained in Section 3.2.1,
         is to use the approximation:

                                     Leq = Ldn(daytime) " 3
         This approximation may be used if the difference between the daytime and
         nighttime levels is the typical one where the nighttime level is approxi-
         mately 4 dB lower.  It has been assumed in this example that sufficient
         measurements were taken to determine that the day-night difference in L
         of that area is a typical one.  Thus, the 3 dB correction term has been^
         applied and the results entered in Table E-ll.

              Next, since not all residents are exposed to exactly eight hours of out-
         door noise, the data in Table E-ll are adjusted to the appropriate Le
         values, using Table E-12.  For example, the population of 7,521 e>
         4 hours of L  /4N = 73 dB has the equivalent of an L  /8\ exposure of 70 dB.
         These L  's nave been entered into the cells of Tablt^E-ll in brackets
         ([ ]). T$ow the number of people exposed to various levels of noise over
         L	,0. = 75 may be read from Table E-ll:
                                             Population Distribution as a
                    L.QJ                      Function of 8-hour L

                     76                           (2542 + 691)  = 3,233
                     77                           (694 +287)   =   981
                     78                                            176
                     79                           (225 + 97)   =   322
                     80                                              0
                     81                                          	67
                                              TOTAL              4,779
              PHL may thus be computed from equation 13 as:
                                               P(L  )i . H(L
                                       PHL =  -    -
         where P(L  ) is the population distribution as a function of 8-hour L^^
         (shown above), H(L  )  is the corresponding weighting function shown onTTable
         5 of the main textTor all ?(!•),  where H(  ).  ^ O, and P-y^-i  is the
         population for the severe healfn effects area, i.e. , the sunrof all people exposed
         to an L  ,g. greater than 75 dB.  Using the information in Table E-ll, the PHL is:

                                                         Table E-11

              Sample Data Presentation  for  Example the Airport:   Information required to calculate the PHL

                                                  Residential Population
                                                                     [Exposures in Lg  (8)]
                                                                     Time Outdoors
                                 4 hr
                                                           6 hr
           8 hr
         12 hr
u* ,
                                                                                       Corresponds to Figure E-14

          Table E-12
Conversion of I^q(X) to Leq(8) and Leq(24)
                     -  9
                     -  6
              I«q(8) -  3 = I«q(24) " 8
                     -  2 =
                          = Leq(24) - 5

                        3 = Lt>al?A\ - 2

P.    =  [[(67)x(0.9)]+[(0)x(0.625)]+[(322)x(0.4)]+[(176)x(0.225)]+[(981)x(0.1)]+[(3233)x(0.025)]]

                          PHL - ,!?5f-= 0.09

         2.5  Conclusions of the Noise Analysis

              The purpose of the noise analysis  is  to compare  the number of people
         affected by  the noise levels with and without  the project.  To do this
         comparison,  the single number indices are  used.  In tabular form, they are
         shown  in Table E-13.  These  indices  show:   177 more people  in the residential
         areas  will be fully adversely affected  by  the  noise created by the project
         in 1985 (when compared with  the  existing situation or with  1985 without  the
         project); 18,810 more people in  the  residential areas will  be fully  adversely
         affected by  the noise from the project  in  2001 (when  compared with the existing
         situation);  11,950 more people will  be  fully adversely  affected by the noise
         created by the project in  2001 (when compared  with 2001 without the  project).
         In addition, the PHL for the year 2001  increases from 0 to  0.09.



         How are  the noise  levels on Figures E-2 and E-4 combined?

         Using Table E-14,  combine  the  noise levels shown on Figures  2  and  4  by determining
         the difference between  levels  at the  same point, and adding  the  appropriate  amount
         from the table to  the higher level.

                                                     Table E-13
                                                Summary of Airport  Example
Without the Project
and 1985 2001
5,787 12,647
0.183 0.185
- -
0 0
With the Project
1985 2001
5,964 24,597
0.188 0.360
- -
0 0.09
Between the
Existing &
1985 with
the Project
Between the
Existing &
2001 with
the Project
Between 1985
without the
Project and
1985 with
the Project
Between 200
without the
Project and
2001 with
the Project
177 11,950
For 1985: 0.0306 For 2001: 0.9449

                              Table E-14
                                             Number of decibels
        Difference between                     to be added to
        Levels in decibels                      Higher Level
     For example, the noise level at the first row of houses is 60 dB in
Figure E-2 and 65 dB in Figure E-4, a difference of 5 decibels.  Table
E-14 shows that for a difference of 5 dB, approximately 1 dB should be
added to the higher level in order to derive the total level.  Therefore,
the noise level at the first row of duplexes in Figure E-5 is computed as
66 dB.  Similarly, the noise level contour at the second row of houses is
64 dB.  In summary:

               Add to
Duplex Row
Figure E-2
Figure E-4
1.2 or 1
1.5 or 2
2.1 or 2
2.6 or 3
2.6 or 3
2.1 or 2
1.5 or 2
1.2 or 1
1.5 or 2
1.2 or 1
1.5 or 2
Figure E-5



Highway Noise

            U. S. Environmental Protection Agency, Office of Noise Abatement
                  and Control.  Comparison of Highway Noise Prediction Models.
                  EPA 550/9-77-355.  May, 1977.

            U. S. Environmental Protection Agency, Office of Noise Abatement
                  and Control.  Highway Noise Impact.  EPA 550/9-77-356.
                  May, 1977.

Aircraft Noise

           U. S. Environmental Protection Agency, Office of Noise Abatement
                 and Control.  Calculation of Day-Night Levels (L.J
                 Resulting From Civil Aircraft Operations.EPA 550/9-77-450.
                 January, 1977.

           U. S. Department of Transportation, Federal Aviation Administration,
                 Office of Environmental Quality.  Federal Aviation
                 Administration Integrated Noise Model.  April, 1978.

           U. S. Department of Transportation, Federal Aviation Administration,
                 Office of Environmental Quality.  Federal Aviation
                 Administration Integrated Noise Model Version I.
                 FAA-EQ-78-01.  January, 1978.

           U. S. Department oif Defense, Air Force Aerospace Medical Research
                 Laboratory.  Community Noise Exposure Resulting From Air-
                 craft Operations;  Computer Program Description, AMRL-TR-
                 73-109, November 1974.  Computer Program Operator's Manual,
                 AMRL-TR-73-108, July 1974.  NOISE MAP 3.4 Computer Program
                 Operator's Manual, AMRL-TR-78-109, December 1978.

Rapid Transit Noise

           U. S. Department of Transportation, Urban Mass Transportation
                 Administration, Office of Technology Development and
                 Deployment, Office of Rail and Construction Technology,
                 Noise Rating Criteria for Elevated Rapid Transit Structures.
                 Report No. UMTA-MA-06-0099-79-3.  May, 1979.

Transmission Line Noise

            Ccmber, M. G., and L. E. Zaffanella.  "Audible Noise."
                  In:  Transmission Line Reference Book 345 KV and above.
                  Published by Electric Power Research Institute,
                  3412 Hillview Avenue, Palo Alto, CA 94304, 1975.

            Driscoll, D. A., and F. G. Haag.  "Prevention and Control of
                  Environmental Noise Pollution in New York State."
                  In:  Proceedings of a Workshop on Power Line Noise as
                  Related to Psychoacoustics.  Published by the Institute
                  of Electrical and Electronics Engineers, Inc.,
                  345 East 47th St., New York, NY 10017, 1976.

            Molino, J. A., G. A. Zerdy, N. D. Lerner, and 0. L. Harwood.
                  "Psychoacoustic Evaluation of the Audible Noise from
                  EHV Power Lines."  To be published.

            Kolcio, N., J. DiPlacido, P. M. Dietrich.  "Apple Grove 750 KV
                  Project — Two-Year Statistical Analysis of Audible Noise
                  from Conductors at 775 KV and Ambient Noise Data."
                  In:  IEEE Transactuibs ib Oiwer Appartus and Systems, 1977.

Taken together, these references provide enough information to predict
transmission line noise under different conditions.  Of the four references,
the first is the most useful.


                           RESPONSE TO COMMENTS ON

     Guidelines for Preparing Environmental Impact Statements on Noise
                           Background and Summary

     In 1972, the Office of Noise Abatement and Control of the U.S. Environ-

mental Protection Agency requested the Committee on Hearing, Bioacoustics

and Biomechanics (CHABA) of the National Academy of Sciences to prepare a set

of procedural guidelines for evaluating the environmental impact of noise.

In response, CHABA formed Working Group 69 for this purpose, chaired by

Dr. Henning von Gierke.  The Working Group completed its task in early 1977,

submitting the report entitled Guidelines for Preparing Environmental Impact

Statements on Noise (referred to in the rest of this report as the Guidelines),

     During the summer of 1977, copies of the Guidelines were distributed to

federal agencies and other interested parties, with a request for comments.

On June 30, 1978, a request for comments was published in the Federal Register

(43 Fed. Reg. 28549).

     As of 22 September 1978, a total of 45 letters, many with detailed

attachments, had been received in response to these requests for comments.

Thirty-three different organizations sent comments, of which 13 were indus-

trial concerns or trade organizations, 4 were state agencies, and 16 were

federal agencies (Table 1).  In addition, comments were received either by

EPA or by the National Academy of Sciences from three individuals (Kryter (2),

Ward, and Young), from the chairman of Committee E-33 (on Environmental

Acoustics) of the American Society for Testing and Materials, and from

Transport Canada.

                                   Table 1
                          Commenting Organizations
     Columbia Gas Systems Service Corp.  (Columbia Gas)a
     Consolidated Edison Co.  of N.Y.  (Con Ed)
     Dames and Moore
(3)b Edison Electric Institute (EEl)
     Ford Motor Co.  (Ford)
     General Motors  Corp. (GM)
(2)  Motor Vehicle Manufacturers Association (MVMA)
     Northern States Power Co. (NS Power)
     R & F Coal Co.
     United Engineers
     Vibra-Tech Engineers
     Wyle Laboratories


     Kentucky Dept.  of Mines  and Minerals (Ky.  Mines)
(2)  New York State  Dept. of  Environmental Conservation (NY Envir)
     New York State  Dept. of  Transportation (NY DoT)
(2)  New York State  Public Service Commission  (NYPSC)

     Dept.  of Agriculture,  Agricultural  Research Service (Ag Res Serv)
(2)  Dept.  of Commerce,  National Bureau  of Standards,  Institute for Basic
        Standards (NBS)
     Dept.  of Defense,  Deputy Assistant  Secretary,  Energy Environment and
                          Safety (DoD)
                       Air Force,  Aerospace Medical Research Lab.  (AMRL)
                       Army,  Construction Engineering  Research Lab (CERL)
     Dept.  of Housing and Urban Development,  Office of Environmental
        Quality (HUD)
     Dept.  of Interior,  Deputy Assistant Secretary  (Dol)
     Dept.  of Labor,  Occupational Safety and Health Admin.  (OSHA)
     Dept.  of Transportation, Assistant  Secretary,  Environment Safety and
                                 Consumer Affairs (DoT)
                              Federal Aviation Administration (FAA)
                              Federal Highway Administration (FHWA)
                              Urban Mass Transportation Administration (UMTA)
     National Aeronautics and Space Administration  (NASA)
     Naval  Construction Battalion Center (NCBC)
     Postal Service
     Veterans Administration  (VA)
.This abbreviation will be used to reference the comments.
 Number of separate letters sent,  if other than one (1).

     Many of those commenting raised excellent points regarding the clarity

or consistency of the publication.  Others questioned the technical material

or the reasoning in the Guidelines.  The major issues addressed by the com-

ments are:

     (1)  the title and purpose of the Guidelines,

     (2)  the inclusion of impulse noise and vibration,

     (3)  the treatment of noise increases in quiet areas, including effects

          on wildlife and other animals,

     (4)  the use of yearly day-night sound level,

     (5)  the interpretation and empirical basis for calculations of level-

          weighted population, and

     (6)  the organization and clarity of the document itself.

