AMRL-TR-73-91

EPA-550/9-73-001-B
              PREDICTION OF NIPTS DUE TO
             CONTINUOUS NOISE EXPOSURE


                  DANIEL L. JOHKSON, MAJOR, USAF


             AEROSPACE MEDICAL RESEARCH LABORATORY
                           JULY 1973
                    JOINT EPA/USAF STUDY
               Approved for public release; distribution unlimited.
            AEROSPACE MEDICAL RESEARCH LABORATORY
                  AEROSPACE MEDICAL DIVISION
                  AIR FORCE SYSTEMS COMMAND
              WRIGHT-PATTERSON AIR FORCE BASE, OHIO

-------
                                       NOTICES
 When US  Government drawings, specifications, or other data are used for any purpose other than
 a definitely related Government procurement operation, the Government thereby incurs no respon-
 sibility nor any obligation whatsoever, and the fact that the Government  may have formulated,
 furnished,  or in any way supplied the said drawings, specifications, or other data, is  not  to be
 regarded by implication or otherwise, as in any manner licensing the holder  or any other person or
 corporation, or conveying any rights or  permission to manufacture, use, or sell any patented in-
 vention that may in any way be related  thereto.

 Organizations  and individuals receiving announcements or reports via  the Aerospace Medical Re-
 search  Laboratory automatic mailing lists should submit the  addressograph plate stamp on the
 report envelope or refer to the code number when corresponding about change of address or can-
 cellation.

 Do not return  this copy. Retain or destroy.

 Please do not request copies of this report from Aerospace Medical Research  Laboratory. Additional
 copies may be purchased from:

                               National Technical Information Service
                               5285 Port Royal Road
                               Springfield, Virginia 22151
This  report has  been reviewed and cleared for open publication and/or public re-
lease by the  appropriate Office of Information  (01)  in accordance with APR 190-17
and DODD 5230.0.   There is  no objection to unlimited distribution of this  report
to  the public at large, or  by DDC to  the National  Technical Information Service
(NTIS).

This  technical report has been reviewed and is  approved for publication.


FOR THE COMMANDER
   -r

Osfi/
HENNINOsfi/ von GIERKE, Dr. Ing.                 ALVIN F.  MEYER, JR.'
Director,  Biodynamics and Bionics Division    Deputy Assistant Administrator
Aerospace  Medical  Research Laboratory            for Noise Control Program
                                                  United States Environmental
                                                    Protection Agency
AIR FORCE/56780/14 August 1973 — 2500

-------
                              DOCUMENT CONTROL DATA - R & D
      (Security cl*»iitlcmtion of f/fft>, body of abmtrmct and Indexing annotation must b» antarad when tfta overall report 13 ctufiHted)
I. ORIGINATING ACTIVITY (Cotpotflt author)
 Aerospace Medical Research Laboratory, Aerospace
 Medical Division, Air Force Systems Command,
 Wright-Patterson Air Force Base, Ohio 45433
                                                    2«. REPORT SECURITY CLASSIFICATION
                                                         UNCLASSIFIED
                                                    26. GROUP
                                                             N/A
3. REPORT TITLE
 PREDICTION OF NIPTS DUE TO CONTINUOUS NOISE EXPOSURE
4.. DESCRIPTIVE NOTES (Typ* ol report and Inclusive detee)
8. AUTHOR(S) (Flral nmme, middle Inltlml. laet name)

    Daniel  L.  Johnson, Major, USAF
«. REPORT DATE
M. CONTRACT OR
b. PROJECT NO.
c. Task No.
* Work Unit
July 1973
GRANT NO.
7231
723103
16
^m. TOTAL NO. OF PAGES 7b. NO. OF REFS
66 16
»a. ORIGINATOR'S REPORT NUMBER(S)
AMRL-TR-73-91
»t>. OTHER REPORT NO (Any other munbere that may be aaatgrtad
thl* report) •
EPA-550/9-73-001-B
 10. DISTRIBUTION STATEMENT
                      Approved for  public release; distribution unlimited
 II. SUPPLEMENTARY NOTES
                                              12. SPONSORING MILITARY ACTIVITY
                                              Aerospace Medical Research Laboratory,
                                              Aerospace Medical Div,  Air Force Systems
                                              Command, Wright-Patterson AFB, OH 45433
II. ABSTRACT
  In support of the main document,  "A Basis for Limiting Noise Exposure for Hearing
  Conservation," this report  compares the relationship of noise exposure to Noise
  Induced Permanent Threshold Shift (NIPTS) as predicted by  the currently available
  works of Passchier-Vermeer,  Robinson, Baughn and Kryter, and the yet unpublished
  work of the National Institute of Occupational Safety and  Health.  The works of
  Passchier-Vermeer, Robinson,  and  Baughn are selected since these are the only works
  that completely predict  the relationship between NIPTS and noise exposure for
  various audiometric frequencies,  sound pressure levels and population percentiles.
  The predictions of these three methodologies are averaged  in order to provide one
  single relationship between continuous noise exposure and  NIPTS.  This relationship
  is presented in various  ways so that the effect of noise exposure on hearing can be
  viewed in more than one  way.   Discussion concerning the type of frequency weighting,
  the equal energy rule, and  long duration exposures is also provided.
DD
 FORM
i NOV ee
1473
                                                              Security Classification

-------
                               PREFACE


      The Biodynamics and Bionics Division of the Aerospace Medical Re-
search Laboratory was given the responsibility under an Interagency Agree-
ment with the Environmental Protection Agency, to develop a document which
would serve as a basis for limiting noise for purposes of hearing conserva-
tion.  The preparation of this document was accomplished by the University
of Dayton Research Institute  (UDRI) under Contract F33615-72-C-1402.
The Aerospace Medical Research Laboratory efforts in support  of this pro-
ject were included under Project 7231-03-16, "Auditory Responses to Acous-
tical Energy  Experienced in Air Force Activities. "

      In order to resolve certain issues that developed during preparation of
the primary document,  the material of this supporting document was develop-
ed.  This document does not cover all facets of the  relations  between hear-
ing and noise exposure, and should be used only in conjunction with the
primary  document "A Basis for Limiting Noise Exposure for Hearing Con-
servation" (AMRL-TR-73-90) (EPA-550/9-73-001-A).

      Acknowledgement is made of the assistance provided by Dr.  H. E. von
Gierke, Dr. C. W. Nixon and Capt.  David Krantz of the Biodynamics and
Bionics Division.
                                   ill

-------
                          TABLE OF CONTENTS
                                                                  Page
  I    INTRODUCTION                                               1
 II    RELATION OF NOISE TO HEARING LOSS                      1
      A.  Relation of Noise to Hearing Loss for Constant SPL
          for 8 Hour Working Day                                    1
          1.   Exposure Situation of Data Base                         1
          2.   Selection of Data Base                                  2
          3.   Other  Methods                                        10
          4.   Simplication of Data                                   10
          5.   Details of Selected Methodologies                      10
          6.   Manipulation of Data                                  19
          7.   Considerations                                        32
          8.   Risk of Noise Relative to Hearing Level Exceeding
              a Predetermined Level or Fence                       39
          9.   Percent of the Population with more than 5 dB
              NIPTS at 400.0 Hz Versus 8 Hour Noise Exposure
              Level                                                 44
         10.   Selection of Limit for 8 Hour Day                      44
         11.   Criticism of Kryter's Method                          44
         12,   D-Versus A -Weighting of Frequency                   50
         13.   Duration  of the Exposure                              50
         14.   Estimation of the Accuracy in Relating NIPTS to Noise
              Noise  Exposure                                       55
      B.  Requirement for "Quiet"                                  58
in    SUMMARY                                                   58
IV    CONCLUSIONS                                               60

      REFERENCES                                               61
                                   iv

-------
      PREDICTION OF NIPTS DUE TO CONTINUOUS NOISE EXPOSURE
I.     INTRODUCTION
      This report was written to support certain parts of the criteria document,
"A Basis for Limiting Noise Exposure for Hearing Conservation".  Specifically,
several different predictive methods are presented that estimate the effects
of noise on hearing.  The predictive results will then be manipulated until
they are reduced to a format that allows  a basis for administratively proposing
a specific noise limit.
      This report relies on the main document (AMRL-TR-73-90) for defini-
tion of terms, arguments concerning impulsive noise, relationships between
Temporary Threshold Shifts (TTS) and Noise Induced Permanent Threshold
Shift (NIPTS), etc.
      Method of Attack.  With respect to  NIPTS, the duration, spectrum and
intensity of the noise exposure,  the sensitivity of the individual,  and the life-
time noise exposure history of the individual are all important parameters.
With this many parameters, it is predictable that there are varied opinions
as to how NIPTS will develop in a group of people exposed to noise.  If one
adds to the problem various interpretations  of what constitutes a significant
hearing loss, then it is not surprising that a resulting jumble of noise limit-
ing criteria will develop.  The intent of this supplement is not to be inter-
pret what constitutes a significant hearing loss until such interpretations are
required in order to suggest a recommended limit.  Therefore, major em-
phasis will be placed on the relationship  of NIPTS to noise for various popu-
lation percentiles.


H.    RELATION OF NOISE TO HEARING LOSS
      A.   Relation of Noise to Hearing Loss for Constant SPL for 8 Hour
          Working Day
          1.  Exposure Situation of Data Base. This situation is the basis
of much of the human data with respect to actual hearing loss.  Therefore it
is this situation that by necessity anchors any criterion which will relate
hearing loss to noise.   Once this point is selected, exposure duration is then
handled such that shorter or longer exposures are expected to be as noxious
as the 8 hour exposure.  The 8 hour permissible exposure point,' therefore,
must be set with great care.  Since this 10 the heart of the report,  a consid-
erable amount of detail will be presented that will hopefully allow selection
of permissible noise exposure for an 8 hour day.

-------
          2.  Selection of Data Base.  Various researchers have made an
attempt to develop a predictive relationship between noise exposure in the
8 hour working day and the resulting hearing losses.  The relationships were
investigated and either accepted or rejected based on whether or not they
(a) allowed calculation of NIPTS at various percentile points and (b) consider-
ed at least speech frequencies (. 5, 1 and 2 kHz) and the audiometric frequency
of 4  kHz.  The methods of Passchier-Vermeer,  Robinson and Baughn satisfy
these restrictions.
             Passchier-Vermeer1 s method is attractive in that it correlates
the data of many different reports.  Inclusion of her method thus provides a
rather broad data base (see Table 1 for a  summary of her sources).  A weak-
ness of her method is that for much of her data base only the 25,  median,
and 75 percentile levels  of the population  were provided.

             Robinson's method provides one mathematical relationship (the
hyperbolic tangent) which is adjusted for the audiometric frequencies con-
sidered and the percentile levels used.  The method's strength is  that it allows
calculation of predicted NIPTS for a wide  variety of conditions.  A criticism of
the method might be that it uses  only one careful study of an otologically
screened population of British subjects.  Such a population may not be typical
of average US population.  It is also difficult to visualize how the hyperbolic
tangent could be a best approximation to NIPTS for all frequencies and condi-
tions.  Nevertheless, Robinson's methodology is well conceived and provides
an additional .data base.
             Baughn's data provides superior insight into how NIPTS develops
at various percentile points,  not just the median*  It has also been used as
the basis for the ISO standard.  Its weakness,  as typical with many industrial
studies, is that some residual TTS will have been measured  since an occasion
only 20 minutes recovery was allowed before audiometric testing was performed.
Lack of recovery would tend to make the predicted NIPTS too high.  A second
problem is that the control (or non-noise exposed group) must be considered
to have been exposed to 78 dBA or less.  Therefore from Baughn's data
alone,  it would be impossible to show that the  78 dBA exposure was not in
itself causing a significant NIPTS.

             In summary, all three methods have both strengths and weak-
nesses and it would be hard to say which of the three methods (Robinson's,
Passchier-Vermeer1 s or Baughn) gives the best estimates of the true situ-
ation.  Therefore,  the predicted NIPTS values were tabulated for  each method
and compared.  The results,  as seen in Table 2, speak for themselves.   In
general, there are not large (greater than 10 dB) differences between the
three methods.  Most differences are less than 5 dB. For this reason, all
three methods  were used to derive predicted values of NIPTS.  The final
prediction is the average of the NIPTS of  each method; and, as a consequence,
should give a final  result that is not unduly influenced by the weakness of any
single method.

-------
                           TABLE 1
    Work Included In Passchier-Vermeer's (1968) Analysis
W.  Burns, R. Hinchcliffe, T.S.  Littler,
An exploratory study of hearing loss and noise exposure in textile
workers.
The Ann.  of Occ. Hyg.  1_ (1964) 323-333.

R. Gallo, A. Glorig,
P.T.S. changes produced by noise exposure and aging
Am. Ind.  Hyg. Ass.  Journal 2,5  (1964) 237-245.

The relations of hearing loss to noise  exposure
A Report  by subcommittee Z 24-X-2  (1954)  34.

N. E.  Rosenwinkel, U. C. Stewart,
The relationship of Hearing Loss to Steady-State Noise Exposure
Am. Ind.  Hyg. Ass.  Quart. J£,  (1957)  227-230.

J. Nixon,  A.  Glorig,
Noise Induced P. T. S. at 2000 and 4000 Hz.
J.A. S.A.  33(1961)  904-913.

W.  Taylor, J.  Pearson, A. Mair,  W. Burns,
Study on noise and hearing in Jute weaving
J.A. S.A.  37 (1964) 113-120.

B. Kylin,
T. T. S. and auditory  trauma following exposure to steady-state noise
Acta Oto-Laryng.  Suppl.  152  (I960).

F. v. Laar,
Results of audiometric research at some hundreds of persons,  working
in different Dutch factories
Publication:  A. G. /S.A.  C 23 of N. I. P. G. - TNO.

A. Spoor,
Presbyacusis values  in relation to noise-induced hearing loss
Int. Aud.  6 (1967) 48-57.

C. W. Kosten and G. J. van Os,
Community reaction criteria for external noises
The Control of Noise, NPL-Symposion no.  12, P.  373-382, HMSO
1962.

-------
        TABLE 2a
Predicted N1PTS for 75 dBA





^
H
5
A N
u -
^ •— 4
V
O, in
03 ^
n
.•«
i
1 .
a< fy-
w ""1
<


90

75

50
25
10
90
75
50
25
10
90
75
M 50
N
25
10
90
75
MI 50
91 25
10
90
75
Ml 50
VD| 25
10
90
75
00
50
25
10
er-Vermeer
'A
u
m
CO
it)
|li


0.0

0.0

0. 0
0.0
0.0
2. 5
1. 5
.5
0.0
0..0
0.0
0.0
0,0
0.0
0.0
10.0
6.0
2.0
Q.O
0.0
2.0
2.0
2.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0
c
o
to
• r*
ja
o
K
10 Year

. 8

. 5

.3
. 2
. 1
1.5
.9
.6
.3
.2
1.6
1.0
.6
.4
. 3
3.6
2.3
1.4
.8
.5
2.5
1.6
.9
.5
.3

_
-
-
_

j3
U)
ft
a
ra


—

••

-
-
—
-
_
_
-
_
_
.
-
-
_
-
-
-
-
_
-
-
• ^
-
—
„
•,
_
_
V
0)
o
t>
Passchier-'
20
0.0
0.0
0.0
0.0
0.0
2.5
1.5
. 5
0.0
0.0
0.0
0.0
0.0
0.0
0.0
10.0
6.0
2.0
0.0
0.0
2.0
2.0
2.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0


Robinson
Year
1. 1
. 7
.4
.3
.1
2.1
1.4
.8
.5
.2
2.0
1.3
.8
.5
. 3
5.2
3.4
2.0
1.2
.7
3.6
2.3
1.4
.8
.5
—
_
_
_
_


fl
A
cd
CQ

_
_
-
-
-
—
_
-
.
-
„
-
-
_
-
—
H
-
-
•
.
„
-
.
-

_
_
_
—
to
V
i
Passchier-Ve:
40
0.0
0.0
0.0
0.0
0.0
2.5
1.5
. 5
0.0
0.0
Q.O
0.0
0.0
o.o
o.o
10.0
6.0
2.0
0.0
0.0
2.0
2.0
2.0
1.0
0.0
0.0
0.0
0.0
0.0
0.0

Robinson
Year
1.6
1.0
.6
.4
.2
3.0
2.0
1.2
.8
.4
3.0
1.9
1.1
.7
.5
7.5
4.9
3.0
1.8
1.1
5.2
3.4
2.0
1.2
.7
—
_
_
-
_

bo
1

^
.
-
_
-
„
-
_
_
-
^
-
.
.
-
^
_
-
.
•
.
-
.
-
.
_
-
-
_
           4

-------
        TABLE 2b
Predicted NIPTS for 80 dBA



^^
ffi
u .

cu m
m
•4
A tu
y •*
a, M
CO -^
in


90
75
50
25
10
90
75
50
25
10
90
75
M 50
(VJ
25
10
90
75
Wl 50
'1 25
10
90
75
&
50
*° 25
10
90
75
col 5°
°° 25
10
chier-Vermeer
W
      C
       I       s     4         3      §
       «       .5     M         a     .S
       »       -°     2          »     ja
       at       o     n          rt      O
       PH       tf     fl          p^     gj
         20 Year _          40 Year _

      0.0    2.2     0.0        0.0    3.2   0.0
      0.0    1.4     0.0        0.0    2.1   0.0
      0. 0     .8     0. 0        0. 0    1.2   0. 0
      0.0     .5     0.0        0.0     .7   0.0
      0.0     .3     0.0        0.0     .4   0.0
      3.5    4.0     1.0        3.5    5.6    .8
      2. 5    2.6     1.0        2. 5    3. 7    .7
      1.5    1.6      .9        1.5    2.3    .5
       .2     .9      .8         .2    1.4    .4
      0.0     .6      .7        0.0    1.0    .3
      0.0    3. 8     -         0.0    5. 5
      0.0    2.4     -         0.0    3.6
      0.0    1.5     -         0.0    2. 2
      0.0     .9     -         0.0    1.3
      0.0     .5     -         0.0     .8


     13.8    9.3    4.1       13.8   12.9     .9
      9.9    6.2    3.9        9.4    8. 7    1.5
      6.0    3.8    3.7        5. 5    5. 5    2.2
      1.0    2.3    3.1         .5    3.4    2.8
      0.0    1.4    2.5        0.0    2.1    3.1


      3.8    6.6     -         3. 8    9. 3
      3.6    4.3     -         3.6    6.2
      3.4    2.6     -         3.4    3. 8
      2. 1    1.6     -         2. 1    2. 3
       .8    1.0     -          ,8   .1.4

       .6            -          .6
       .4                       .4
       .2            -          .2    -
       .2            -          .2
       .2                       .2

-------
u
CO  .
  in
                                       TABLE 2c

                               Predicted N1PTS for 85 dBA
              f4                          fe                        U
              «                          »                        «
              4)                          0)                        0

              U                          »                        »
              •rj       O       _,         *d       O      rt          *r}      O
              o       2      ,2          u       2      •*          o      S
              «0       .5       M          09      .5      J?          W      .H
              00       J3       P          TO   ,   J2      ••:          *      J3
              
-------
  n
  >
   H
  8
•8 3
o -:
  in
                                         TABLE 2d

                                 Predicted NIPTS for 90 dBA
               t*                         M                        t*
               0)                         «                        4>
               V                         V                        V

               4)          "               4)                        4>
               ^                         ^                        ^
               I                          I                         I
               o)      d                  v       d                v      d
               s      g       d          2       g     «          s      i     d
               8      .S       •§>          S      .3     f          85-4
                  10 Year20 Year40
90     2.4      4.2    5.5        3.2    5.1    6.9        4.5    8.6    7.3
75     1.6      2.4    3.8        2.4    3.1    4.9        3.8    5.4    5.5
50      .8      1.5    2.6        1.6    2.0    3.3        3.1    3.5    3.2
25      .6      1.0    2.0        1.4    1.1    2.6        2.8    2.1    3.0
1°      .5      .8    1.8        1.2     .8    2.2        2.5    1.4    2. 5
90     7.3     7.8    11.6        8.3    9.8    9.8       9.5    13.8    6.6
75     6.4     5. 1     8. 5        7.0    6.0    8. 5       8. 2     9. 8    6. 1
50     5.1     3.3     6.3        5.7    4.5    7.1       6.9     6.7    5.4
25     3. 7     2. 1     4. 4        4. 3    2. 8    5. 7       5. 5     4. 3    5.9
10     2,3     1.5     3.3        2.9    1.9    4.7       4.1     2.7    6.2
      90     6.8     8.8      -         9. 2   12.2     -        13.4   16.4
      75     4.6     5.8      -         7.0    8.3     -        11.2   11.5
      50     2.4     3.6      -         4.8    5.2     -        9.0    7.4
      25     1.6     2.1      -         4.0    3.1     -        8.2    4.6
      10      .8     1.4      -         3.2    2.0     -        7.4    2.9
      90    23.6     18.8   30.1       23.6   24.0   18.7      23.6    29.5   46
      75    20.8     13.4   22i7       20.8   17.8   19.2      21.3    22.9   7*8
      50    18.0      8.7   17.4       18.0   12.1   18.6      18.5    16.3  lo!4
      25    13.2      5.5   11.5       13.2    7.8   14.9      13.7    10.9  14.8
      10     8.4      3.5    7.7        8.4    5.1   12.4       8.4     7.3  17.'4
      90    18.3     14.2     -        18.3   18.8     -        18. 3   24.0
      75    15.6     9.8     -        15.6   13.4     -        15.6   17.8
  Ml  50    12.9     6.2     -        12.9    8.7     -        12.9   12.0
    1  25     6.. 7     3.8     -         6. 7    5.5     -         6.7    7. 8
      10      ..5     2.4     -          .5    3. 5                .5    5.1
      90     8.9       -       -         8.9             -         8.9
      75     6.7       -       -         6.7             -         6.5
  •gl  50     4. 5       -       - -        4. 5      -      -         4.5
      25     4. 5       -                 4. 5             -         4.5
      10     4. 5       -       -         4. 5             -         4. 5