The responses to many of these are inter-related conceptually, although

they are written in separate and distinct sections in the body of this report.

     We begin with the view that the comments about the title and purpose are

correct:  noise is only one of many impacts to be considered in any environmental

assessment.  The Guidelines are intended to help people treat noise impacts in

order that a decision-maker can understand them better, and can incorporate them

more fully in the comparative evaluation of all of the factors in the decision.

     To facilitate such a treatment of noise impacts, the material in the

Guidelines has been reorganized around four topics:  (1) the general adverse

response to audible noise; (2) hearing damage and other severe health effects

from higher level audible noise; (3) environmental degradation in rural and

natural areas; and (4) special considerations for ultrasound, infrasound,

impulse noise, and vibration.  For the first topic, the approach in the

Guidelines is followed, and an effort is made to improve comprehension of

"level-weighted population."  For the second,  the weighting function for loss

of hearing is based explicitly on at-ear exposure, which may be determined by

8-hour Lgq, 24-hour Leq, or outdoor L(jn depending on circumstances.  For

the third topic, we are using the concept of environmental degradation but not

providing a procedure to calculate a  single-number index for it.  For the

fourth topic, we have tried to clarify when each of the special noises is to

be considered, and the reasons for the approach taken.

                       Response to Comments on Guidelines
Background and Summary


Overall Assessments
     Favorable Comments
     Unfavorable Comments

General Comments
     A.  Simplification
     B.  Example Applications
     C.  Accountability

     1.  Comments on Title

Front Matter
     2.  Membership of the CHABA Working Group

Chapter I:  Introduction
     3.  Inclusion of wildlife
     4.  EIS
     5.  Purpose of Guidelines
     6.  Inclusion of Vibration
     7.  Mandatory Guidelines?
     8.  Conflict With Other Procedures
     9.  Federal Environmental Impact Review Process
    10.  Rationale
    11.  Hearing and Day-night Average Levels
    12.  Table 1-2
    13.  General Audible Noises
    14.  Temporary Changes
    15.  Feasibility of Forecasts
    16.  Chapter IV Omitted

Chapter II
    17.  Noise Due to Project
    18.  Fig. II-l
    19.  Noisy Sources



                              Contents  (Cont'd)
Chapter III
    20.  Fig.  III-l                                                      26
    21.  Project Life Span                                               26
    22.  Five-year Intervals                                              27
    23.  Noise Impacts Only                                              27
    24.  Simplified Procedures  for Temporary  Change                       27
    25.  Uncertainties                                                   28

Chapter IV
    26.  Any Change in Impact                                             29
    27.  Level of Detail                                                 29
    28.  Noise Measures:   YDNL                                            30
    29.  Other Noise Measures                                             36
    30.  Base YDNL                                                       36
    31.  Estimating DNL                                                  38
    32.  Summary Tables                                                  40
    33.  Property Line Limits                                             41
    34.  Land Use Categories                                              42
    35.  Special Situations                                              42
    36.  Residential Average Number                                       43
    37.  Tables IV-2 and IV-3                                             43

Chapter V
    38.  CSEL for Impulse Noise                                          45
    39.  Measuring CSEL                                                  46
    40.  CDNL                                                            47
    41.  Infrasound                                                      47

Chapter VI
    42.  Percent Highly Annoyed                                          49
    43.  Table VI-1:  Criterion Levels  for
         Various Land Uses                                               50
    44.  Table VI-2, 3, and 4:  Summaries of
         Human Effects                                                   51
    45.  Response Independent of  Source                                  52
    46.  Degradation of Environment                                       54
    47.  Effects of Impulse Noise                                        55
    48.  Infrasound                                                      56
    49.  Vibration Criteria                                              57
    50.  Blast Noise                                                     58

Chapter VII
    51.  Single Number Impact Index                                       59
    52.  Level Weighted Population                                        60
    53.  PLH and Other Abbreviations                                      61

                              Contents  (Cont'd)
    54.  Population Weighted Loss  of Hearing                             62
    55.  Normalizing Weighting Function                                  63
    56.  Sound Level Weighting Function                                  63
    57.  Weighting Function for Loss of  Hearing                          64
    58.  Change in LWP                                                  64
    59.  Temporary Noise Environments                                   65
    60.  Vibration Impact Index                                         65

Chapter VIII
    61.  Noise Re Other Impacts                                         66
    62.  Structures Exposed                                             66
    63.  Sound Level Weighted Area                                      66
    64.  Descriptive Qualitative Evaluations                             67
    65.  New Population Exposed                                         67

Appendix A
    66.  Terminology                                                    69

Appendix B
    67.  Exponential Functions                                          69
    68.  Population Weighted Hearing Loss                                70

Appendix C
    69.  ISO/TC 108/SC 2/WG 3                                           70
    70.  Figure C-4                                                     71


     The comments are presented in the order one would encounter  them  in  read-

ing through the Guidelines.  Comments on  the same topic from several sources

have been listed together, and appear the  first time a topic is encountered.

In addition to the many comments on specific material, nearly half of  those

commenting on the Guidelines included an overall assessment of the report.

These are summarized first.  Also, several general comments were  offered  which

do not pertain to a specific part of the Guidelines.  These are listed and

answered following the overall assessments, before turning to the more spe-

cific material.

                              Overall Assessments
     Of the 33 different organizations which sent comments, 16 made explicit

statements about the overall usefulness or appropriateness of the Guidelines.

Eight of these were favorable; eight were negative, calling for delay or

withdrawal of publication.  Extracts or precis of all 16 of these general

comments are listed here.  Most of these contained particular questions or

comments as well as the overall appraisal.  It should be kept in mind in

reading these that the praise was not without reservation, nor the criticism

without reasons.

     Of the eight favorable comments, six were from federal government

agencies, and two from industry.  No federal agencies sent unfavorable over-

all comments.  Half of the eight suggestions for delay or withdrawal of publi-

cation came from public utilities—three from the electric power industry and

one from a gas company.  Of the remaining four, two were from New York State

agencies, one from another state agency and one from a manufacturer.


     "OSHA applauds this report and agrees that the suggested guidelines will

be useful in a wide variety of applications."  (Occupational Safety and Health

Adminis trat ion)

     "We agree with the approach...; the Guidelines should provide consider-

able assistance to the Navy..."  (Naval Construction Battalion Center)

     "I believe that it will be suitable for its intended function...[if]

the preparing group will update it as needed."  (Ametek)

     "We think it is an excellent report, representative of the best state-

of-the-art today."  (National Bureau of Standards)

     "As a technical guideline, this document provides excellent direction

for evaluating noise impacts."  It should be published only as a guideline

and not as a regulation.  (U.S. Postal Service)

     The Guidelines "is a fine document which contains the necessary noise

impact criteria in one place."  "...it is an important document which should

get wide dissemination."  (Dames and Moore)

     "Overall we believe that the guidelines can serve a useful purpose in

assisting agencies in their assessments of the environmental effects of

noise."  (National Aeronautics and Space Administration)

     "As you can see, most of our comments are minor in nature due to the

overall high quality of the Report."  (Urban Mass Transportation Adminis-


Unfavorable Comments

     "We have serious reservations regarding the validity of the Guidelines

and urge that U.S. EPA not adopt them as a 'uniform national method for

noise impact assessment* (page iv)."  (New York State Department of Environ-

mental Conservation)

     "EEI recommends that EPA not publish or endorse the Guidelines for the

reasons cited in the attached comments."  (Edison Electric Institute)

     "NSP recommends that EPA not publish or endorse the Guidelines for the

reasons stated."  (Northern States Power Company)

     "...we must strenuously recommend that this document, which is in effect

a regulation, be required to have included an economic impact statement prior

to adoption."  (Kentucky Department of Mines and Minerals)

     "The docket should remain open for six months to allow for the completion

of these existing studies [at Battelle Columbus Laboratories] so that they may

be made available to EPA and the scientific community."  (Ford Motor Company)

     "Con Edison recommends that the Guidelines not be published or endorsed

by the U.S. Environmental Portection [sic] Agency (EPA) for the following

reasons."  (Consolidated Edison Company of New York)

     "...there are a number of serious flaws related to the reliability of

the impact assessment which I believe make it unwise to publish the Guidelines

in their present form."  (State of New York Department of Public Service)

     "...Columbia strongly recommends that the Committee withdraw its Proposed

Guidelines until after EPA prepares an L^n standard..., prepares an Economic

Impact Statement..., submits it for public review and comment, and publishes

a final recommended L,jn standard."  (Columbia Gas System Service Corporation)

                              General Comments

A.  Comment (Simplification)

     Dol, HUD, and UMTA all say that the Guidelines are complex and hard to

follow.  UMTA complains about the foreword references, e.g., from Chapter  IV

to VI and VII, which make the material hard to follow.


     The revision attempts to overcome these problems, primarily by putting

all material necessary for the treatment of a specific type of noise envir-

onment in one section.

B.  Comment (Example Applications)

     NASA, DoD, and NYSOoT all state that if EPA intends these Guidelines

to be widely used, they should demonstrate the merit of the procedures with

case histories, real-world applications, or field tests, which can show their

usefulness in the decision-making process.  Dames and Moore echoes this in

their observation that the Guidelines are "not as applicable to the site

specific EIS as I would like," i.e., to those prepared "for local developers,

industrialists, or mine owners."


     More realistic examples are included in the revised version, to cover a

fairly wide range of possible applications.

C.  Comment (Accountability)

     New York Envir. points to "... a fundamental weakness in the environmental

impact assessment process, it ... fails to hold the project sponsor account-

able for the assumptions and predictions upon which the decision was based."


     This is, of course, a weakness of the EIS process, not simply of these

Guidelines.  Hopefully, the Guidelines (in both original and revised forms)

improve the probability of reasonably accurate assumptions by making those

assumptions explicit, at least in the noise assessment.  In controversial

projects, weak assumptions are usually brought out by project opponents during

public hearings.  Thus, while the Guidelines do not solve the accountability

problem, they do help to minimize it.


1.   Comment
     Four federal agencies (HUD, DoD, DoT, FHWA) suggested that the existing

title of this publication is not appropriate, or is misleading.  There

was considerable agreement on the reasons for this assertion among the four.

EIS's do not often address noise alone (FHWA, HUD).  Noise is only one of

many environmental concerns in an EIS (FHWA, HUD, DoD, DoT).  Noise assessments

are done in contexts other than simply an EIS (FHWA).

     DoT proposes that "the title be revised to suggest that it provides guide-

lines for noise analysis and assessment as part of the comprehensive EIS

process."  Two other agencies suggested new titles:

     "Environmental Assessment of Noise" (FHWA); and

     "Guidelines for Noise Impact Evaluation in the Environmental Impact

Analysis Process" (DoD).


     These points are all well-taken.  Revision of the title is necessary.

The title of the revised document is "Guidelines for Noise Impact Analysis."

                                Front Matter

2.  Comments on the Membership of the CHABA Working Group

     Some concern was expressed that the composition of the Working Group

which prepared the Guidelines was too limited.   Five different writers made

this point, in varying degrees of strength.   "The Guidelines were prepared

by a select group of 13 people who do not represent a consensus of the

acoustical profession.  Almost half of Working  Group 69 are employed by the

Federal government, and about one-quarter work  for a government contractor.

Working Group 69 did not include a representative from a local or state

pollution control agency, nor any representation from the manufacturing,

construction, transportation, or utility industries.  We recommend that

the Guidelines be reviewed by a consensus standards writing organization,

such as the ASTM E-33 committee..." (EEI; a very similar comment from Con

Ed).  "The WG was clearly loaded with persons who have already been ordained

in the Total Energy religion."  "I do hope...that in the future the composi-

tion of Working Groups is made broad enough to  guarantee discussion of

alternative theories and approximations..." (Dixon Ward).  "The EPA should

refer the comments submitted to the docket to a committee comprised of per-

sonnel with technical expertise (expanded from  the original CHABA group)"

(Ford).  "[N]one of the members of CHABA working group 69 is an established

authority in the area of blasting or impulsive  vibrations ..."  "Almost every

recommendation made in the proposed standard ... is contrary to the recommen-

dations of the U.S. Bureau of Mines and other authorities after 20 years

practical research ..." (Vibra-Tech).