-------
   Iff
  s
u
O -*
0>
ft
w
   fc^
  5
OT
                                          TABLE 2e

                                 Predicted NIPTS for 95 dBA
                                         ft>
               0)                         4)                        Q)

               i                         *                        t
               *                         h                        ti
               2      g                  0      c                 v     «
              2gg          22S          2     g
               S      -5      -S          8     .«      f          85
               m      J3       J3          SO      JO      3          WpQ
               rtOrt          rt      °      .5          cSO
              & _ « _ W          &      «      «          PH     tf
                 10 Year	          20 Year                 40 Year
      90      5.6    8.1     9.7        7.9    12.6   11.7      11.6   15.1   12.9
      75      4.5    5.1     6.8        6.8     8.4    8.2      10.5   10.2    9.1
      50      3.4    3.3     4.6        5.7     5.6    5.5       9.4    6.9    6.1
      25      2. 3    2. 1     3. 7        4.6     3. 5    4. 5       8. 2    4. 6    5.0
      10      1.2    1.4     3.0        3.5     2.3    3.7       7.0    3.0    4.1
      90     12.1    13.0    17.6       13.7    17.7  14.1       15.3   10.6   11.1
      75     10.8     9.1    13.0       12.5    12.9  12.0       13.6   14.3    9.3
      50     9.5     6. 2     9.4       11.3     9.1   9.9       11.9   10.0    7.7
      "     7-8     4-°     6.3       9.6     6.0   7.9        9.2    6.8    8.6
      10     ^ 1     2. 7     4. 7       7.9     3. 7   6.6        7.5    4. 4    9.0
      90     12.4   14.9       -        18.2    19.6     -        27.6   24.9
      75     9.1   10.3       -        14.9    14.1     -        24.3   18.6
  Wl  50     5. 8    6.6       -        H.6     9.2     -        21.0   12.7
    '  25     2.6    4.0       -         8.4     5.8     -        17.8    8.3
      10     0.0     2.6      -         4.2     3.8     r        13.6
                                                                       5.4
      90    31.4   27.7    41.2       31.4    33.1   21.3       31.4   38.1     5.8
      75    29.7   21.2    31.7       29.7    26.6   23.6       29.7   32.0     9.8
      50    28.0   14.8    23.7       28.0    19.5   23.1       28.0   24.7    12.7
      25    24.5    9.8    14.1       24.5    13.4   18.1       24.5   17.8    19.4
      10    21.0    6.5     9.8       21.0     9.1   15.5       21.0   12.6    23.9
      90    25. 7    22. 2      -       25. 7    27. 7     -        25. 7  33. 1
      75    22.1    16.3      -       22.1    21.2     -        23.1  27.6
  !g|  50    18.5    10.9      -       18.5    14.8     -        19.5  19.5
      25    n-4     6.9      -       11.4     9.8     -        12.4  13.4
      J0     4.3     4.5      -        4.3     6.5     -         4. 3   9. 1
      90    15.1     -        -       15,1      -      -        15.5
      75    12.1              .       12.1      -      -        12.5
      50     9.1     .        .        9>1      .      _         9<5
      25     9.1              _        9.!      .      .         9>5
      10     9.1              -        9.1      -      -         9.5

-------
        TABLE 2f

Predicted NIPTS for 90 dBA



"5"
*%
u -
0) .
£•«>
w .
to
V
43 "2
o "P
0,™
to "^
irv



90
75
50
25
10
90
75
50
25

10
90
75
«| 50
N| 25
10
90
75
MI 50
^1 25
10
90
75
£
50
° 25
10
90
75
S
GO
50

25
10
schier-Vermeer
u
«s


2.4
1.6
.8
.6
.5
7.3
6.4
5.1
3.7

2.3
6.8
4.6
2.4
1.6
.8
23.6
20.8
18.0
13.2
8.4
18.3
15.6
12.9
6.7
.5
8.9
6.7
4.5

4.5
4.5
§
09
1

O
10 Year

4.2
2.4
1.5
1.0
.8
7.8
5.1
3.3
2. 1

1.5
8.8
5.8
3.6
2.1
1.4
18.8
13.4
8.7
5.5
3.5
14.2
9.8
6.2
3.8
2.4
—
-
_

-
-
rt
"rt
P K
O



1.0
4.3
4.3
3.3
1.3
_
-
-
^

—
-1.0
4.0
4.0
3.0
0
5.0
11.0
9.0
3.0
1.0
-8.0
2.0
3.0
4.0
7.0
—
-
_

-
-
JH
4>
g
Paaschier-Ven
20
3.2
2.4
1.6
1.4
1.2
8.3
7.0
5.7
4.3
2.9
9.2
7.0
4.8
4.0
3.2
23.6
20.8
18.0
13.2
8.4
18.3
15.6
12.9
6.7
.5
8.9
6.7
4.5
4.5
4.5


Robinson
Year
5. 1
3. 1
2.0
1.1
.8
9.8
6.0
4.5
2. 8
1.9
L2. 2
8.3
5.2
3. 1
2.0
24.0
17.8
12.1
7.8
5.1
18.8
13.4
8.7
5.5
3.5

-
-
*•
-


a
15
P Jj
DC v
°o

10.6
8.0
5.0
2.0
1.0
mm
.
_
_
-
14.0
16.0
9.0
5.0
2.0
26.0
24.0
20.0
15.0
13.0
10.0
19.0
18.0
12.0
6.0

-
-
-
.
                                 V
                                 V
Passchier-Ve
40
4.5
3.8
3.1
2.8
2.5
9.5
8.2
6.9
5.5
4. 1
13.4
11.2
9.0
8.2
7.4
23.6
21. 3
18.5
13.7
8.4
18.3
15.6
12.9
6.7
.5
8.9
6.5
4.5
4.5
4.5
Robinson
Year
8.6
5.4
3.5
2. 1
1.4
13.8
9.8
6.7
4.3
2.7
16.4
11.5
7.4
4.6
2.9
29.5
22.9
16.3
10.9
7.3
24.0
17.8
12.0
7.8
5.1
_
-
-
-
-
NIOSH Data
34 Year

13.3
13.0
6.3
2.6
1.3
mm
• _
-
-
-
20.0
27.0
12.0
4.0
5.0
-10.0
14.0
20.0
7.0
5.0
-3.0
12.0
22.0
9.0
10.0
_
-
-
-
-

-------
          3. Other Methods.  The National Institute of Occupational Health
and Safety (NIOSH) also presented data which have not been smoothed.   Table
2f has some of these same data incorporated for comparison.  This data base
was not used because (1) it only predicts NIPTS for 90 dBA, (2) the sample
size was very small (22 workers for some of the age groups),  and (3) gome
type of smoothing of the data would be required in  order to make it a pre-
dictive method.   The data is presented in Table 2f in order to  show (1) that
raw data requires treatment (such as provided by Robinson, Passchier-
Vermeer  or Baughn) before it is useful, and (2) the NIOSH data is  not out of
line with the predictive methods used in this report. There is, however,
one method  in the literature which differs greatly with other methodologies.
This is Kryter's latest work published in the Journal of the Acoustical Society
of America, 1973.
             Figure 1 shows a plot of predicted NIPTS values for  each of the
three selected methods as  well as Kryter's predicted values.  Of all the studies
compared,  only Kryter does not seem to be in general agreement with the
three methods selected.  Therefore, a special discussion of his method i>
included.  At this point,  however,  attention will focus only on  the methods of
Passchier-Vermeer, Robinson, and Baughn.
          4. Simplification of Data.  Now that three different  methods have
been selected,  the question remains as to how to use the data.  The data are
simplified to three curves  (representing different philosophies of what and
whose hearing should be protected) for three audiometric  frequencies.  Two
curves are  the expected NIPTS (maximum and a 10 year exposure  point) of
of the  sensitive  ears on the 90 percentile points with respect to SPL.  The
other curve is the average NIPTS expected during  40 years of  exposure as
averaged  over all the population percentiles.  This third curve is approximated
closely by the median NIPTS level after 20 years of exposure.  The three
audiometric frequencies presented were speech (average of 0. 5, 1, and 2 kHz),
speech (average of 0. 5,  1, and 4 kHz) and 4 kHz.   A Table relating percent
of population with more than a 5 dB NIPTS at 4000  Hz versus exposure is also
developed.  The data are presented in the sequence in which reduced so that
a user may, at his discretion,  stop and use as a basis of his decision the data
one or more steps before the manipulation that provides the final curves
discussed above.

          5. Details of Selected Methodologies.

             a)  Passchier-Vermeer (1971)
                 Passchier-Vermeer results are in graph form (see Figure  2).
Tables 3 and 4 are then used to calculate the effects of age and the correction
necessary for considering different percentile levels.  The details of the
calculations of the values in Table 2 are as follows:
                                    10

-------
  70
  60
  50
£40
Q
LJ

y
o
LJ
CC
QL
30
  20
   10
    40
       SPEECH ^(.5,1,2 KHz)
            75PERCENTILE

         8hr. DAILY EXPOSURE
            FOR 40 YEARS
                                 KRYTER
                                 PASSCHIER
                                 VERMEER
                                 BAUGHN
                                 ROBINSON
                             O	O

                             A	A

                             D	O
         50
60
90
100
                                                        10
                       70      80
                       SPL(dBA)

* KRYTER CONSIDERS THIS HEARING LEVEU FROM TABLE in [KRYTER (ie)]J
                        Figure 1
                            11

-------
                                                                4000 Hz
CD
TJ
O
I!
 o
o^
o
        MEDIAN HEARING  LOSS CAUSED  BY
        EXPOSURE TO NOISE  FOR 10 YEARS,
        AS A  FUNCTION OF THE NOISE RATING
        FOR 500  TO  2000  HERTZ.
                                               94
104
                                   dBA
                           Figure 2
                              12

-------
                          TABLE 3
                  (from PaBsehier-Vermeer)
Frequency    j  Increase;  of  1)5070 in relation to 050$ (T = 10)
* .v vr vi *•» w • • ™t/
for exposure time a
500 HR 2 #
1000 "
2000 »
3000 "
4000
6000 "

8000 "

2,5
10
1
0
0.
0.28 (NR-92)
0
0.37 (NR-92)
of at least 10 years 'j
per year
n
it
n
n
n
n
n
n
I
1