     The primary reason that EPA has requested comments on the report of

the CHABA Working Group was to provide an opportunity to consider additional

viewpoints and expertise in revising the proposed Guidelines.  By carefully

reviewing comments submitted, and revising the Guidelines accordingly, any

problems of the kind raised are hopefully overcome.  There would seem little

reason to rely on a consensus standards writing organization, since this docu-

ment is not a standard.  It is a compilation of procedures for improving com-

munication about noise impacts, in hopes that decision-makers will be able to

incorporate them more fully into decisions.  (Part of the cause of the mis-

understanding of this intent is perhaps addressed in comment 5 regarding the

purpose of the Guidelines.)

                          Chapter I;  Introduction

3.  Comment (p. 1-1, para 1:  inclusion of wildlife)

     "We recommend deletion of Sections VI-E, VIII-A-1, and references to

'wildlife* and 'animals' on pages 1-1, III-2" (NS Power, EEI).  "One cannot

assume that identical noise criteria can be utilized for people as well as

for wilderness animals.  While there is little technical literature on this

subject, our terrestrial observations indicate that common species, i.e.,

squirrel, deer, seem unaffected by man's noises whereas certain others such

as bald eagle and timber wolf are highly sensitive to all disturbances by

man" (NS Power).   "Protecting wildlife and domestic animals from noise vibra-

tion levels equivalent with human inhabitants is absurd" (Ky Mines).  "There

is so little information on the effects of noise on animals, and effects

assumed in the Guidelines are gross speculation.... We recommend that EPA

sponsor additional studies on the effects of noise effects of noise [sic] on

animals..." (EEI).  "A statement is even made to essentially ignore the impact

of ultrasound (>20 kHz) on animals, even though it is well known that dogs are

very sensitive to ultrasound" (United Engineers).  "The proposal...'exposure

level identified to protect man will also protect animals' appears reasonable

but would be unduly restrictive if it were established as a maximum noise

restriction for animals" (Ag Res Serv).


     If these Guidelines are to serve as an aid to people who have found

noise to be a significant consideration in a proposed project, the document

should be comprehensive.  It should also, however, be based on known facts,

and these are admittedly scarce for animals and wildlife.  Qualitative

statements about the effect of noise on animals are simply all that can be

made.  (Examples are the statements NS Power makes about eagles and wolves,

and United Engineers makes about dogs.)  If sensitive animal populations

are present in a project area, they should not be overlooked in the noise

assessment.  On the other hand, if there is no information that suggests

noise problems for animals, it is not reasonable to assume they are more

sensitive to noise than people are.  In general, then, the effects of noise on

animals will be listed as a "special situation" in the tabulation of impacts,

and will be described qualitatively in the text of the noise impact analysis.

(See the EPA publication Effects of Noise on Wildlife and Other Animals

(NTID 300.5, Dec. 1971).)

4.  Comment  (p. 1-1, para 1:  EIS)

     The reference to Environmental Impact Statements (EIS) is both incomplete

and misleading.  "[T]he report does not give the uninitiated reader any clue

to what an EIS is, its objectives, and what it stems from" (HUD).  The Guide-

lines "fail to take into consideration the mandates of NEPA in regard to

when a 'detailed statement1 or 'Environmental Impact Statement' is required"

(Columbia Gas).


     These comments are both correct.  The Guidelines were written for people

involved in the NEPA process (or similar ones), so that a detailed descrip-

tion of that process was not required in the document.  Therefore, the re-

vised version has not expanded the discussion of the NEPA process.  Instead,

it has provided a broader statement of the rationale for these Guidelines.

(See comment 5 also).

5.   Comment  (p.  1-1, Purpose of Guidelines)

     Most of those who suggested altering the title also suggested revising

the purpose of the Guidelines, as did others (FHWA, DoT, OoD, FAA).  The

purpose should read, "This report offers guidelines for the environmental

assessment of noise" (FHWA).   "[T]he purpose of the document should be to

assist those evaluating the environmental impact of proposed actions and

to address noise as a contributing factor..." (DoD).


     This section has been rewritten to make exactly those points.  The

revised section is similar to the following statement by CEQ:  "The NEPA

process is intended to help public officials make decisions that are based on

understanding of environmental consequences, and take actions that protect,

restore, and enhance the environment" (Council on Environmental Quality,

Proposed amendments to regulations; Federal Register, June 9, 1978, p. 25232

Sec. 1500.l(c)).   The purpose of the revised version, then, is to present

procedures which can be used to express noise impacts in terms which are easily

understood by decision-makers, so that those impacts can be fully incorporated

in the comparative evaluations which constitute the decision.

6.   Comment  (p.  1-1, para 2:  inclusion of vibration)

     Three writers, all from the electric power industry (Con Ed, EEI, NS Power),

suggested that the material on infrasound, ultrasound, and vibration should be

deleted from the Guidelines,  because "the 1972 Noise Control Act gave EPA

authority to control noise only within the audible frequency range" (Con Ed).

Vibra-Tech also requested deletion of this material,  on the grounds of the

Working Group composition.  (See Comment 2.)


     NCA 1972 defines environmental noise as "the intensity, duration, and the

character of sounds from all sources" (Sec. 3.11) but does not further define

'sound'.  The question of whether or not there is sound if it cannot be heard

is an age-old semantic one, and no attempt is made to answer it here.  Since

the revised Guidelines are offered simply as an aid to decision-making, and

their use is not mandatory, we feel it is entirely appropriate to offer

guidance for a variety of "non-audible situations."

7.  Comment  (p. 1-1, para 2:  mandatory Guidelines?)

     "One may assume that by adoption the guidelines take the character of a

'standard* and their use [becomes] mandatory" (NASA).  EEI expands on this

notion:   "EPA endorsement of the Guidelines would constitute de facto

regulation of environmental noise emissions from fixed facilities, which

appears  to be beyond the authority granted EPA."  Several of the comments

seemed to indicate a fear that these Guidelines would become mandatory,

thus entailing double work where other forms of noise assessment are also

required (FHWA, Ky Mines), or that they would be interpreted as noise level

standards by field personnel (Con Ed, EEI).  Ky Mines also stated that since

the Guidelines are in effect regulations, there should be an economic impact

statement prior to their adoption.


     Given the purpose stated in comment 5, it should be obvious that these

guidelines are merely an aid for quantifying noise impacts, and should not

be considered as standards.  A discussion in this regard is included in the

preface. It should be noted that the methods in the Guidelines are reasonably

consistent with the procedures followed by most agencies at present.  Because

the Guidelines are offered as only an aid to planning and decision-making, and

are not mandatory requirements,  no economic impact statement is necessary.

8.  Comment  (p. 1-1, last para:   conflict with other procedures)

     Columbia Gas suggested that  "If the Committee desires the Proposed

Guidelines to be used by federal  agencies," delete the statement that

"Agency's guidelines would be expected to take precedence over those pro-

posed in this document" "where there is conflict."  On the other hand,

NASA and FAA supported the statement in the Guidelines,  and the FHWA goes

even further, saying that this is an important point which "should be stressed

in the summary."  NYSDOT points  out that because of this provision "the

Guidelines would have little if  any application to our Department's activities"

since NYSDOT noise studies "presently follow procedures  outlined in Federal

Highway Program Manual 7-7-3." Dol on the other hand observes that "... lack

of uniformity complicates efforts in seeking to comply with all applicable

noise control requirements," and  refers specifically to  quite different

requirements in Oregon and Washington.


     The response to comment 7 is appropriate here also.  Because, the Guide-

lines are for the most part consistent with procedures of other agencies, the

question of making them mandatory should not arise.   Furthermore, it should

be noted that many components of  the procedures used or  cited within the Guide-

lines have been standardized both nationally and internationally. The revised

Guidelines strongly recommend relating impacts to people, and suggest how to

do this, but they do not require  a specific form of analysis.  The point made

by NASA, that "agency procedures need to retain sufficient flexibility to

be applicable to special conditions not addressed in these guidelines," is

a reasonable compromise position.

9.  Comment  (p. 1-2, para B.2; and Table 1-1:  federal environmental impact
             review process)

     DoT suggested that Table 1-1 should be deleted or modified, "since the

EIS review process varies widely among Federal agencies."  DoD, on the other

hand, suggests, "Additional drafting is needed to properly describe the NEPA



     Expansion of the context of the guidelines should obviate the need

for a detailed discussion of the NEPA process. The procedures are useful

any time an analyst needs to investigate noise impacts.  Specification

or description of all such occasions is unnecessary.  Section I-B and Table

1-1 have been deleted.

10.  Comment  (Sec. I.C.:  Rationale)

     FHWA agrees completely with the rationale as stated.  NBS would add the

phrase "consistent with reasonable accuracy" to "should be uniform and as

simple as possible."


     These comments are accepted.  Other portions of the rationale have

been revised to be consistent with the new title and purpose.

11.  Comment  (p. 1-4:  hearing and day-night average levels)

     DoD notes that the  'Levels' document "does not endorse L^n as a suitable

technique for assessing the potential for hearing loss" but recommends Len


instead.  NBS raises similar concerns in conjunction with pages VI-8 to VI-11 of

the Guidelines,  asking,  "Ought not the 75 dB threshold be identified as an Leq(8)?

...Also, the need to use Leq(24) rather than L^n should be made explicit."


     Von Gierke's letter (6 Dec 77) responding to the DoD comment indicates

that Working Group 69 used L
stated in the note to Figure VI-2 in the original version), the user of the

revised Guidelines may wish to use outdoor L
longer than six months, but less than 10 years.   Dol suggests that the

short term/long term division be at 2 years,  and that the upper limit on long

term temporary changes be 25 years.  Dames and Moore, on the other hand,

suggests 3 years as the upper limit of temporary change.  Dol's reasoning is

that "it is common practice in the electric utility industry to install a

temporary transformer unit for a period of from six months to two years;" and

that "most electrical facilities have a useful service life of at least 25

years." NBS points out a further inconsistency in that the Guidelines state (on

p. III-9) :  "Temporary changes in which the daily DNL is less than 90 dB may

modify, simplify, or eliminate full Noise Environment Documentation...."


     First, since the Guidelines do not seem to employ the distinction

between long and short term anywhere, the revision will speak only of

temporary changes.  Second, 2 or 3 years seems a reasonable upper limit on

a usual temporary change, with the possible exception of some major construc-

tion efforts.  For example, Dol speaks of "temporary transformers" in place

for up to 2 years.  The 90 dB (Ldn) cut-off is a typographical error,

which should be 70 dB (to be consistent with Fig. II-l).  However, a special

cut-off for temporary projects is not used in the revised document, consistent

with the new screening approach (see response to comment 18). Temporary

changes will call for a less involved noise impact analysis, as discussed in

comments 24 and 59.

15.  Comment  (p. 1-9, IV-8:  feasibility of forecasts)

     For permanent changes, the Guidelines require "a projection of popula-

tion and land use...over a twenty-year period."  Dol points out that "the


linear nature of [electrical] transmission facilities makes it impossible

to make the long range projections of adjacent populations and land uses,"

and suggests that such facilities "should be exempted from evaluations of

the effects of noise on projected adjacent populations and land uses."  DoD

says the requirement for ^10% accuracy (p. IV-8) is unreasonable in view of

the 20-year horizon.