NR < 92
NR > 92
NR < 92
NR > 92
                          TABLE 4
                   (from PaB«chler-Verme«r)
NR for 500
{o 2000 Hz
75
80
85
90
94
98
Number of decibels to bo addod io Degj, in order io calculate Dyetf
500 Hz
0
0
0
0
0
0
1000 Hz
0
0
0
0
0
o.5
2000 Hz
0
1
2
3
4.5
7
3000 Hz
0
0
2.5
4.5
4.5
4.5
. 4000 Hz
4
3.5
3
2
0.5
0
6000 Hz
0
1
2.5
3.5
4
5
8000 Hz
0
1
2
^
3
3

NR for 500
to 2000 Hz
75
80
85
90
94
98
Number of decibels to bo substrocted from Dg£ ;in order to calculate bft% J
500 Hz
0
0
0
0
0.5
1.5
1000 Hz
0
0
0
0
0.5
1.5
2000 Hz
0
0
0.5
3
4
5
3000 Hz
1
1
2.5
3.5
3.5
3.5
4000 Hz
5
5
5
4
2
1
6000 Hz j SOCO Hz |
1
!•*
7
7.5
8
0
0
S
0
0
                              13

-------
Reference:  "Hearing Loss Due to Exposure to Steady-State Broadband Noise. "

(1)  Converted N. R.  into dBA by adding formula dBA = N. R. + 4.
(2)  Procedure used was outlined in pages 23-25.
(3)  Noise-induced shift of hearing level (Dx), not approximation of noise
    induced hearing loss (D'x) was calculated.
(4)  (Dx) values were  obtained from Figure R35-A and Tables A and B.
(5)  For 75 dBA,  the curves of R35-A were extended slightly by straight lines.
(6)  Speech hearing loss was obtained from averaging Dx for 500,  1000, and
    2000 Hz frequencies.
(7)  Since no method was suggested in her original report for estimating the
    10 and 90 percentile levels, the corrections used to estimate the 25 or
    75 percentile levels were doubled in order to  approximate the 10 or 90
    percentile levels.  The  error of this approximation will be less than
    10 percent for a normal distribution.  This is in agreement with
    Passchier-Vermeer's supplement (1969) to the main report.
      In her 1971 paper "Occupational Hearing Loss", Passchier-Vermeer
does provide NIPTS values for the 10 year exposure point.  These values
agree with the approximation used in this supplement.

             b)  Robinson

                Robinson provides a formula and a set of Tables (see
Tables  5 and 6) which can be used to calculate  NIPTS.  A nomogram is also
presented which allows calculation of hearing levels of noise-exposed popu-
lations  since the presbycusis correction is included.   Details of the calcu-
lations used to obtain the values of Table 2 are as follows:

Reference:  "The Relationships Between Hearing Loss and Noise Exposure. "
(1)  Used LA = dBA.
(2)  Used procedure outlined on page 18 except that the formula:

                                LA +  10 LOG T/TO -I- Un - Xi
      H = 27. 5
1 + TANK
                                             15
    was used instead of the nomogram.
(3)  Table 5 (page 6 of reference) was used to find \i for TO = I year.
(4)  Table 6 (page 7 of reference) was used to find Un, which relates H to a
    percentile of the population.
(5)  T = time of exposure in years and H = noise induced hearing loss.
(6)  Speech hearing loss was calculated from averaging H for 500,  1000 and
    2000 Hz frequencies.
             c)  Baughn

                Baughn presents  a set of Tables (see Tables 7 and 8) that
give the actual hearing levels of 8 different age groups for 9 percentile levels
under three exposure conditions.  Considering the 78 dBA group as non-
exposed groups, the calculations are as follows:
                                   14

-------
              TABLE 5
Frequency parameter  X in H-function
           (from Robinson)
Audiometric
frequency
(kHz)
0. 5
1
2
3
4
6

X (dB)
To = * year
130.0
126.5
120.0
1 14. 5
112.5
115,5
               15

-------
               TABLE 6
Percentile parameter  u in H-function
(from Robinson)
Percentile n
"Sensitive ears"
1 *
2
3
5
7
Deoile 1 0
15
20
Quartile 25
30
40
Median 50
60
70
'Quartile 75
80
85
Deoile 90
93
95
98
99 *
"Resistant ears"
u

13.8
12*1
11.1
9.8
8.7
7.6
6.0
5*0
4*0
3.1
1.5
0
- 1.5
- 3.1
- 4.0
- 5.0
- 6.0
- 7.6
- 8.7
- 9.8
-11.1
-12.1
-13.8
 Extrapolated.
                  16

-------
                                           TABLE 7                        (from Baughn)

                     INTERPOLATED AND EXTRAPOLATED FROM FIELD
                                     (Speech (.5,  1, 2 kHz)
Int.        AGE 18 - 23              ASH 24 - 29               AGE 30 -  35             AGH 36 - 4!
Pec.            5173                     3173                      S17T
Points    SO   8590    95      £0    S5~~  90    95      80    ST~  90    95       80
  1      -.8    -»4   -.1    .4      .2   1.0   1.7   2.7      .6   1.4   2.5   3.9      1.0   2.0   3.1

  2       .55   1.1   1.4   1.2     1.7   2.6   3.4   4.6     2.2   3.1   4.3   6.0      2.6   3.S   5.1

  3      1.5    2.1   2.4   3.0     2.8   3.8   4.6   5.0     3.3   4.3   5.6   7.5      3.8   5.0   6.5

  4      2.5    3.1   3.5   4.1     3.8   4.9   5.8   7.5     4.4   5.5   6.9   8.9      4.9   6.3   7.8

  5      3.7    4.4   4.8   5.5     5.2   6.4   7.4   9.0     5.8   7.0   8.6  10.8      6.4   7.9   9.6

  6      4.3    5.1   5.6   6.4     6.0   7.4   8.6  10.4     6.7   8.1  10.0  12.5      7.4   9.2   11.1

  7      5.0    5.9   6.5   7.4     7.0   8.6  10.0  12.2     7.8   9.5  11.6  14.6      8.6   10.7   13.0

  8      6.0    7.1   7.7   8.9     8.4  10.3  11.9  14.5     9.3  11.3.  13.8  17.4     10.3   12.7   15.5

  9      7.8    9.2  10.1  11.6    10.3  13.4  15.5  18.9    12.2  14.7  18.1  22.7     13.4   16.6   20.2


          AGE 42 -. 47               ACE 48 - 53               AGE 54 - 59               AGE 60 -  65

  1      1.6    2;7   3.9   5.3     2.7   3.7   4.9   6.7     4.1   5.3   6.8   8.4      6.8   8.0   9,2

  2      3.3    4.6   5.9   7.6     4.5   5.7   7.2   9.2     6.2   7.6   9.3  11.2      9.3   10.7   12,1

  3      4.5'   5.9   7.4   9.2     5.9   7.1   8.7  10.9     7.7   9.2  11.0  13.1     11.0   12.6   14.1

  4      5.7    7.2   8.8  10.8     7.2   8.6  10.3  12.7     9.2  10.8  12.8  15.0     12,8   14.5   16.1

  5      7. 3   8.9  10.7  12.9     8.!)  10.4  12.3  14.9    11.1  12.9  15.0  17.5     15.0   16.9   18.7

  6      8.5   10.3  12.4  15.0    10.-!  12.1  14.3  17.3    12.9  15.0  17.4  20.3     17.4   19.6   21.7

  7      9.9   12.0  14.4  17.4    12.1)  14.0  16.6  20.1    15.0  17.4  20.3  23.6     20.3   22.8   25.2

  8    J1.8   14.3  17.2  20.8    14.:;  16.7  19.8  24.0    17.9  20.8  24.2  28.2     24.2   27.2   30.1

  9    15.3   18.7  22.5  27.1    18.7  21.8  25.8  31.3    23.3  27.1  31.5  36.8     31.5   35.5   39.3

-------
00
TABLE 8
4000 Hz
Int.
Dec.
Points
1
2
5
A
5
6
7
8
9
1
2
3
4
5
6
7
8
9
AGE 18 - 23
78
.37
1.44
2.34
3.28
4.1
5. OS
6.85
7.95
10.7
AGE
8.32
13.5
18.2
22.1
26.0
30.2
35. 6
41.3
50.4
W
2.09
4.09
5.66
7.13
8.7
10.6
13.05
16.18
23.66
42 - 47
19.3
26.7
31.6
36.1
41.0
45.1
50.0
56.2
64.8
92
2.8
5.74
8.40
11.48
14.0
17.5
21.8
26.6
37.2
23.9
30.8
35.9
41.4
46.0
50.1
54.3
59.8
66.2
AGE
78
1.37
3.46
5.46
7.46
9.1
11.10
14.2
16.8
22.2
AGE
12.1
18.7
23.9
28.5
32.8
37.4
42.6
48.9
58.4
24 - 29
"TTC
5.7
9.88
12.9
15.96
19.0
22.4
26.6
28.5
45.03
48 - 53
24.1
31.6
37.1
41.8
'46. 4
50.6
54.8
60.3
67.7
92
7.67
13.15
17.54
23.01
27.4
32.61
38.91
46.03
61.38
30.5
37.1
41.7
46.7
50.8
54.4
58.4
63.0
69.1
AGE
78
3.0
6.15
9.15
11.87
14.3
17.12
21.31
25.17
32.6
AGE
17.2
24.5
30.9
35.3
40.1
44.9
49.7
56.1
65.0
(from Baughn)
30 - 35
— B6
9.77
15.6
20.09
23.99
27.9
32.09
37.11
43.52
53.01
54 - 59
30.3
38.2
43.4
48.1
52.3
56.5
60.1
64.9
71.7
92
12.6
18.9
23.8
30.5
35.0
39.9
45.9
52.5
64.8
37.3
43.4
47.2
51.6
54.9
58.2
62.6
65.9
71.4
78
5.2
9.6
13.4
16.8
20.0
23.6
28.6
33.4
42.4
24.0
32.3
39.2
44.1
49.0
53.9
57.8
63.7
70.6
AGE 36 - 41
4 V
"SET
14.7
20.9
25.8
30.4
34.9
39.1
43.2
51.0
60.0
AGE 60 - 65
35. S
43.6
48.1
52.0
55.9
59.3
63.7
66.5
72.7
92
17.9
24.4
29.6
35.7
40.6
45.5
50.3
56.0
64.1
44.0
50.4
53.3
56.3
58. 6
62.1
66.2
09.7
7S.O

-------
(1)  Use Table 7 (6a of reference) and Table 8 (9 of reference) from Baughn's
    data.
(2)  NIPTS for speech was considered as the difference in hearing of a certain
    percentile of people, who are exposed to a noise level greater than 80 dBA
    minus the. hearing  level of that  same percentile  of people who are exposed
    to only 80 dBA.
(3)  Percentile levels were  given in units of 10 percent only.  The 25 and 75
    percentile points were obtained by averaging 20 and 30, and 70 and 80 per-
    centile values, respectively.
(4)  The data was given by age groups with 6 year differences. Linear inter-
    polation was used where necessary to obtain exposures for 10, 20 and
    40 years.
(5)  HL values for 4000 Hz at 80, 85 and 90 dBA calculated from  Baughn's
    data by linear interpolation between the 78 and 86  dBA data points or the
    86 and 92 dBA data points.  Values at 95 dBA were obtained by linear
    extrapolation from the 86 and 92  dBA points.  NIPTS due to some exposure
    level, e. g. , 85 dBA,  was calculated as the HL at  85 dBA minus the HL
    at 78 dBA for the same  percentile and age  group.

          6.  Manipulation of Data. These values were manipulated and
simplified as follows:  Tables 9,  10 and 11 were constructed by averaging
the NIPTS values of Table 2 over a 40 year lifetime (age 20 to age 60).
After the NIPTS values were averaged over time for various population
percentiles, the results were averaged over the total population.  A graphic
method was used to calculate "Average NIHL during 40 Years Exposure".
The 0, 10,  20 and 40 year data points were plotted on  graph paper.  The area
under the curve drawn through these  points was measured and then divided
by 40 to obtain the "average NIHL during 40 Years'  Exposure." A graphic
method in which the . 9, • 75,  . 5, .  25, and . 1 percentile points were plotted
was used to calculate "Average Loss of Total Population During 40 Years of
Noise Exposure".  The area under  the resultant curve was measured and
normalized to obtain the desired value.
             From this average, Table 12 was developed.  Tables 13 and
14 come directly from the data of Table 2.  Table 13 provides the expected
NIPTS after 10 years of noise exposure that will not be exceeded by 90 per-
cent of the  population (.9 Percentile level). Table 14  depicts the maximum
NIPTS that will be encountered during a typical 40 year exposure which starts
at age 20.  Normally this occurs at 60 years o£ age, but for 4000 Hz,
Passchier-Vermeer's  method shows  that this occurs after both 10 and 40 years
of exposure time, while Baughn's data indicates that this occurs  at the 10
year exposure point.
             The resulting  NIPTS  values of Tables  12, 13 and 14 are now
averaged over the three methods.  This grand  average is presented in Fig-
ures 3-8.  Figures 3, 4 and 5 compare the 3  different ways (Max NIPTS,
. 9 percentile; NIPTS after 10 year  exposure, . 9 percentile; and  average
NIPTS of total population during 40 years) of considering the data at three
                                    19

-------
                              TABLE 9
                Average NIPTS during 40 Years Exposure
                          1/3 (.5,  1, 2 kHz)
dBA
      Passchier-Vermeer   0
 80   Robinson
      Baughn

      Passchier-Vermeer
 85   Robinson
      Baughn
 Population Pe re entiles
.9      .75     .5     .25
.1
      Passchier-Vermeer   3.0
 90   Robinson
      Baughn

      Passchier-Vermeer   9.2
 95   Robinson
      Baughn
Average
Loss of Total
Population
0
2.0
0
0
1.3
0
0
.8
0
0
.4
0
0
.2
0
0
.9
0

.9
3.6
2.8
.5
2.4
2.0
.2
1.4
1.3
.2
.8
1.1
.1
.5
.9
.4
1.6
1.6

3.0
5.5
6.0
2.3
3.2
4.3
1.6
2.1
3.0
1.4
1.2
2.3
1.2
.9
1.9
1.9
2.5
3.5

9.2
11.0
10.2
6.3
7.5
7.2
5.5
4.4
5.0
4.4
3.1
3.8
3.5
2.1
3.4
5.8
5.2
5.7
                                 20

-------
              TABLE 10

Average NIPTS during 40 Years Exposure
          1/4 {. 5,  1, 2, 4 kHz)
Population Percentiles
dBA

80


85


90


95


Passchier-Vermeer
' Robinson
Baughn
Passchier-Vermeer
Robinson
Baughn
Passchier-Vermeer
Robinson
Baughn
Passchier-Vermeer
Robinson
Baughn

3.
3.
*

5.
6.
5.

8.
9.
8.

q
4
6
8

1
3
1

1
3
,8

14.7
15,
,8
13.3
.75
2.5
2.3
.7

4.0
4.2
4.0

6.9
6.4
7.2

12.1
11.7
10.7
.
1.
1.
•

2.
5
5
5
7

,9
2.7
3

5
4
6

11
7
8
.5

.7
.3
.0

.1
.7
.5
.25
• •
4
I

1
1
3

4
2
4

9
5
6
2
.8
,6

.6
.6
.0

.3
.7
.9

.4
.3
.9
.1
0 .
.6
.6

.3
1.0
2.7

3.0
1.9
4.3

7.9
3.6
6.4
Average
Loss ot Total
Population
1.4
1.7
<

2
3
3

5
4
6

11
8
9
.7

.9
.2
.7

.7
.9
.3

.1
.5
.0
                    21

-------
                               TABLE 11
                  Average NIPTS during 40 Years Exposure
                                4000 Hz
                         Population Percentiles
dBA
 80
                                                      Average
                                                      Los a of Total
Passchier-Vermeer 13.8
Robinson
Baughn
      Passchier-Vermeer  17.8
 85   Robinson
      Baughn
      Passchier-Vermeer  23.6
 90   Robinson
      Baughn

      Passchier-Vermeer  31.4
 95   Robinson
      Baughn
.
-------
                         TABLE 12
                Average Loss of Total Population
                  during 40 Years of Exposure
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
                                1/3  (.5,  1,  2 kHz)
75
-
-
-


75
-
-
«•


75
-
-
-

80
0
.9
0
.3
1/4
80
1.4
1.7
.7
1.2

80
5.5
4.2
3.0
4.2
85
.4
1,6
1.6
1.3
(.5, 1
85
2.9
3.2
3.7
3.2
4000
85
10.6
7.4
10.0
9.3
90
1.9
2.5
3.5
2.6
, 2, 4 kHz)
90
5.7
4.9
6.3
5.6
Hz.
90
17.0
12.0
14.7
14.6
95
5.8
5.2
5.7
5.5

95
11.1
8.5
9.0
9.5

95
26.9
18.3
19.0
21.6
                             23

-------
                          TABLE 13
                 Noise  Induced  Hearing  Loss
               90 Percentile  Level  -  10  Years
                                1/3  (.5,  1,  2  kHz)
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
Passchier-Vermeer
Robinson
Baughn
Average
75
0
.8
0
.3

75
2.5
1..5
0
1.3

75
10.0
3.6
0
4.5
80
0
1.5
0
.5
1/4 (.
80
3.5
2.7
1.3
2.5

80
13.8
6.6
5.3
8.6
85
.9
2.8
2.5
'2.1
5, 1, 2 &
85
5.2
5.0
6.5
5.6
4000 Hz
85
17.8
11.6
18.6
16.0
90
2.4
4.2
5.5
4.0
4 kHz)
90
7.3
7.8
11.6
8.9

90
23.6
18.8
30.1
24.0
95
5.6
8.1
9.6
7.8

95
12.1
13.0
17.6
14.2

95
31.4
27.7
41.2
33.4
                           24

-------
                          TABLE 14
        Maximum Hearing Loss from Noise .9 Percentile
                                 1/3  (.5.  1,  2 kHz)
 Passchier-Vermeer
 Robinson
 Baughn
 Average
 Worst Case
 Passchier-Vermeer
 Robinson
 Baughn
 Average
 Worst  Case
Passchier-Vermeer
Robinson
Baughn
Average
Worst Case
*This maximum value is  for  10  years
maximum occurs at 40 years).
                             25
75
0
1.6
0
.5
Use

75
1.9
3.0
0
1.6
Use

75
10.0*
7.5
0
5.8
10.0
80
0
3.2
0
1.1
Robinson
1/4 (.
80
3.5
5.6
1.3
3.5
Robinson

80
13.8*
12.9
5.3*
10.7
.13.8
85
1.1
5.8
3.9
3.6
's Data
5, 1, 2 &
85
5.2
9.5
6.5
7.1
•s Data
4000 Hz
85
17.8*
20.5
18.6*
19.0
20.5
90
4.5
8.6
7.3
6.8

4 kHz)
90
9.5
13.8
11.6
11.6


90
23.6*
29.5
30.1*
27.7
30.1
95
11.6
15.1
12.9
13'.?.


95
15.3
19.6
17.6
17.5


95
31.4*
38.1
41.2*
36.9
41.2
(Otherwise the

-------
          SPEECH  (.5,1,2)
                                           MAX. NIPTS
                                            .9PERCENTILE
                                           NIPTS  AT 10 YR
                                            .9PERCENTILE
                                           AVERAGE NiPTS
                                           OF TOTAL POPULATION
                                           DURING 40YRS.
    75
80       85       90
      SPL(dBA)
95
100
Figure 3. Predicted NIPTS Averaged over the Methodologies of
        Passchier-Vermeer, Baughn and Robinson.
                          26

-------
  20
   15
m
3  10
E
z
         SPEECH i-(.5,l,2t4)
                                                MAX
                                                NIPTS .9PERCENTILE
                                    NIPTS AT IOYR
                                    .9 PERCENTILE
                                    AVERAGE NIPTS OF
                                    TOTAL POPULATION
                                    DURING 40YRS
                             JL
    70
75
80
    85

SPL(dBA)
                                      90
95
                                          100
   Figure 4.  Predicted NIPTS Averaged over the Methodologies of
             Passchier-Vermeer, Baughn and Robinson.
                                 27

-------
   30
CO
  20
OL
   I S
   I w»
   10
              4000 Hz
                                          MAX
                                          NIPTS.9PER.
                                          NIPTS.9PER.
                                          AT 10 YEARS
   AVERAGE NIPTS
   OF TOTAL
   POPULATION
   DURING 40 YRS.
   70     75     80     85      90
                        SPL(dBA)
95
100
 Figure 5.  Predicted NIPTS Average over the Methodologies of
           Passchier-Vermeer, Baughn and Robinson.
                              28

-------
      25
      20
    UJ
    o:
log
  K-X
    UJ
<^to
K6-K
       10
        70
               75
                                                 4000 Hz
                                                  (.5,1,2,4,^)
                                                  (.5,1,2KHz)
80      85      90     95

     SPL(dBA)
100
                            Figure 6.
                               29

-------
  50r
  35
  30
LU
  25
h-
UJ
o
Sao
G_
0>
CL
2
   10
                                         4000HZ
                                         ^(5,1,2,4 KHz)
                                         I (.5,1,2 KHz)
    70     75     80
                         85     90
                           SPL(dBA)
95     100
                    Figure 7.


                       30

-------
  40
   39
y  30

P
2
UJ
O

5  25
Q.
O>


CO
Q.

2
    70
                                          4000 Hz
                                            .5,1,2,4KHz)
75
80
  85     90
SPL(dBA)
                                                100
                       Figure 8.