     In fact, transmission lines are not unique in finding it difficult

to make 20-year population projections.  Even urban planners have great

difficulty (and no guarantee of success) in doing this, particularly at the

level of areal detail often required for noise impact analysis.  Nonetheless,

it is important for a noise impact analysis to try to ascertain the long-

range implications of the project.  Twenty years is suggested as an appropriate

time span.  To avoid these prediction problems, the revised version suggests

using, where possible, the local community's land use plans or zoning designa-

tion for the areas in question, whenever such plans exist.  Making this change

overcomes the criticisms (from HUD), that this document disregards community

goals, objectives, and plans.  In locales without such plans, forecasts should be

attempted.  When even this is not possible, the reasons need to be stated, since

it is almost always desirable to specify long-range as well as immediate


16.  Comment  (p. 1-10:  Chapter IV omitted)

     An introduction to Chapter IV is absent from this summary (UMTA).


     The revised guidelines will contain appropriate introductory material.


                                 Chapter II

17.  Comment  (p.  II-l:  noise due to project)

     The Guidelines states that the screening chart  "is based on the relation

between the existing noise environment and the noise environment after the

project is completed."  NBS suggests that "after the project" was intended to

say "due to the project."


     NBS is correct that the expected future noise  levels for the screening

procedure (and for the base YDNL for impact assessment  (p.  IV-4)) are meant

to be calculated for the project alone,  not for it  added to the existing

levels.  (Note that impact analysis is based on total resulting levels, how-

ever.  See response to comment 18.)  This seemingly  minor editorial comment

points up the fact that it is extremely difficult to write these guidelines

concisely to apply equally well both to projects which  produce noise and to

projects which reduce noise..  The revised version discusses the screening

diagram only in the context of projects which may increase noise.  This, of

course, does not preclude use of the'methodology presented in the revised

Guidelines for those projects which may reduce noise.

18.  Comment  (Fig. II-l)

     There were over a dozen separate comments on the rationale for or clarity

of Figure II-l, the Screening Diagram, and the text  material explaining it.

The bulk of the comments are addressed to the levels identified for screening.

FHWA says the "levels shown are too low," because all highway projects would

exceed the 40 dBA lower limit.  Dames and Moore say  the screening "philosophy

encourages degradation of rural population density areas."  Dol points out


that the 40 dB threshold does not take into account areas where the existing

level is low, and the unique nature of wild lands, where even minor increases

in noise can "disturb the values for serenity possessed by these areas"

(Ool).  DoD simply asks "What is the basis for the 40 decibel standard on

page II-l and the 50 and 60 dB standards on page II-3?" This is echoed in

Con Ed's contention that these thresholds will become de facto standards

(discussed in comment 7).  Both Con Ed and EEI state that the figure

conflicts with NCA 1972 in that local governments are to set local noise

limits.  Con Ed also asserts that "there should be no need for a noise

impact statement when the expected noise emissions are below...local

ambient sound levels."  EEI concurs.  Columbia Gas questions whether the

screening levels are consistent with NEPA Sec 102(C) which requires a

detailed statement on '"...major federal actions significantly affecting

the quality of the human environment."1  In line with this, Columbia notes

the reliance on the Levels document as the foundation for this screening

procedure, and asserts that the Levels document is being misinterpreted.

Dol also picks up on the apparent reliance on the Levels document, noting

that these "very stringent" levels were "derived based on the most sensitive

human response" and "also have a margin of safety associated with them."

     A number of comments address the clarity of the diagram.  DoD says it is

"difficult to understand," and that it "must be proven both scientifically and

practically."  NBS points out that both axes have the same label—expected

YDNL—which is confusing.  NASA points out that the double labelling of the

horizontal axis is confusing, which NBS would correct by putting the relation-

ship between population density and YDNL elsewhere.  NASA also requests more

discussion of "the rationale for exempting potentially noisy projects from

detailed noise environment documentation."


     The shaded portion in the upper right of Fig. II-l, labelled "measurement

required" drew a number of comments.  Columbia Gas notes that the title of the

figure says measurements "are advised," so at the very least, the shaded area

should be so labelled.  HUD, NASA, and NBS all ask why measurement is needed

in this particular area.  HUD goes on to suggest that "measurements should be

required based on the situation...(e.g. cases where there are aggregate

sources, where available models are not so reliable, etc.)."


     This diagram is obviously central to the content of the Guidelines, and it

is important, in responding to these comments, to keep in mind the totality of

the Working Group's approach.  The starting point for clarification of Figure

II-l is the labelling of the axes.  In his 6 Dec 77 reply to USAF comments, von

Gierke states, "This figure would be clearer, perhaps, if the ordinate was

labelled 'Expected Yearly Ljn of the Project Being Evaluated* and the abscissa

was labeled 'Existing or Expected Yearly L,
       PROJECTS '
        .DEPENDS ON
        -FUTURE OF
                       EXISTING 1^)

     Given these labels, we must next address the levels themselves.  Von

Gierke's letter also states, "[T]his figure is based on the observation that

if the existing background is 10 dB or more [greater] than the level produced

by a new noise source, the new noise level will increase less than one-half a

decibel." This explanation obviously applies to the diagonal line in the ori-

ginal Figure 11-1, which extends from the point representing project levels

of 40 and existing levels of 50 up to the point representing project levels of

55 and existing levels of 65.  This explanation of the 10-dB-down diagonal is a

reasonable answer to Con Ed's assertion that if project levels are anywhere

below ambient, no noise impact analysis is needed.  Con Ed and EE1 apparently

did not consider the fact that such situations would produce up to a 3 dB

increase in the ambient.  The reasons for the 10 dB difference are stated in

the revised text.  There may well be cases where departure from a rigid appli-

cation of such screening levels is warranted, which the user can explain given

a knowledge of the reason for those levels.

     However, Columbia Gas rightly questions, in essence, whether a one-half

decibel increase constitutes "significantly affecting the human environment."

Von Gierke's letter goes on to say, "Thus the screening values were set con-

servatively so if a mistake was made, it would be requiring some additional

noise analysis for a project that turns out to cause little environmental

impact with respect to noise." Working Group 69, then, was operating, in

effect, on the basis of those sections of NEPA which mention "...the critical

importance of restoring and maintaining environmental quality..." (Sec.

101(a)), and which call for attaining "the widest range of beneficial uses of

the environment without degradation..." (Sec. 101(b)(3)).  This seems an

appropriate perspective for a document which is intended to be a comprehensive

guideline for assessing noise impacts.  If there is any possibility of such

impacts, the method should be applicable.

     In light of the comments on rural areas or wild lands, there seems a

good argument for extending the 10-dB-down diagonal below new project levels

of 40 dB.  Von Gierke's letter (6 Dec 77) states, "To some extent the 40 dB

is arbitrary.  It was based on...practical judgment and experience....  We

would be interested in specific cases where this was too high."  The comments

appear to identify such cases, and to make the point that the screening dia-

gram should not ignore such situations, even though subsequent analysis may

show the impact to be quantitatively small.  The revised screening diagram

allows for assessment of noise impact even for low project noise levels in

such cases, in recognition of the fact that quiet is a national resource, and

is one of the features of natural areas which makes them uniquely valuable.

It should be noted that such areas are valuable because they are rare:  this

region of the screening diagram will not often be encountered.  The revised

diagram uses a cut-off level of 45 dB for proposed projects in these quiet

areas.  This level was set by combining the 10 dB down reasoning from the

previous discussion with the level of 55 dB (L. ) identified as protective

of public health and welfare in the Levels document.  This 45 dB (L, ) level

is not, however, intended as a standard or regulatory level in any sense. The

expanded discussion of the rationale for the 10 dB difference should overcome

the tendency to read specific examples as standards, the way DoD apparently

did in their comment.  The fact that the screening line is based on environ-

mental degradation in no way conflicts with any local noise limits.  The

purpose of an environmental assessment is to allow a decision-maker to

evaluate comparatively effects on many factors, along with costs and, for

example, services provided. The fact that noise is below local limits does not

mean a project is good: some other project may offer lower noise levels, along

with different levels of other effects,  costs, and benefits.  The same reason-

ing applies to the FHWA comment:  noise  effects should be a consideration in

all highway proposals if noise levels will increase as a result of the project.

The fact that the levels are not high enough to warrant federal cost sharing

of noise abatement efforts is not an indication that no noise impacts exist.

     The termination of the diagonal line at project levels of 55 dB needs

some justification, which the existing document does not provide.  In the re-

vised version, the diagonal line has not been terminated at 55 dB.  Instead,

projects with levels above 55 dB, but below the diagonal, are indicated as

needing documentation depending on the likely future emissions of the existing

noise sources.  In that region of the diagram, potential abatement of other

noise sources may reduce existing levels, making the proposed project the major

noise problem.  This argument is made explicit in the revised version.  It

should be recognized that screening in this region needs to be done with some

judgment.  For example, the probability  of significant noise reductions of the

other sources during the life of the project could be stated.  If it is low

(as in central cities, where a multitude of noise sources are present), then a

project in this region of the screening  diagram (new Fig. 3) may not require

further noise environment documentation.  If existing levels are 70 dB and are

not expected to be abated, and one works through the analysis for a project

producing 55 dB, the result will be a finding of no impact.  Since there seems

little point in doing a lot of work for  a preordained outcome, the diagonal

line has been extended above project levels of 55 dB.  It terminates at 75 dB

project levels because these are levels  at which health effects are of concern.

     It is difficult to address the DoD comment that this diagram must

be scientifically and practically proven.  Hopefully, a better explanation

of the rationale for the diagram will suffice.

     On the matter of the advised or required measurements, there is no

available explanation for the location of the shaded portion.  In the

revised version, that shaded portion of the diagram is omitted.  Instead,

field measurement is identified as the preferred approach, and there is

a discussion of situations in which other estimation procedures might

reasonably be used.

     In keeping with our aim to further clarify and simplify the Guidelines,

we have deleted from the screening diagram the distinction between permanent

and temporary projects, such that above the screening line (new Fig. 3), a

full noise environment documentation is recommended for all projects.  This

decision is simply a recognition that nowhere do the Guidelines describe what

a "modified NED" is to consist of.  It seems clear that documentation for a

temporary project would have at most only a two year, rather than twenty year,

horizon, and that this would be the time-saving modification apparently de-


19.  Comment  (p. II-3, noisy sources)

     NASA suggests that some guidance would be useful on how to estimate the

effects of existing especially noisy sources.


     Agreed.  In the revised version, references are made to existing (federal)

computer or manual techniques for estimating airport or expressway noise.

                                 Chapter III

20.  Comment  (Fig. III-l)

     HUD says, "Since this [figure] is the key to the process, the structure  of

the guidelines should follow [it] more closely and the outline should be  keyed to

the Figure."  HUD also states that this figure does not recognize community plans,

goals, or objectives.  Animals appear as the top item in the detailed considera-

tions, and HUD suggests people should come first.  (HUD's comment was made about

Chapter VIII, but since that chapter is based on the order of this figure, it is

addressed here.)  NBS says, "A.4. shows exit of Ldn < 55 before examining possible

infrasound.  Also B:  no path if only infrasound or ultrasound is a problem."


     The organization of the revised report has been based on this diagram, as

suggested.  Some re-ordering of the diagram was necessary, both to correct

the specific errors noted by the comments, and to improve the organization

of the Guidelines.  Community plans have been brought into the Guidelines,

but it proved better to do so in the text rather than with one or two words

in this diagram.  (See comment 15.)

21.  Comment  (p. III-3:  project life span)

     DoD points out that some actions may be obsolete or eliminated in  10 years

or less, so that a 20 year evaluation is unnecessary.


     Agreed.  The impact assessment should be confined to the life-span of the

project, with the exception that when one is comparing alternatives with  different

life-spans, the longest one should be applied to all of the alternatives.


22.  Comment  (p. III-3:  five-year intervals)

     United Engineers writes, "The evaluation of  impact after 5,  10,  15,  and  20

years of operation can be speculative at best, and  is definitely  time-consuming.

...[I]t would appear to be less useful than potentially misleading."