                         31

-------
selected audiometric frequencies.  It is these sets of figures, along with a
set of Hearing Risk tables and one other table to be discussed later, that are
considered sufficient to select the permissible A-weighted SPL for the 8 hour
noise exposure.  Before such a selection is made, however,  certain other
observations should be considered in detail.
          7.  Considerations.

              a)  NIPTS at 4000 Hz may decline with exposure for the very
sensitive ears, while increasing for resistant ears.  Figures 9, 10, 11 are
a plot of the Hearing Levels of Baughn's data for  .9, .5, and . 1 percentile
levels.   Figure  12 is a plot of the difference between 85 dBA exposed groups
and 78 dBA exposed groups. As expected, during the first years of exposure
the sensitive ears (. 9  percentile) show a large increase in NIPTS while the
resistant ears (. 1 percentile ) show little increase.  After 40 years  of ex-
posure,  the situation is completely reversed.  If only the effect on the sen-
sitive ears is considered,  the NIPTS for the noise resistant ears could be
improperly neglected.
              It was for this  reason that the  "average NIPTS during 40 year*"
was calculated.  For instance, using the results for Table 11 for 85 dBA,
Baughn's method gives approximately 12 dB average NIPTS for the sensitive
(.9) ears and approximately 8 dB average NIPTS  for the resistant (. 1) ears.
Apparently the  entire population,  not just some super-sensitive individuals,
are significantly affected  by noise during some part of their lifetime at the
4000 Hz  audiometric frequency.  Essentially,  Table 11 was prepared to
show this effect.
              One of the  obvious reasons for the decline of NIPTS is seen
from Figure  11.  As the total loss of hearing increases, regardless of the
reason,  the influence of noise diminishes as there is only so much hearing
to be lost.  The unanswerable question that remains is "what causes such a
large hearing loss as evidenced by Baughn's (78 dBA) supposedly non-noise
exposed  group? "  Is it aging,  pathological conditions,  non-occupational noise
exposure greater than 80  dBA, the fact that 78 dBA may still be capable of
causing a very  significant loss in sensitive ears,  or some combination of
these factors?  Figure 13 is a plot of Baughn's 78 dBA {. 9) population versus
the 1960-62 Public Health Survey (PHS) data.   For the most part,  Baughn's
78 dBA (. 9) group shows less hearing  loss than the PHS group, until age 50,
at which point the  two groups become equal.   One can conclude that Baughn's
78 dBA (. 9) group does not differ significantly from the general population.
Baughn did not  screen for pathological conditions, so one would definitely
expect that such conditions would be an influence  in both groups.  The effect
of aging  cannot be neglected.  The rate of hearing loss for both the 78 dBA
group and the PHS (. 9) group is approximately 1. 5 dB/yr.  Such a steep
increase does not  occur for median hearing levels for  4000 Hz once  a certain
age is  reached  (such as 50-70 years).   It may not, therefore, be so  unlikely
that for this sensitive 10 percent of the population, aging alone causes a very
                                    32

-------
    80
    60
-J O
> <£
LLJ
CD
Z ^
DC ~
< d)
  *
    40
    20
                   FROM BAUGHN
               10 PERCENT HAVE BETTER HEARING
                 THAN LEVEL INDICATED (.1)
95dB(A)

90dB(A)

85dB(A)


78dB(A)
              20      40      6O

                  AGE (YRS)
                  Figure 9
                    33

-------
                 FROM  BAUGHN
80r
    60
^O
•^ co
LJ »-«
C£> O
2 Cft
CC

  5
    40
  - 20
         50% HAVE BETTER HEARING THAN
            LEVEL INDICATED (MEDIAN)
                             95dB(A)
                             90dBlA)
                             85dB(A)

                             78dB(A)
              20       40      60
                    AGE (YRS)
                   Figure 10.
                    34

-------
     60
LJ —
> O
LU CO
     40
     20
                   FROM BAUGHN
         90% HAVE BETTER HEARING THAN LEVEL
                   INDICATED (9)

                                  90dB(A)
                                  85 dB(A)
                                  78 
-------
             m
               / k
U)
         bJ
         LJ
40
         m5
         til O
         y ¥ 30
         " 00
         LU Q
         §iio
         cp cx)
         E
         LU
         X
               0
FROM BAUGHfJ
   4000 Hz
         10
       20
30     40
AGE(YRS)
50
60
70
                               Figure 12.

-------
   CO
     CO
     •a
^o
H f£
-   60
^5
t CD
£D "O
  CO

  « ^O
UJ O
i
^ u
LU UJ

JL. CD
20
  i
  I
  i
                  FROM  BAUGHN

                       4000 Hz
               20       40

                    AGE(YRS)
                            60
                                         90dB(A)
                  Figure 13.
                     37

-------
  80r
                      4000 Hz
          90% HAVE BETTER HEARING THAN LEVEL INDICATED(.9)
  60
O
v>
u
> 40-
UJ nw>
o
2
UJ
                                        O—O BAUGHN 78 (.9)
                                        X—X PUBLIC HEALTH
                                              SURVEY 1960-62
                                    I
                   20              40
                           AGE (YEARS)
60
                             Figure 14
                                38

-------
 significant change even in the early years.  These arguments are not brought
 forth to prove that the rapid loss of hearing at 4000 Hz for this segment of the
 population is not largely due  to noise exposure,  but rather to emphasize the
 converse; over-protecting the population against noise exposure to prevent  the
 rapid rise in hearing  loss at  4 Hz for 10 percent of the population may be
 entirely futile.  Such  over-protection could easily come about if one made
 the assumption that the 78 dBA  is the main cause of the large hearing losses
 in the sensitive 10 percentile.

               b)  Selection of a standard deviation for sensitivity to hearing
 loss.  Figures  12 and 13 demonstrate the difficulty of considering only mean
 data at some exposure time and from these data estimating various percentile
 levels by assuming a  standard deviation.  In order to predict Baughn's data,
 the standard deviation must be constantly changed for increasing exposure
 time.  This emphasizes the care that must be taken if a noise limitation is
 selected to protect 90 percent of the population instead of the median.   The
 90 percentile points can be seriously misestimated.

           8.  Risk of  Noise Relative to Hearing  Level Exceeding a Predeter-
              mined  Level or Fence. Up to this point discussion of hearing
 risk, as it relates to an increase of the  numbers of individuals who show a
 hearing loss greater than some  fence value, has not been undertaken.  The
 use of hearing risk as it relates to fences  has been used for some time.
 One of the major drawbacks to the use of fences, however, is that a single
 fence only considers or protects hearing of individuals whose hearing is al-
 ready near the fence values.  Since fences have  customarily been set relative-
 ly high with respect to the median hearing level, the hearing of the majority
 of the population is not considered.

               Simply  stated,  the object of the fence is not  to protect the
 excellent hearing from becoming just good, but the fair hearing from becom-
 ing bad.  The argument that the excellent hearing will automatically be pro-
 tected if the fair hearing is protected may not be true.  Figure 15 is such a
 counter example.  Thus the use of hearing risk  should not be the only basis
 for selecting a noise limit for hearing conservation.  Nevertheless hearing
 risk is one way to give meaning to NIPTS values and for this  reason Tables
 15 and 16 were prepared.  Table 15 shows the hearing risk in percentage
 as calculated by Robinson.  The 87, 92 and 97 dBA values were taken direct-
 ly from Robinson and the 80 dBA values  were calculated using his method.
 Table 16 shows the same data as calculated from Baughn's curves.  A typical
 curve from Baughn's data  is shown in Figure 16.  The data agree well only
if a 10 dB is  added to each of  Robinson's fence values.  This,  as proposed
by Robinson, will  account for the fact that Robins on1 a data  have been care-
fully screened for pathological hearing losses while Baughn's data have  not.
Baughn's data, in  this  regard, will certainly be more typical of the normal
population exposed to non-occupational noise. Therefore,  the 10 dB correc-
tion will be added  to Robinson's  fence values in this report.
                                   39

-------
        HYPOTHETICAL  SITUATION
       X=5dB
BEFORE
 NOISE
FENCE
25 dB
                    ,.5%
     -10 0  10  20 30
                FENCE
     B
      -10 0  10 20 30
        AFTER NOISE 10% HEARING RISK

     ,n * o,  MORE SENSITIVE POPULATION
     IU'° /0  AFFECTED MORE THAN LESS
        _  SENSITIVE.
                FENCE
                         AFTER NOISE 10% HEARING RISK
                 k vlo-5% ALL EAR AFFECTED WITH THE
                 NT      SAME AMOUNT OF NIPTS.
      -10 0 10 20 30
         B
      -10  0  10 20 30
            THERE  IS A 5 dB
            DIFFERENCE BETWEEN
            THE MEAN HEARING
            LEVEL DEPENDING ON
            WHICH ASSUMPTION
            USED.
                      Figure 15
                        40

-------
                   TABLE 15

               Robinson's Method

Noise risk for population at various ages for exposure
at constant noise level commencing at age 30.
Fence
Height
(ISO)
80 20
87 10*
92
80 25
87 15*
92
80 30
87 20*
92
80 35
87 25*
92
Noise
22 25
2
3
6
1
2
3
0
0
1
0
0
0
2
5
10
2
2
5
0
1
2
0
0
1
Risk
30
4
8
15
2
4
8
1
2
4
0
0
2
(%) at age
40 50 60
6
14
22
3
7
15
1
4
8
0
1
4
9
17
28
6
13
23
3
7
14
1
3
8
10
18
28
9
19
31
6
13
24
2
7
14
   *Use these fence values for non-pathological
population.
                         41

-------
  TABLE 16
Hearing Risk
Baughn's Data
Noise Fence
Level Height
dBA (ISO)
85 15
90
95
85 20
90
95
85 25
90
95
85 35
90
95
22
10
14
22
4
7
13
1
4
6
0
0
1
25
10
14
25
9
13
26
3
8
13
0
1
3
30
10
18
30
13
22
37
5
13
21
1
2
5
40
11
21
30
14
26
38
7
17
29
2
3
8
50
10
15

12
22
36
9
19
32
3
6
12
60 70



8
17
24
11
20
29
6
13
22
       42

-------
lOOp
                                                             100
                                                             95  m
                                                                 Q
                                                                 ID
                                                                 CO
                                                             90  g

                                                                 -
                                                                 UJ

                                                             85
                                                             80
                                                             GLORIG'S(I960)

                                                             NON-NOISE
                                                             EXPOSED

                                                             GROUP (2518)
           10
20
30      40      50

  AGE(yrs)
70
                            Figure 16.
                                 43

-------
          9.  Percent of the Population with more than a 5 dB NIPTS at
              4000 Hz Versus 8 Hour Noise Exposure Level. Since In general
the audiometric frequency at 4000 Hz is the most sensitive indicator of hear-
ing changes, a special table was derived to indicate the percentage of the
population expected to exceed a measurable NIPTS (greater than 5 dB) for a
daily 8 hour noise exposure of more than 40 years.  The expected NIPTS
for each of the Sound Pressure levels was calculated  or obtained graphically.
The NIPTS values of the three methodologies (Passchier-Vermeer, Baughn,
and Robinson) were averaged for the various percentile points.  These points
were plotted on probability paper and a line was drawn through them with a
French curve.   The intersect point with the 5 dB NIPTS line gives the per-
cent of the population that will exceed a measurable hearing change at that
exposure level. Table 17 is a summary of such data.