     In many instances, the noise impacts may not occur at  the same time  for

all alternatives.  In such instances, it is important to know the relative

timing of the impacts in making a decision among  the alternatives.  In

other cases, the comment is correct.  The revised version documents the

reasons for this suggestion of evaluating impacts at five year intervals,

and suggests that judgment be used regarding its  usefulness  in each case.

23.  Comment  (pp. HI-4,5:  noise impacts only)

     FAA points out that this section appears to  place noise as the sole

criterion in an EIS, which is contrary to statutory requirements.


     Agreed.  This has been revised, as discussed in comments 4 and 5.

24.  Comment  (p. III-9:  simplified procedures for temporary change)

     NBS states, "Reference to Chapter VII is circular:  Chapter  VII  refers

the reader back to this paragraph."


     The Guidelines are not clear on when to modify, simplify, or eliminate

the noise environment documentation, simply saying  those things may be done

"as is reasonable" for "temporary changes in which  the daily DNL  is less  than

90 dB."  The 9.0 is merely an uncorrected typographical error, and should be

70, to conform with Fig. II-l.  The revision (comments 14 and 18) suggests

that all projects, temporary and permanent, conduct the same documentation,

which may be simplified as Chapter VII describes for temporary projects.

25.  Comment  (p. 111-10:  uncertainties)

     The Guidelines states that in all cases a discussion of probable uncer-

tainties must be provided, and suggests using two estimates:  the best and

worst cases.  NBS suggests "stating explicitly that use of 90% confidence

intervals (if it can be estimated)" is acceptable.   R&F Coal quotes this part

of the Guidelines and says, "The report then proceeds to recommend specific

analytic methods which are predicated on 'worst* case data and assumptions."


     Given the nature of the uncertainties identified in section III.G. of

the Guidelines, there is almost no possibility of the kind of statistical

precision which would allow one to speak of 90% confidence intervals.

Inclusion of such a phrase in the Guidelines would therefore be misleading.

Users are of course welcome to use confidence intervals where feasible.

     R&F Coal's comment is made primarily  in the context of the impulse noise

and vibration material, where there is in  fact some validity to it.  The text

has been revised to make explicit the point which R&F Coal makes,  and to

encourage the project proponent to conduct a best case analysis as well.

                                 Chapter IV

26.   Comment  (p. IV-1:  any change in impact)

     FAA asserts that the threshold levels which require a  full noise  analysis

are  unreasonable:  "... to describe any change in the  impact of noise  on



     Although there is some merit to this assertion, it  is  unfortunately

impossible to utilize a higher cut-off level without also requiring  a  full

noise analysis.  That is, the alternative is to say (for example)  a  full noise

analysis is necessary only if more than 200 people are exposed to  an increase

not  less than 3 dB (or the LWP increases by 50).  But  in order to  know whether

these limits are met, one must perform a noise analysis.  FAA's comment seems

impossible to incorporate.  (See also response 18.)

27.   Comment  (p. IV-1:  level of detail)

     FAA also asserts that "the level of detail required is unreasonable

(section IV)."  NYSDOT says, "the methods and procedures ... are much  too

complex and cumbersome for normal project development  activities."


     Although this may be true for projects whose impact turns out to  be

minor, it is felt that for projects with major impacts anything less than

is suggested in the Guidelines would not suffice.

28.  Comment  (p.  IV-2:  noise measures:  YDNL)

     Fourteen different organizations commented on the assertions culminating

in the identification of YDNL, the yearly day-night average sound level, as

the primary measure to be used.  None of the comments were laudatory—from 5

industrial, 2 state,  and 6 federal agencies, plus 1 individual.  The comments

dealt with A-weighting, I measurement problems, and yearly averaging.

     NASA asserted that A-weighting is not "the best indicator of complaints

and annoyance due to low frequency noise (30-60 Hz)."  Dol notes that Oregon

uses "octave band frequency requirements."  Dol also suggests that the Guide-

lines are not useful for electric transmission facilities, because of the

"continuous 120 Hz low frequency hum of transformers and intermittent corona

noise from high-voltage lines."  NYPSC and NS Power make similar comments.

Dixon Ward notes that "the use of A-weighted sound level and of total A-weighted

sound energy...are all quite arbitrary and challengeable."  MVMA also makes the

point that the use of A-weighting is an assumption which "has not been widely

accepted," and which, according to one JASA article they cite, "is inadequate

for estimating speech interference....  MVMA does not propose a more suitable

scale..., but simply points out the inadequacy....  This example demonstrates

a need for further research."

     MVMA objects to Leq/Ldn» on the grounds that "some short duration, but

highly annoying sounds" are not properly treated by them, and cites as a case

in point the EPA background document for proposed bus noise regulations.

Ford makes a similar point, saying that use of a single noise descriptor,

while commendable "... because of its simplicity, may in fact be misleading

and result in factually inaccurate noise assessments."  MVMA also sent

copies of a report by Battelle Laboratories which they assert substantiates

their position that "the energy average noise approach does not appropriately

account for the diverse effects of different noise sources."  FHWA simply

states "our opposition to the Ljjn descriptor and the reasoning for our

position is well known" to ONAC.  NYSDOT prefers "a critical hour approach."

     With respect to measurement, both Con Ed and EEI note that there are

no consensus measurement standards for Leq and L^n, and Con Ed asserts

that publication of a document relying so heavily on such descriptors is

therefore premature.  Dol also wants more information on measurement tech-

niques.  Both EEI and DoD take exception to the stipulation that nighttime

levels need to be measured from 0000 to 0700 and 2200 to 2400 hours of one

calendar day.

     Yearly averaging was objected to for both pragmatic and theoretical

reasons.  Both United Engineers and NY Envir question how annual L
than continually changing noise descriptors.   NYPSC asserts that there is no

basis for assuming that people mentally average a year's noise "before deciding

whether to become annoyed—."  DoD notes that the military has always used

the average "busy day" concept, and requests  that this exception be written

into the Guidelines.

     The final comment on L,jn was from UMTA,  which expresses concern that

NEF, Leq, and CNEL received little or no treatment in the report.


     Data suggest that A-weighting is reasonable to use in almost all cases.

For that reason it has been widely accepted.   MVMA1s second point, that

knowledge is not perfect, is a truism.  One cannot, however, refrain from

action until complete knowledge is available.   Certainly, some simplifica-

tions are necessary for a general document such as this, and in making them,

some sensitivity to subtleties is always lost.  It is our opinion that the

advantages of simplicity outweigh the loss of potential information.  Further,

if one has certain knowledge that there are special characteristics of the

proposed noise source, and that these constitute a problem above and beyond

the A-weigh ted level, one should certainly include a discussion of these

points in the assessment.  The revised version makes this explicit, and the

original version recognizes it in the use of  C-weighting for impulse noise.

However, there are as yet no definitive published studies documenting the

added impact in the community of pure tone (or low frequency) components.

Without such it is clearly premature to advocate the use of special proce-

dures for such noises.

     As to the problems with L
Their traffic noise prediction method requires information on traffic

density. As this varies over the day, additional predictions are required.

Accurate convolution requires complete specification of exceedance levels

for individual sources.  Until it can be shown that added information on

community response can be gained from this extra effort, there seems little

reason to undertake it, and to nullify all of the gains in simplification

(including uniform noise prediction approaches) which have been made in

recent years.

     MVMA's comment on the use of other than L
reasonable assumptions, and have been made explicit in the revision.  In an

effort to reduce the problems of measurement, the revision explains yearly

day-night average sound level as being analogous to the traffic engineering

concept of annual average daily traffic.  That is, they are meant to be

those levels measured under average conditions, or, if conditions vary

during the year, weighted averages of levels at the different times of

year.  To reduce the proliferation of new sets of initials, the revision

uses Ldn(y) and yearly L{jn instead of Lrfny and YDNL.  This is consis-

tent with the treatment of time in other metrics (e.g., Leq(8))» and with

the Acoustic Terminology Guide previously developed by this office.

     With respect to problems of estimation in rural areas, there simply

is not enough known about noise levels in such areas, since measurement is

routinely conducted only in the absence of wind, rain, and other natural

sounds.  Whenever possible, measurement of existing levels is recommended.

     The comment by Dames and Moore is not fully understood.  If the daily L
     DoD's use of average busy day, rather than annual average, means at most

an increase of 1-1/2 dB in the resulting value, and would appear to be

consistent with the underlying rationale.  There seems no need to explicitly

acknowledge their variation, nor to prohibit it, particularly since the Guide-

lines are not to be construed as standards.

     UMTA's request for a wider variety of descriptors runs counter to the overall

purpose of these Guidelines which is to present noise impacts in a way which

enhances their understandability and promotes their consideration in decisions.

The smaller the number of technical descriptors used, the more likely is this

to be accomplished.

29 .  Comment  (p. IV-3:  other noise measures)

     Both NYPSC and NASA call for further elaboration of the use of average

sound level over specific, short time periods.  This need for elaboration is

also evident in HUD's question:  "Why introduce SEL for individual noise

events...?"  MVMA's comment is also appropriate here, on the use of other than

    by EPA in the background document for the proposed bus noise regulation.

     This seems a reasonable request.  The revision includes a clearer

statement of when the shorter term metrics are appropriate (instead of "In some

instances it is desirable...").   It also indicates where to get information so

that these metrics may be utilized in ways equivalent to the LWP approach.

30 .  Comment  (p. IV-4:  base YDNL)

     The discussion of 'base YDNL1 caused 8 comments, best summarized by that

from HUD:  "We find discussion of base YDNL confusing."  In particular, HUD

questions the need for a common base among alternatives, and the requirement

for a 20 dB range of YDNL.  The reliance on the Levels document to justify the

55 dB limit is objected to by both Columbia Gas and Dol (see comment 18 also).

The statement that base YDNL will never be greater than 55 dB draws 3 complaints.

DoD quotes EPA ONAC as having "stated that 'L
rather Chan to introduce 55 dB, or a 20 dB range,  to define the base Ldn(y)«

This would answer comments from HUD, Columbia Gas, Dol,  EEI, DoD, and Dames

and Moore.  It will still be emphasized in the revision, however, that a

range of noise levels needs to be considered.  Twenty decibels is an example

of a good range to use.  Practical feasibility, raised by DoD and CERL, may make

it advisable to relax this base level or range in many instances.  Aircraft

noise prediction models, for example, are much more uncertain below an L
Table B-3 from the 'Levels document1 except that the levels have been decreased

by five dB without comment."  In its place they "urge use of the residual sound

level," as in ISO-R1996.


     The estimation procedure is derived in Population Distribution of the

United States as a Function of Outdoor Noise Level (550/9-74-009); this is

referred to in the revised Guidelines.  That derivation identifies the standard

error of the estimate as roughly 4 dB.  The ^10 dB mentioned in the Guidelines

is therefore somewhat misleading:  the 95% confidence interval about an estimate

would be less than +6 dB.  The reliability of the relationship is approximated

by the correlation coefficient of 0.723 between L
levels is strongly advised for most projects.  If measurement is not feasible,

then this table and its underlying equation are available.  If these indicate

levels above 55 dB, and especially when levels are above 65 dB, the project

analyst needs to consider the probability of future reductions of those levels

due to noise abatement efforts.  If there is such a probability, the impact

of the project noise should be calculated against these reduced future levels,

rather than against existing levels.  (See the discussion of use of the screen-

ing diagram above project levels of 55 dB (L^), in the response to comment

18; and the discussion of appropriate comparisons for the single number index,

in the response to comment 51.)

32.  Comment  (p. IV-9:  summary tables)

     HUD suggests it is not really necessary to include all land areas in

the tables, just the noise sensitive uses, both existing and planned.