               It must be emphasized that this method is approximate only
and is very sensitive to errors in the basic data. To  emphasize this vari-
ability Table 18 was constructed in the same way as Table 17 except each
individual methodology was used alone;.

          10.  Selection of Limit for the 8 Hour Day.  Data have been presented
that should allow the setting of a maximum allowable  noise  exposure (8 hour)
based on several considerations.  The considerations  emphasized in this
report have been:  (a)  average NIPTS of total population during 40 years,
(b) NIPTS not exceeded by  90 percent of the population at any time during
their exposure history, (c) percent of the population with a measurable hear-
ing change at 4000 Hz, (d)  hearing risk as determined by a permissible hear-
ing loss or fence.   If desired, other considerations can be developed from
the data.  It is suggested that any recommended noise exposure be accept-
able with respect to all selected considerations.
         11.  Criticism  of Kryter's Method.
              a) From Figure 1 it is obvious that there is a very large dis-
parity between the predictions of Kryter and that of other researchers.
While Kryter may make  some valid points,  it is believed that there are
enough basic errors or inconsistencies in his methodology to make his re-
sulting predictions  invalid.  Therefore his NIPTS predictions were not con-
sidered in this document.

              b) Faults  and Inconsistencies of Kryter's Method

                 (1) Kryter arrives  at the conclusion that a non-noise ex-
posed population is that population that has not been exposed to a continuous
8 hour noise of 55 dBA.  This is based on extrapolation from Baughn1 s Data
and the Public  Health Survey of 1962.  The faults of this method are:

                    (a) Baughn1 s data are for 92, 86, and 78 dBA.  From
just these 3 points which span a range of 14 dB only it is very questionable
that it is justifiable to extrapolate another 23  dB downward to determine
                                   44

-------
                 TABLE 17
Derivation of % of Population with greater than
      5dB NIPTS after 40 years exposure.
L
eq
in
U "o
rt Q) a\
fc c •-<
W H X»
< s s
._ 2 o
CO *J S
H *3 «
S u
o

.9
.75

.5
.25

.1
% of Population with more
than 5 dB NIPTS
72


3.8
2.2

.7
.4

0

4
75


5.8
3.6

1.7
•6

.4

15
80


9.2
6.5

4.4
2.2

1.7

44
82


11
8.4

6.4
4.2

3. 1

66
85


13. 5
11. 5

9.8
7.8

5.2

92
                     45

-------
                       TABLE 18
Precent of Population with more than 5 dB NIPTS versus L
                    Individual Methods
                                                       eq
Le 5 dB NIPTS *
4000 Hz
Pas schie r -Ve rmee r,
Unmodified
Passchier-Vermeer
Straight Regression
Line
Baughn
Robinson
72
4
14
0
N/A
12
75
15
28
1
N/A
17
80
44
50
21
N/A
54
82
66
66
50
34
66
85
92
78
75
77
83
                            46

-------
where the threshold SPLi that causes NIPTS is located.  Furthermore,  most
of the three points do not even align in a straight line, thus requiring the
extrapolation be made by a series of complex curves (see Figure 17).

                     (b) Kryter uses two different reports, which probably
have different biases,  to determine the "NIPTS Threshold. " In fact Baughn
admits that he had a systematic error of at least 5 dBA  and perhaps  more in
his absolute thresholds.   For instance TTS was a problem as Baughn had to
test people during working hours.  The problems do not unduly jeopardize
the validity of Baughn1 s data when compared with itself as at least some of the
biases will be expected to cancel.  But when Baughn1 s data are compared to
other data,  such differences will not tend to cancel and must be fully con-
sidered.  Looking at the PHS curves and Baughn1 s 78 dBA curves versus age,
(Figure 18), it can be  noted that they look very similar except Baughn1 s
78 dBA  curve is displaced upward by 10 dBA.   Kryter would attribute this
upward  shift to the fact that the  78  dBA exposure was still causing a  sub-
stantial hearing loss.  But plotted also in Figure 18 is the median of  Baughn's
pre-exposure audiograms of new 18 year old  employees. Note that even for
this group,  there is still an 8 dB variation in  the Public Health Survey data
and Baughn1 s.  This variation shows that there were indeed systematic
differences between the studies. These differences may have  come from
audiometric techniques, differences in the population of this midwest area
versus the nation as a whole, or some other subtle bias; however, it is clear
that the 78 dB exposure is not,   a priori, the cause of the 10 dB discrepancy
between Baughn1 s data and the Public Health Survey data.

                     (c) In order to demonstrate the sensitivity of Kryter's
method  to systematic  error between the two sets of data, consider that the
hearing levels of Baughn1 s subjects were systematically 10 dB too high.
This 10 dB  error has significant implications with respect  to Kryter's NIPTS
threshold prediction.   See Figure 17 for a typical correction if Baughn's data
are reduced by 10 dB.  Such a 10 dB reduction now brings the  "NIPTS Thres-
hold" up to  75-80 dBA with far less extrapolation.  This puts Kryter more in
line with other researchers. It should also be apparent that the gain in "NIPTS
Threshold" was 20-25 dB for a  change of only 10 dB in Baughn1 s raw data.
This indicates that with an arbitrary fence of  so many dB,  the results obtained
are very sensitive to the absolute thresholds  of the data used.   One only has
to look at the literature to  see how often a 10  dB or greater difference has
occurred between researchers as to what is the median threshold level.  The
10 dB difference between the 1951 ANSI standard and the 1969  ANSI standard
for the  speech frequencies is an obvious example.  It should be noted that
even if the systematic difference in Baughn's  data was as small as 5 dB, which
is the minimum amount of  error predicted by Baughn, Kryter's methodology
would still predict that the threshold of the effect is at 65-70 dBA, not 55 dBA.
Therefore,  even if one would agree with Kryter that his methodology is adequate,
one must correct his threshold  value of 55 dBA by at least  10-15 dBA and
probably much more.
                                   47

-------
                            FROM KRYTER'S SIR
ACTUAL DATA
 POINTS
                         1	1	1
                  HlJS26dB OR GREATER
                - (Average 
-------
vO
     C3
     s
     Ul
          60
          50
          40
          30
     UJ t
     XS20
     o
     111
          10
         -10
                                        SPEECH (.5,1,2)
               Baughn's Pre-Exposure Audiograms
               of New Hire 18 yr. old Bnployees
                   10
20
30
40       50

AGE  (YRS)
                                        BAUGHN(78dBA EXPOSED
                                                     GROUP)
                                                  PUBLIC HEALTH
                                                 SURVEY (MALE
                                                  RT EARS)
                                                                   KRYTER
                                    ROBINSON (NON-NOISE EXPOSED)
60
70
80
                                               Figure 18.

-------
                 (2)  On Figure 18, Kryter's recommended presbyacusis
curves are plotted along with Robinson's.  Note that Robinson's values are
below Kryter's.  Yet Robinson has found that NIPTS for speech (0. 5,  1 and
2 kHz) essentially disappears for less than 75 dBA exposure.  This does
not fit with Kryter's assumption that 75  dBA is causing a very significant
shift in hearing.

                 (3)  Another inconsistency  of Kryter's NIPTS predictions
can be seen if these values are compared to the actual hearing levels of
Baughn's workers.   Figure 19 is such a comparison.  Somehow Kryter has
taken Baughn's data and manipulated the data such that the predicted NIPTS
is the same as the total hearing loss of these individuals.  Since hearing
loss consists of both NIPTS and aging, the only way to predict such a large
value of NIPTS, as I see it, is to predict that hearing will not change with
age.  This is clearly wrong, of course,  and even Kryter predicts 15 dB loss
from  presbyacusis  at age 65.

         12.   D-Versus A-Weighting of  Frequency. At first glance, the  use
of a D-weighting scale instead of an A-weighting might seem attractive.  The
D-weighting added approximately a 10 dB penalty to the frequencies that  are
more likely to cause NIPTS at the super-sensitive 4000 Hz audiometric
frequency. If one's goal is to protect the 3,  4 and 6 kHz frequencies  equally
with the lower frequencies of 0. 5,  1 and 2 kHz,  then perhaps the D-weighting
would be desirable.  However, D-weighting also emphasizes the frequencies
above 5600 Hz by 6-9 dB, and thus would tend to give  these high frequencies
more influence than they properly deserve.  The very low frequencies are
also emphasized more.  Thus protection of the speech frequencies of 0. 5,
1 and 2 kHz is slightly deemphasized. Qualitatively,  the argument reduces
to this:  if one desires that the risk of hearing loss should be equal for the
speech frequencies of 0. 5, 1 and 2 kHz and for the frequency of 4 kHz, then
the D-scale may be a slightly better approximation.  If  one is willing to
allow 5 dB more loss at 4 kHz than at the speech frequencies (0. 5,  I  and 2)
then the dBA is the better approximation.  The general feeling among most
investigators is that the frequencies of 0. 5,  I and  2 are aornewhat more
essential; therefore it is recommended that the A-scale be used for purposes
of hearing conservation.  The D-scale can be used to  predict the effects  of
noise on hearing, but the proper adjustments must be made  to provide the
same safety to the lower speech frequencies.
         13.  Duration of the  Exposure.

             (1) Less than 8 hours.  The relationships between NIPTS and
SPL discussed up to this point have been based on an 8 hour working day ex-
posure.  The auditory system can tolerate higher SPLs provided that the
exposure time is shorter (6).  It is not entirely clear, but it is suspected
that the SPL should be reduced: if the ear is exposed to noise for durations
greater than 8 hours.
                                   50

-------
  70r
   ».
  50
CD
T3

CO
  40
Q
LU
o
LJ
oc
o_
  20-
   10
   0
    40
SPEECH3(.5,l,2KHz)
    75 PERCENTILE

  8hr. DAILY EXPOSURE
     FOR 40 YEARS
                                    HEARING LEVEL OF
                                    THE EMPLOYEES OF
                                    BAUGHN'S STUDY
                                    (RE 1964 ISO-USE
                                    SCALE AT LEFT AS
                                    HEARING LEVEL)
                          KRYTER
                          PASSCHIER
                          VERMEER
                          BAUGHN
                          ROBINSON
                                                     A

                                                     D
 50
60
90
100
no
                   70     80
                   SPL(dBA)

* KRYTER CONSIDERS THIS HEARING LEVEL (FROM TABLE III
            [KRYTER (16)])
                        Figure 19
                           51

-------
                  The decision as how to relate SPL to duration in order to
obtain equally noxious noise exposure depends upon how the auditory damage
progresses with time.  Three popular theories are  equal energy (ISO stand-
ard for example),  equal pressure (Kryter for example) or a compromise
between equal energy and equal pressure (NIOSH for example).   The equal
energy rule predicts an equal hazard if the  SPL is reduced 3 dB for each
doubling of duration (SPL varies  inversely as 10 log t).  The equal pressure
rule dictates that the  SPL must be reduced  6 dB for each doubling of time
(SPL varies inversely as 20 log t).  The NIOSH compromise suggests that the
SPL should be reduced by 5 dB for each doubling of time (SPL varies inverse-
ly as 16. 6 log t).  The selection of one  rule over another is not a trivial
question.   For instance, considering the 8 hour exposure as the baseline,
equal pressure allows the permissible SPL for a one-minute exposure to be
27 dB higher than  that allowed for equal energy.