     EEI makes a similar comment, proposing that evaluations be made only

beyond the projects' property line, since buffer zone acquisition is an

accepted way of abating noise in the community.  EEI demonstrates with an

example that "a very small change in the estimated source sound power level

results in very large changes in the area between consecutive isopleths,"

e.g. a 3 dB increase from 100 to 103 dB (re: 10    watts) leads to a 100%

increase in the area between the 55 and 60 dB contours.  The fact that this

could lead to an over- or under-estimation of the true impact "serves to

cast doubt on the validity of the methodology."


     Certainly EEI's suggestion, that impacts be tabulated only beyond the

property line, is a reasonable one, and is made clear in the revised text.

(An airport, for example, may attempt to contain the L
which it will not.  The second premise is unlikely to hold.  Most noise limits

represent a trade-off between costs and benefits, for the general community.

They will therefore almost always be set at levels at which some adverse noise

effects may still result.  The Guidelines, however, are meant to result in the

noise impacts being described in a manner which allows a decision-maker to

make those trade-offs anew for each potentially noisy project.  There will be

instances, of course, where the appropriate decision-maker deems it not

worthwhile to make these trade-offs anew, and to use the existing local noise

limits.  The Guidelines are merely pointing out that even where noise limits

do exist, major projects often warrant a reconsideration of the trade-offs.

34.  Comment  (land use categories)

     HUD is concerned that the categories used do not consider local plans or

community goals, and also expresses interest at the determination that the

residential population is those "who sleep for four or more hours per day in a

residential area" (p. IV-10).


     These categories are offered as examples.  The revised text emphasizes

use of local categories.

35.  Comment  (p. IV-10:  special situations)

     CERL notes that the separate treatment of special situations (such as

schools) can lead to excessive work, and gives the example of O'Hare Airport,

and a base DNL of 55, for which the "area is measured in one or more thousands

of square miles."


     The concern seems justified (although it is probably easier to count

schools within a contour than, to estimate the total residential population).

It needs to be remembered that the treatment of special areas is offered as a

guideline for what is desirable, and that the project circumstances must

always be considered in using the Guidelines.

36.  Comment  (p. IV-11:  residential average number)

     The Guidelines, recognizing that the population varies over time, suggests

using a time-weighted average number of people for places such as churches,

parks, and stadiums, but specifies that this concept should not be used for

residential areas.  United Engineers suggests that this will overestimate

impact "for residential areas during the working day."


     On the face of it, this seems a reasonable comment.  However, when one

considers the data which lie behind the population weighting function, the

Guidelines are seen to be correct.  The social survey data consist of responses

from a wide variety of people with respect to time spent at home. Thus the

data on average response have already incorporated, for residential areas,

the time-averaging which is recommended in the Guidelines for other areas.

37.  Comment  (Tables IV-2 and IV-3)

     UMTA suggests including the category "industrial/commercial employees...

to maintain consistency" by dealing with people.  Con Ed notes, and EEI

agrees, "The level of accuracy for calculating Nil as reported in [these

tables—i.e., 3 significant figures] is beyond the accuracy of the input

data."  NBS notes that Tables 2d and 3d are transposed.  HUD asks why these ;

tables contain (in the definition of average population) a new day-night split

(at 0800 and 2000 hours).


     The inclusion of employees in the tables which summarize the impact is

a good suggestion.  Con Ed and EEI are correct in limiting Nil to two digits,

as Table VI1-3 in the Guidelines indicates.  NBS is correct about the trans-

posed tables.  HUD's confusion stems from their inferred comparison with the

^dn split-  The reason for the different time periods is explained more fully

in the revised text.

                                  Chapter V

38.  Comment  (p. V-2:  CSEL for impulse noise)

     Dames and Moore requests more explanation of CSEL.  NBS notes that CSEL

is to be used if peak pressures "are greater than about 105 dB," and asks "is

this to be A-weighted pressure?  C-weighted?  unweighted?"  DoD notes this

same cut-off, and says it is inconsistent with p. V-3, where 80 dB(C) is

indicated, "since the C-weighted sound exposure level for a blast  is approxi-

mated by the maximum C-weighted slow level."  Vibra-Tech says, "The C-weighted

network simply does not respond to much of the low frequency energy present in

air blast and which can produce a marked response in structures.   ...US Bureau

of Mines has recommended the use of a linear frequency system...," which most

commercial recording systems use.  NYPSC also points out that the Bureau of

Mines has recommended against C-weighting.  On the other hand, CERL provides

three papers which "further support the use of C-weighting to assess sonic

booms, air blasts, and other similar impulsive sounds."


     More explanation of C-weighted sound exposure level (Lgg) is  included in

the revision.  The 105 dB is meant to be unweighted, and is consistent with the

screening level identified on page II-l.  However, subsequent discussion has

led to the use of a 120 dB peak level.  DoD's comment is incorrect:  the

difference between the peak level (unweighted) of an impulse and the Lgc has

been empirically found to be about 20 and 30 dB for blast noise and sonic

booms, respectively.  However, the confusion probably derives from DoD's

reading of the 105 dB as "maximum C-weighted slow level," since the text was

not explicit.  The decision by Working Group 69 to use C-weighted L§ appears

to have been based in large part on work by Kampermann Associates (for EPA)

and by CERL.  This work is referenced in the revision, with better documenta

tion of the reasons for using
39 .  Comment  (p. V-3:  measuring CSEL)

     NBS says, "Implications of the 2 second contiguous time period require-

ment are unclear", and also questions the crest factor requirement of ANSI

SI .4, referring to IEC-79A.  AMRL reports on an attempt to use the approach

identified on page V-3, which was less than successful, suggesting problems

with the method.  R. W. Young questions the need for the thresholds included

in the 'Note*.


     The detailed discussion of measurement of impulse noise arises because

there is as yet no standard procedure for this.  In general, for impact

assessment, measurement is not the problem; prediction is.   For impulse noise,

however, the most effective prediction technique usually consists of measure-

ment of existing events similar to those in the proposed project.  Hence,

measurement becomes an issue for the Guidelines.  However,  rather than being

a part of the main text, this material is included in Appendix A in the


     The 2 second time period arises in an effort to specify a criterion

which could be used with unattended equipment to distinguish impulse noise

(such as blasting or sonic booms) from other high-energy noise events (such

as jet aircraft flyovers).

40.  Comment  (p. V-3:  CDNL)

     HUD indicates confusion as to whether CSEL or CDNL should be used for

impulse noise.  Young says, "My advice is to forget CDNL entirely."  Both

DoD and United Engineers (UE) question the "physical basis" (UE) for adding

CDNL to DNL, "since they are two distinctly different measures" (DoD).  UE

suggests that since this is an attempt to combine community reactions, the

combination should be done "during the impact-assessment stage, using non-

dimensional ratings."


     This portion of the text has been revised to clarify the reasoning

involved.  In large part, the conversion of C-weighted sound exposure level

to C-weighted day-night sound level is based on the desire to be able to

simplify impact assessment in situations with more than one impulse noise

occurrence per day.  Procedures for combining the community effects of impulse

noise and general audible noise are not spelled out in the revised version,

although it does recognize the need for these.

41.  Comment  (p. V-4:  infrasound)

     R & F Coal points out that "there is considerable confusion...on the method

of measuring mine blasts....  Because the frequency spectra of air blasts con-

tains [sic] infrasonic components it would appear that three different measure-

ments are required; 0.1 to 5 Hz, 5 Hz to 20 Hz, and C-weighted slow response."

They go on to conclude (paraphrasing a line from the preface), "With respect to

the use of explosives in surface mining the methods recommended are generally

not practical, certainly not economic, of doubtful accuracy...."


     The confusion they identify is present in the Guidelines.   It was the

intention of Working Group 69 that  blasting be covered by CSEL measurements,

and that the infrasound material was not meant to  apply in addition, but to

apply, alone, to other circumstances.   This is made clear in the revised text,

                                 Chapter VI

42.  Comment  (p. VI-2:  percent highly annoyed)

     Ford notes, "This choice was based on the assumption that high annoyance

provides a reliable summary measure of other adverse effects of noise....

This is an assumption which has not been carefully tested in the research

literature."  NBS says, "'percent highly annoyed* is really being used in a

very technical sense (e.g. the top two points of a 7 point annoyance scale)."

Consequently, they suggest "inventing a clearly technical term," or relating

this to "some estimate of just plain 'annoyed'."


     In the sense that there are few if any published papers testing this

assumption, Ford is correct.  The statement in the Guidelines is more an asser-

tion than an assumption, however, indicating what response will be used to indi-

cate general adverse reaction.  Further, the conceptual basis for this assertion

can be traced to Borsky's (1970) model of response to noise, which starts with

perception of the noise, moves to activity interruption or interference, which

results in annoyance, and may subsequently lead to complaints or other actions.

Thus annoyance is seen as the result of the direct impacts of noise, and conse-

quently a good summary indication of them.  While this conceptual framework has

never been proven, it has also not been disproved or discarded.  In fact, there

is some explicit support for this model in the report which Ford sent with

their comments (Hall, Birnie, and Taylor, Community Response to Road Traffic

Noise, p. 55; emphasis added).

          Four main findings emerge from these results.  Firstly, they
     show that annoyance due to road traffic is significantly related to
     the adverse impacts of the noise which an individual has experienced.

     To this extent the link between impacts and annoyance in the concep-
     tual model is supported.  Secondly,  however,  the results show that
     this link is by no means a simple cause and effect relationship.
     The proportion of the variation accounted for by the impacts whether
     treated singly or in combination indicates that intervening factors
     play an important role in determining annoyance.  Thirdly, the results
     indicate that reported health effects are in general more strongly
     related to annoyance than are the various types of activity inter-
     ference.  Fourthly, the results reveal no major differences in the
     strength or pattern of the relationships for reactions to road
     traffic noise in general as compared with reactions to truck noise.

     The NBS comment seems to arise because of inadequate explanation in the

Guidelines of the source of the generalized function (Schultz's synthesis).

First, 'highly annoyed1 is not a "technical" term.  It is plain language

which is meant to be interpreted as such.  In six of the 11 surveys in the

synthesis, the scales were fully labelled, such that "those people were

counted as 'highly annoyed' who said they were highly annoyed" (Schultz,

JASA, p.  380).  The remaining 5 surveys did not use fully labelled scales,

hence judgment was used to equate those scales to the other 6.  Second,

Schultz (1978) provides several good reasons for not using "just plain

annoyed," many of which are borne out by other publications (pp. 378-379).

43.   Comment  (Table VI-1:  criterion levels for various land uses)

     DoD says these "are practically unreasonable....  If published this way,

it will probably be used as a standard."  Dol notes that the table does not

provide a category for wilderness areas,  national parks, and other quiet places,

and feels they should be included.  Or. Kryter suggests that it should be made

explicit that the level for school buildings is a one-hour Leq, not 24-hour,

and also feels that the 60 dB level is too high.  NASA asserts, "It may be

questionable whether structural attenuation can be adequately described with

three discrete values," and raises several other questions about attenuation and

masking of outdoor noise as received indoors, since most of these levels are

rationalized on the grounds that "the noise-sensitive activity in question is

usually indoors" (Guidelines, p. VI-3).


     It is doubtful that this table belongs in the Guidelines.  Dr. von Gierke's

reply (6 Dec 77) to the DoD comment indicates, "In some respects this table

is a carry over from earlier versions....  In the final guidelines, this table

could easily be eliminated."  In fact, this table would appear to contradict

the rationale for the screening diagram (comment 18).  The table will be

deleted from the revised Guidelines.  The revised Guidelines will refer the

reader to appropriate material in other documents.

44.  Comment  (Table VI-2, 3, and 4:  summaries of human effects)

     The fact that the basic information in these tables is drawn from the

Levels document bothers Dol, on the grounds that the Levels analysis included

a margin of safety.  Dr. Kryter raises two objections (in correspondence with

Dr. von Gierke):  "(1) the omission, without better justification, of the

benchmark function showing that 17% of the people are highly annoyed from

aircraft noise at an Ldn of 55; and (2) ...the failure to use as a criterion

measure, at a minimum, an hourly Leq related to specified noise-speech areas

in a classroom or lecture hall...."  MVMA, on the other hand, wishes it to be

made very clear in the guidelines that "(1) they reflect essentially what was

concluded previously in the Criteria and Levels Documents of 1973 and 1974,

(2) little subsequent work has been done to substantiate those documents (see

'Guidelines' reference list), and (3) new information may demonstrate that

those Documents and the Guidelines may not be correct."


     These tables do contain new data,  and Dr.  Kryter's suggestion is certainly

reasonable.  A fuller explanation is provided in the revision.  Further, the

material needs to be available in more  detail if it is to be easily usable.

Kryter's concern about schools is a case in point.   If we expect relationships

such as these to be used to describe impacts on schools, the relationship

needs to be available in more detail than 10 dB intervals.   When this detail

is provided, it is clear that MVMA's comments are largely incorrect.  Sub-

stantial new work on the effects of noise on residential communities has

been done in the past five years, and is incorporated in these Guidelines.

Dol's concern about the use of the Levels document  is a legitimate one.  It

is made clear in the revised version that what  is being used is the data

summary of Appendix D, not the policy material  of identified levels.

     Kryter's comment about hourly Leq  is further reason to expand the

discussion of the short-term metrics (comment 29),  and to expand the material

in paragraph 4 of page VI-2, describing how to  treat non-residential urban

land uses.

45. Comment  (p. VI-8:  response independent of source)

     The Guidelines state that Schultz's data confirm the assumption that

the relationship between annoyance and  noise is independent of noise source.

Ford notes, "This is a strong claim considering that the Schultz synthesis

involves only transportation noise surveys," a  concern which is echoed by

United Engineers.  Ford's position is summarized thus:  "What has to be

questioned is whether our present state of knowledge about the relationships

between exposure to noise and human response supports the reliance upon a

single weighting function and specifically the  one  described in the Guidelines."


Ford submitted a copy of a report from McMaster University, which indicates

lower percentages highly annoyed by road traffic than Schultz's curve would

suggest.  From this, Ford draws the implication that, "because most of the

studies defining the Schultz curve are aircraft related, his average curve

does not accurately represent response to road traffic noise."  In addition,

they cite several studies which show differences in response to road traffic

and aircraft noise, and point out that differences in the wording of the

questions reviewed by Schultz "serve to complicate further a synthesis of the

results."  MVMA asserts that the logic in the Criteria document, "heavily

biased to aircraft exposure," is "carried into the 'Guidelines'."  On the

other hand, Dr. Kryter says that "serious, substantial errors appear to have

been made . .  . with respect to the effects of aircraft noise."  Both Kryter

and Ford suggest that more than one function or set of criteria may be

required.  Ford also suggests that the Guidelines should be "updated on a

semi-annual basis so that it is a truly living document."


     As noted earlier (comment 28), one cannot wait for complete information

before attempting to assess the environmental impacts of noise.  At this

point in time there are no published studies which quantify any differences

in response to different noise sources, nor are there any studies relating

average response to noise level for non-transportation sources.  (Dol states

that a study is in progress on electric power transmission line noise, but

its results are not yet available.)  In the absence of any solid information

to the contrary, the Guidelines have adopted a single relationship for all

noise sources.  If information becomes available which can be used to identify

a different relationship for certain sources, that should certainly be used.


In that respect the Guidelines are open to revision as new information becomes

available, although that may not occur on a semi-annual basis as Ford suggests.

     On the specific points raised, responses are as follows.  In a more

recent paper, presented at the International Congress on the Biological

Effects of Noise, in Freiburg, Germany, in October 1978, Schultz constructed

separate curves for the 4 road traffic studies,  the 6 aircraft surveys, and

the rail study in his synthesis.  His conclusion is, "The road and air

traffic results are extremely close; while the railway data are perceptibly

lower than the other two, one cannot claim a significant difference."  The

importance of the differences in wording of questions, noted by Ford, is

possibly responsible for differences between the McMaster results and Schultz's

synthesis.  McMaster asked simply for a rating of the noise; many of the others

asked how annoyed the respondent was.  Social survey experience suggests that,

relatively, a higher degree of response will be elicited by the latter ques-

tion.  Also, Schultz has suggested that the difference arises because the

McMaster study uses a bipolar scale whereas most others have used simply an

annoyance scale (JASA, 1978, p. 403).

     MVMA's assertion that the Guidelines are simply a repeat of data in

the Criteria document is obviously incorrect: eleven of the thirteen studies

in Schultz's synthesis were not included in the Criteria document.  Kryter's

comment arises from the fact that the Guidelines do not explain Schultz's

synthesis well. The revised Guidelines are explicit about the reasons for

rejecting the earlier curve.

46.  Comment  (p. VI-11:  Degradation of environment)

     DoI would like some guidelines for assessing as a special case the

noise impacts where no people are living, such as wild lands.  NBS remarks,


"1C doesn't seem appropriate to use [data based on daily residential living]

to quantify impact on areas 'to which urban dwellers wish to go for an escape

from city noise....1"  United Engineers makes a similar comment:  "How can

you compute a percent population highly annoyed in an area where no people are

living?"  NYPSC says this approach is inadequate, since "almost by definition,

'degradation1 must consider the extent to which the ambient sound level is



     These comments are all cogent.  (See comments 3 and 18 also.)  For special

areas valued for their quiet, a tabulation of the impacts and a qualitative

description are recommended.  The tables would be similar to those used for

populated areas, but rather than focusing on number of people subjected to

various noise levels, the tables would focus on the size of the affected area.

Sensitive uses, such as for particular animal species, or for wilderness

recreation, could be treated the same way special situations are in the

populated areas.  This proposed approach has the advantage of being able to

identify impacts even when there is no permanent population.  In some cir-

cumstances, it may be useful to reduce this tabulation to a single number,

in which case the population weighting function is the best available indi-

cator of relative impact, and can be used to calculate a "level-weighted area."

If data become available which describe response as a function of noise in such

areas, the Guidelines will be revised accordingly.

47.  Comment  (p. VI-12:  Effects of impulse noise)

     R & F Coal questions the use of sonic boom data to estimate annoyance

for surface mine blasting.  "Because people complain about military aircraft

flying over their homes at supersonic speed it does not follow that they will

complain about a taxpaying, local employer, 'who is producing vital basic

materials just because the decibel level is the same."


     Data specific to mining blasts are not available.  In the revised

version, recognition is given to the fact that estimates of C-weighted sound

exposure level (Lgc) from peak pressure are different for boom and blast

noises.  As a result, booms are seen to cause adverse response at lower peak

sound pressure levels than blasts, for the same Lgc, as suggested by R & F


48.  Comment  (p. VI-15:  infrasound)

     EEI recommends inclusion of "cut-off criteria for vibration, structure-

borne noise, ultrasound and infrasound."  United Engineers notes that "the

fact that infrasound is sometimes related to structural vibration was not

even ment ioned."


     If EEI's comment is intended to point out the absence of a screening

diagram for these considerations, it is well taken.  The numbers in the right-

hand column of Table 1-3, for impulse noise and vibration, are such cut-off

levels, which is stated explicitly in the revised version.  United Engineers'

comment is correct, and has been noted in the revision.

49.  Comment  (Fig. VI-6 and Table VI-5:  Vibration criteria)

     R & F Coal points out that "the annoyance curve suggests that people

will be more annoyed by numerous small blasts than by one blast properly

designed to do the task, and this preference is compatible with our exper-

ience."  They then say that the Guidelines attempt to reduce both the charge

weights and the number of impulses per day, which can only be done by mining

less material, and that is not in the national interest.  Ky Mines asserts

that acceleration measurements "do not correlate effectively with damage

levels," and that the level specified for vibration is "unrealistically



     The comment by R & F Coal provides additional empirical confirmation of

the figure.  The inference drawn from it, however, is a good example of a

misreading and abuse of the Guidelines.  The Guidelines exist to assist one

in quantifying impacts. The question of trading off those impacts against

other concerns is not addressed in the Guidelines.  For coal extraction,

energy concerns may outweigh vibration impacts.  The important point is that

the impacts are identified, and are identified properly.  The correlation of

acceleration with damage is irrelevant:  The Guidelines bases the vibration

criteria on human response, as both the text and Figure VI-6 make clear.

For that reason, the levels undoubtedly are lower than those based only on

damage criteria.

50.  Comment;  (p VI-21:  blast noise)

     R & F Coal notes that the Guidelines describe the formulas given for

blast size against safe distance as "newly proposed and somewhat tentative,"

and ask, therefore, why they should be  used.   R & F Coal also notes that

these are calculated for worst case conditions,  and incorporate numerous

safety factors, such that no allowance  is made for the careful surface mine

operator.  They also claim the resulting charge weights are unduly restric-

tive, and are about one-fifth of what their experience indicates is safe and

reasonable.  They question whether charge weight per delay or total explosives

in the blast is meant in the formula.  DoD also says the formula is overly

conservative, and suggests 250^"w~as the distance beyond which there is

almost certain to be no glass breakage.


     These somewhat tentative guidelines for blasting are offered simply because

there are no other generally accepted,  simplified guidelines.  However, if

the blasting engineer is skilled and careful enough to be able to make more

detailed calculations and to (e.g.) wait for wind changes, then other formulas

can certainly be used.  That is in no way contrary to the intent of the Guide-

lines, which attempt to provide simplified procedures for identifying

potential problems.  (If this equation shows no potential impacts, for example,

there is no need to go into the more detailed analyses by a blasting engineer.)

The factor of five noted by R & F Coal  is probably due in large part to the

safety factors built into the Guidelines equations.  The equations should

specify charge weight per delay, an editorial omission which will be corrected.

The DoD formula appears quite consistent with the equation in the Guidelines,

once one corrects for different units (feet and pounds for DoD, kilometers

and kilograms for the Guidelines).


                                Chapter VII  .

51.  Comment  (p. VII-1:  Single number impact  index)

     HUD disagrees  that  it  is desirable to derive  a  single  number which

represents the integrated impact.  "[T]he approach is  not necessary  for  an

adequate assessment of impact."  FHWA made similar comments.  NYSDOT suggests

that tabulating "the number of persons/residences  at various  levels" would

suffice.  DoD also  concurs  (emphasis in original):   "[N]o criteria are given

to evaluate the results  of  the calculations.  How  will  they be used  in the

decision-making process?" "Until it is tested...,  we cannot concur in its

     From the perspective of standards or limits  (which HUD  and  FHWA both

employ), it is probably true that  the notion of a single number  impact  index

offers little extra information.   For such uses,  the number  of people within

each of the contour bands may be a reasonable end-point of the analysis.

However, for full and thorough environmental assessment, which these guide-

lines are written for, noise impacts must be compared with,  and  traded-off

against a host of other decision criteria.  In those circumstances, a single

number index of noise impacts is obviously preferred over a  full table  of


     In reponse to DoD's comment,  the revised version is more explicit  about

how to use the calculated results.  In particular,  the revision  suggests

which of the single-number  indexes should be emphasized in any given situa-

tion, and suggests bases for comparison of the numbers.  For clarity, the

different situations are identified on a reproduction of the screening

diagram (comment 18).


52.  Comment  (p. VII-2:   Level weighted population)

     Dames and Moore note that references or discussions are needed to

justify the assumptions stated, because they are open to question and criticism.

NASA says, "the assumptions underlying Sound Level Weighted Population are

critical to the guidelines and should be demonstrated experimentally."  FAA

says the paragraph in question "is completely unreasonable."  Noise criteria

should protect people regardless of population density.   HUD says the second

assumption "is absurd".  EEI, in earlier correspondence  on the Fractional

Impact Method (15 March 76), says the assumption is wrong, and while there

"is probably no reliable  relationship..., non-linear values...make more sense

than linear values."  More recent correspondence (29 Nov. 77)  adds to this,

saying the assumption has no verification in the technical literature, and

that the assumption "implies a linear subjective response to noise and would

require a linear sound weighting function." Con Ed echoes EEl's concern for

"verification" of this "incorrect" assumption in the technical literature.

Ky Mines says this is "totally inconsistent with human rights, in that

according to the guidelines a group is entitled to more  peace and quiet than

an individual."  Columbia Gas says this assumption contradicts "the statement

on page B-2 that the rate of change of annoyances with sound level is greater

at high sound levels than at lower sound levels." FHWA points out that LWP

"is a number whose absolute value is meaningless. One of the serious defi-

ciencies of this report is how to explain to the layperson, in terms he will

understand, what the noise impact will be."


     In fact, FHWA has identified the fundamental problem with LWP, which is

behind all of these comments.  In the Guidelines, it is  explained in terms


which make it poorly understood.  This probably arises because it is explained

as a carryover from and minor modification to the earlier Fractional Impact

Method, which did in fact rest on some critical assumptions.  The following

alternate explanation should overcome all of the comments, as will be shown.

     The single number impact index under discussion here applies only to

general audible noise in populated areas, and is calculated in terms of the

previously identified 'general adverse impact1.  In particular, it is based

on the use of a function (Schultz, JASA 1978) expressing percentage of people

highly annoyed as a function of outdoor residential Ldn.  Clearly, if 4% of

people are highly annoyed at 55 dB, and 37% at 75 dB, one can identify the

total number of people exposed to each noise level, multiply by the appro-

priate percentage, and add over all noise levels to identify the total number

of people highly annoyed.  This number is one which is easily explained

to laypeople because it has intrinsic and intuitive meaning.

     If the single number is explained in this fashion, there is much less

chance of the kind of mis-interpretation both FAA and Ky Mines make.

Schultz's work provides the empirical demonstration requested by several.  It

is obvious that linear or non-linear functions are equally possible, and that

this one is identically that discussed in Appendix B.

53.  Comment  (p. VII-4:  PLH and other abbreviations)

     NBS notes, "PLH and PHL are used interchangeably.  Need to pick one and

stick with it."  United Engineers concur.  HUD says, "The non-technical

reader is introduced to such abbreviations as FI, PHL, LWP, Nil, DNL, YDNL,

CSEL, CDNL, TAVL, and SEL, not to mention all their derivations; these may

add to the mystique of the product, but leave much to be desired in a practical

guidance document."



     Certainly when abbreviations are used, they should be used consistently.

The revision will be carefully edited.  But HUD's point is well taken.  Even

careful editing may not improve communication that much.  Perhaps abbreviations

are needed for use in equations,  but an excellent argument can be made for

writing the text in full,  and for minimizing new jargon in Guidelines such as

these.  The revised version uses the symbols set out in the EPA acoustic

terminology guide.

54.  Comment  (p. VII-4:  Population weighted loss of hearing)

     NBS asks why the denominator of PLH includes only those exposed to L
55.  Comment (p. VII-5:  Normalizing weighting function)

     Both Ford and MVMA question "the apparent 'scaling up1 of the Schultz

'percent highly annoyed1 curve in the Guidelines by a factor of approxi-

mately 3" (Ford).


     The revision uses Schultz*s curve directly.  (See comment 52).

56.  Comment  (p. VII-5 and Fig. VII-7:  Sound level weighting function)

     Wyle points out that, with the proposed function, impacts could be

theoretically unbounded, since this function falls less rapidly than do noise

levels.  NBS suggests that the mathematical expression of this function can

be simplified.  Both Wyle and EEI suggest that the function should be zero

below roughly 55 L
annoyed), and (c) in all eleven synthesized surveys, there is only 1 data

point below 45 dB (for which % HA is 0 (Fig. 36).

     The effect of background noise remains a problem, although not as large

a one as NYPSC suggests.  More data (Schultz's) are now available than were

when the Levels Document was compiled.  The arbitrariness of the curvilinear

function is refuted by the fact that it is derived from data, and that others

besides Schultz have found non-linear relationships for many responses of

interest (e.g. Hall, Taylor & Birnie, in the report submitted by Ford).

However, as the Guidelines point out, this curve differs little from the

straight line used earlier by EPA to indicate relative impact.  The re-

striction to outdoor environments noted by Wyle will be made explicit in

the revision.

57.  Comment  (p. VII-8:  Weighting function for loss of hearing)

     NBS says '"loss per ear* implies that one should multiply by the number

of ears... Suggest using 'average dB loss per person'".



58.  Comment  (p. VII-9:  Change in LWP)

     NBS:  "Percent change should be change in sound level weighted popula-

tion, divided by LWP before the change."  United Engineers notes that as

defined this is a ratio, not a percentage.



59.   Comment  (p. VII-12: temporary noise environments)

     Dames and Moore asserts, "YDNL should not be used for temporary noise

environments," because the yearly average may not change much for medium or

small projects.  NYPSC says, "Actual impacts can be substantially greater than

would be predicted using an annual L^n," and implies that a better approach

to assess the impacts is to estimate them as if the project were permanent.


     The Guidelines specifically suggests using both the actual noise during

the project and the yearly average, in order that the decision-maker have

available the most appropriate information.  It is instructive to note (using

the example on page VII-13) that impact calculated from the yearly day-night

sound level (LWP=601) will almost always be greater than impact for the year

calculated from a weighted average of project and non-project noise levels

(LWP=557, from 3/4 yr. of 664 and 1/4 yr. of 236).

60.   Comment  (p. VII-15:  vibration impact index)

     NBS:  "Line 4 should read 'total number of people* (not residences).

Again, however, only the number of people in areas of significant vibration

are included in the denominator."  (See comment 57.)  VWP is Vibration-

weighted population.



                                Chapter  VIII

61.   Comment  (p.  VIII-1:   noise re other impacts)

     DMTA says the summary "could have been enhanced by discussing a mechanism

for evaluating noise impacts relative to other environmental impacts."


     In line with  the revised context and title of  the Guidelines, a discus-

sion of comparative evaluation of impacts would be  desirable here.  However, it

is not the task of these Guidelines to suggest mechanisms for making such


62.   Comment   (p. VIII-1:  structures exposed)

     EEI request specifying a lower limit below which there is no need to

discuss noise induced structural damage.


     It is made clear in the revised version that this consideration applies

only to impulse noise, infrasound, and ultrasound,  not to general audible

noise.  (For levels to use, see comment  48.)

63.   Comment  (p.  VIII-2:   sound level weighted area)

     United Engineers notes that this concept was never defined.


     We agree that the summary is not the place to  introduce new concepts.

This one is discussed earlier in the revised version.  (See comment 18).  As

defined, it may prove useful in two situations:  to reduce tabulations of

unpopulated areas  affected by noise; and for cases  where it is known that an

area will be developed, but exact population projections cannot be forecast.


64.  Comment  (p. VIII-3:  descriptive qualitative evaluations)

     HUD requests "more guidance" on doing descriptive evaluations, "e.g.

examples of questionnaires, where to obtain complaint listings, etc."


     While more guidance might be useful, it is not clear what can be

offered in a short section which would be generally applicable.

65.  Comment  (p. VIII-5:  new population exposed)

     NBS indicates this paragraph is not clear.  "Is not the  'changes in  the

existing environment...1 to be obtained from Table IV-1?  Then what is  the

meaning of comparing to the predictions of Table IV-1?"  Further, NBS takes

issue with the assumption that the new population comes from an area of

Nil .35, suggesting some discussion of likely previous residential area be



     The revised version clarifies this paragraph in line with the following

points.  For projects which entail new populations in an area where noise is

otherwise unchanged, two possibilities arise.  The first is that the existing

noise levels are greater than the levels which can be expected as a result

of the population cluster (Table IV-1); the second is that population noise

will be greater than the existing.  In the first case, noise impacts are

obvious, and the site plan will probably include ways to reduce that impact.

In the second case, it needs to be pointed out that there are noise impacts,

and that, again, the site plan should look to ways to reduce them.  Table

IV-1 can be used to show the expected sound level for the population density

in question, and that can be compared with the levels  predicted on the basis

of the thoughtful site plan.

     On the basis of a somewhat cursory knowledge of the residential migra-

tion literature, we find it unlikely that one could say much useful about

the likely previous residences of a new housing development.  The value

of Nil = 0.35 is based on the calculation in Table VII-3, and is intended to

represent the national average value.  It is explained more clearly in the

revised Guidelines.

                                 Appendix A

66.   Comment  (Terminology)

     NBS has offered suggested improvements to 13 of the terms in part 1.

United Engineers has also made a number of comments on the definitions, as has

DoD.  Isolated points are raised by FAA and NASA.  NBS asks what happened to

section 2 through 4 of this Appendix.


     Each suggested revision will be checked against EPA terminology and other

normal usages.

                                 Appendix B

67.   Comment  (exponential functions)

     NBS asks:  "Why is equation B-6 better than B-5?"


     The Working Group apparently preferred the two exponential functions for

two reasons.  First, at higher levels, this function "has the same rate of

growth as a loudness function" (page B-5).  Second, the asymptotic behavior of

the function at very low noise levels provided a way to quantify environmental

degradation in such areas.  The revised version retains equation B-6 since it

was indistinguishable from equation B-5 in the range of the (Schultz) data.

However, its use is discouraged (as noted in comment 56) below levels of 45

dB (Ldn).

68.  Comment  (p. B-8:  Population weighted hearing loss)

     United Engineers makes several comments regarding calculation of pop-

ulation-weighted hearing loss.  "[T]he time dependence should be kept in

parametric form, rather than averaged out.   This is especially important if

impacts are to be assessed in 5-year increments."  The assessment should be

"of NIPTS due to the new source alone.. .Also,  the assessment was to be for a

20 year time span, not 40 years."


     Data are not available to assess NIPTS in 5-year increments, so there

is no point to leaving in the time dependence.  Assessment of NIPTS for the

new source only would not be adequate, as the resulting level may be higher

than either the source or existing levels.   Comparison of the resulting

level's NIPTS with that of the existing  level might be reasonable.  It is

true that the assessment covers at most  a 20 year span, but the average used

here in fact approximates median response after 20 years.

                                 Appendix C

69.  Comment (ISO/TC 108/SC 2/WG 3)

     Vibra-Tech Engineers main concern "is  the inclusion of arbitrary levels

of vibration alleged to cause damage."  The ISO draft standard, "the basis for

this criteria," "...was so out of step with more widely accepted practical

research done in the U.S.A. and other countries it was never adopted as an I.S.O,

standard by any member country."


     The Guidelines do not adopt the ISO recommendation, either.  Section

4.5 of Appendix C specifically states, "reducing the threshold from 50 mm/sec

to 5 mm/sec does not appear warranted."  The data used to support a reduction

of 50% are given in Fig. C-4, from U.S. Bureau of Mines.  (See next comment


70.  Comment  (Fig. C-4)

     R & F Coal says, "This is a graph from U.S. Bureau of Mines Bulletin

656 which contains 4 points below the 2 ips safe level.  The study group

is in error.  The four minor damage points are not the result of blast damage.

They are from some experiments where eccentric motors were set up in doorways

and inside houses in order to induce damage.  They should not be included at

all."  Hence the 50% reduction of the commonly used 2 ips safe level is

uncalled for.


     R & F Coal is correct that the four points were from eccentric motor exper-

iments.  R & F Coal do not mention that all 160 of the data points were from

that experiment.  To throw out those data is to remove most of the informa-

tion which established the 2 ips level.  Further, a close inspection of

Figures 3.1-3.4 of Bureau of Mines Bulletin 656 shows that careless drafting

left out of Fig. 3.4 a minor damage point at 40 cps and .0015 inches which

was in Fig. 3.1 and would not be covered by the proposed criterion.


                           *4	3_£
                 ••   f      •



           If *( ..€«te«V»( ^ rff^'ffi

   N JL	A/jtee. Jjrywah_/«^«		_%. 17

                             r.  k*«        «g*3