                  There is  a lack of unequivocal NIPTS data that would sug-
gest which rule to use.   Therefore,  equal TTS has been the only method for
assessing equal hazard.  This is why a considerable effort was given in the
main criteria document to the relationship of TTS (via animal and human
studies) to NIPTS.

                  Experimental results have not yet completely clarified the
problem.   Spieth and  Trittipoe (7) indicate that the equal pressure rule pro-
vides equal TTS for high level, short duration exposures.  Ward (8) has
found that equal energy best predicted an equal amount of TTS for chinchilla
during 4 exposure conditions.

                  Some sense can be made  out of the apparent contradictions
if the CHABA curves  are studied.  Figure 20 is a replot of the CHABA
curves  that relate equal TTS at various Sound Pressure Levels (SPL), dura-
tions and audiometric frequencies.  All curves, only for thepurpoees of com-
parison, were related to the same SPL value for the 8 hour duration.  Vari-
ous schemes for relating SPL to  duration are then plotted.  The results show
two main points.   These are,  (1) No  simple function of log t best matches
the CHABA values for all time durations  and (2) the selection of the function
used varies with the audiometric frequency that is to be protected.  At this
time, it is not suggested that a function other than the log t  be used since it
would effectively eliminate  the ability to provide dosimeters and perhaps
unduly complicate the situation.  The use of equal noxious TTS values is not
that firmly secure to  warrant such refinements.   Spieth and Trittipoe results
can be explained,  however,  since the durations with which they were con-
cerned  were  short. For exposures of 16 minutes and less, TTS at 4 kHz
does  start to follow the equal pressure law.

                  Using Figure 20 as a basis,  the decision as to which rule
to use reduces to which audiometric frequencies will be protected.  If 4000
Hz is to be protected,  then  the equal energy rule will be the best approxima-
tion.  If only the speech frequencies  of 0. 5, 1 and 2 kHz are to be protected,
                                   52

-------
01
           150
         cc 140
         m 130
         •a

         CO
         •O

         2 120
laJ
-J
a
2
Z>
           110
           ioo
£  95
o
uj  90

i  85

   80
                            FROM PARRACK(UNPUBUSHED)

                                         l   r~1   ' "
       A-WEIGHTED OVERALL &)UND LEVELS
       .SHORT DURATION LEVELS EQUIVALENT TO 85 dBA(OR OTHER BASE
       LEVELS) WHEN EXPOSURE DURATION IS 8HRS/DAY          D

      - EQUATION S>
                                                     IO
                                                       1*3
                                                );
                                                     a
      ® LEQUIVsLBASE + IOI°9|o| ^EQUAL ENERGY RULE)   O
          = 8HRS/DAY

         t = 8 HRS OR SHORTER DURATION AS SELECTED

         LEQUIV=LBASE
                                                 a
                                                      Q250HzBAND
                                                      C4000HZ BAND
                                                      y 1000 Hz BAND
                                            l  t
                                         I   II   I ii i  I il
                              I
                                   810 16 2432486496126 240
                                                 6080  128  256
                             8hr.       2hr.        30mia

                             RATIOS OF EXPOSURE DURATIONS
                                                  75min225sec.
                                                                  Imin.
                                      Figure 20.

-------
the NIOSH rule of 5 dB change in SPL for each doubling of time ia a very
good compromise.  Either rule will overprotect for short time durations
and as such will add an additional safety factor into any standard for hearing
conservation.  It should be noted that given an exposure level and duration,
Figure 20 can be used to directly predict the relation between such a condi-
tion and the dBA SPL of 8 hours duration that will cause the same amount of
TTS (or therefore NIPTS).  The usefulness of such a figure is limited,  how-
ever,  as typically a total daily noise exposure does not occur in such a simple
manner.  Therefore,  some approximation scheme  such as equal energy must
be used.  Correction  factors for  such variables as the intermittency of the
noise  are then required.

             (2)  Durations more than 8 hours.  There is a noticeable  lack
of actual NIPTS data on 24 hour exposure situations, therefore most of what
is known is  based upon TTS data.
                 Smith et al. (9) exposed groups of men for 25 hours to a
70 Hz tone or a 300 Hz tone at 113 dB SPL,  In general TTS ranged from
0 to 20 dB.  Yuganov  et al.  (10) simulated a 24 hour space mission with an
ambient noise of about 75 dB (not enough details are given to convert to dBA
but a rough  estimate would be 80 dBA) and found a  TTS of  10 to 20 dB with
recovery in 1-2 hours.  Mills (11) exposed himself to a 93 dB SPL signal
for about 30 hours and measured 25-27 dB TTS which required 2-4 days for
total recovery.  Melnick (12) exposed subjects for  16 hours to the 300-600
Hz octave band at 95 dB SPL and found the maximum TTS to be 15-20 dB.
Recovery was complete within 20 hours past exposure.  The Environmental
Protection Agency (EPA) is currently sponsoring research at the Aerospace
Medical Research Laboratory (AMRL) to further investigate this question
with human  subjects.  At this time, however, there is no evidence that the
effect of continuous noise is more noxious that what would be predicted by
use of the logarithm of time.  In fact,  several investigators (Mills,  Melnick)
have suggested that TTS reaches limiting value that may occur between 16-
48 hours.   Studies accomplished on animals (Mills and Talo (13); Melnick
(12); and Carder and Miller, (14) all predict that TTS will reach an asymptote
or a limiting value.  Exposures have been for as long as three weeks to three
months, with the TTS reaching its limit within the first day (Carder and
Miller (14) and Mills  (in Press)).  What is not so clear is the question,
Does hearing damage stop when such a limiting value that is independent of
duration is reached? " Based on Carder and Miller's animal findings that
similar recovering curves occurred once the asymptotic values were reached,
the answer appears to be a qualified yes if the TTS is less than 20-30 dB.
Recent work not yet published (Mills (in Press)) indicates that for greater TTS
than 30 dB,  such recovery may change with exposure time.  Since TTS will
normally be less than 30 dB only for exposures leas than 85 dBA, this  limit
will be considered valid only for exposures less than 85 dBA.  The  signifi-
cance of such a limit is that there may be  little1 difference between  a con-
tinuous lifetime exposure (24 hours exposure daily with no quiet periods) or
24'hour exposures with rest periods in between each exposure. Up to now, the
term 24 hour exposure has been used rather loosely  to mean either case.   We

                                   54

-------
will continue to use it in this context for exposures less than 85 dBA with
the justification that the asymptotic behavior of TTS allows such an approxi-
mation to be made.
                 The equal energy rule would predict that the 24 hour ex-
posure should be 5 dB less than the  8 hour exposure.  The NIOSH rule
would predict an 8 dB difference.  The animal results of Carder and Miller
show better correlation with the NIOSH rule.  The results of Melnick (1972)
on humans show that the equal energy hypothesis gives a better correlation
(it is even slightly conservative).
                 Preliminary results at AMRL have not shown the necessity
of deviating from the equal energy concept.  Therefore a 5 dB reduction in
dBA is considered the best approximation at this time for extrapolating
8 hour data to 24 hours.

                 If the SPL is below the value which causes measurable
TTS at 8 hours, then there is no evidence that there will be measurable
TTS at 24 hours.
          14.  Estimation of the Accuracy in Relating NIPTS to Noise Exposure.

              a) Underestimation Errors.

                 (1)  Worst case  of three methods.

                     Averaging the NIPTS predictions over the three methods
will provide in some cases lower  NIPTS predictions than one method by itself.
In order to estimate the worst conceivable situation, the worst case values
are included in Table 14. This table already consists of the maximum NIPTS
expected for the . 9 percentile level during some part of a 40 year exposure
lifetime.  Therefore selecting the highest predicted NIPTS value of the three
methods should set an approximate upper bound on the possible estimation
of NIPTS.  That  such an upper bound varies at the maximum byr only 4 dB
from the  average provides additional confidence that any prediction errors
in the average data presented are not likely to underestimate the risk of
noise by more than 4 dB.
                 (2)  Percentile estimates.

                     The estimation of NIPTS for some percentile  has been
accomplished by subtracting the hearing level of that percentile of the non-
noise exposed group from the hearing level of the respective percentile of the
noise exposed group.  The . 9 percentile group is thus that group whose
hearing level is worse than 90 percent of the population.  If the . 9 percentile
point moves 10 dB because of noise exposure,  then it is considered that the
. 9 percentile group had NIPTS of  10 dB.  However,  this 10 dB  shift could
have been caused by some of the exposed ears shifting from a . 1 percentile
hearing level to the . 9 percentile  hearing levels before the noise exposure,
then these exposed ears would have received a true NIPTS of 30 dB.  Un-
doubtably there are a few individuals who have this occur.  There is no way
                                   55

-------
to account for such individual susceptability and it must be emphasized that
all estimates are for statistical groups of the population, not individuals.
Changes in the . 9 percentile hearing level is still considered the best indica-
tor of the true NIPTS not exceeded by 90 percent of the population,  however,
for two reasons. First, the .9 percentile in a noise situation normally
does exhibit the greatest shift when exposed to noise. Apparently the people
that make up this group  are those most sensitive to the noise exposure.
Second,  changes in the . 9 percentile hearing level should be considered
more significant in that  the hearing of this group is already worse than 90
percent of the population.   A shift in this percentile point is  thus liable to
have more significance than a shift in the . 1 percentile point.

                     It can be  noted that the average NIPTS over 40 years
of exposure circumvents this problem. The errors introduced in saying that
90 percent of the population will have less NIPTS than some value X when
this NIPTS value was obtained  by changes in the . 9 percentile hearing  level
are difficult to estimate. If the changes in the . 9 percentile hearing level
are small, then one can reasonably expect that the error will be small.
But as stated earlier,  a better way to look at this problem  is to consider
that the . 9 percentile hearing level changes  are the most important measure.
In this light,  we will not unduly worry about this error.
              b) Over estimation Errors.
                 (1) "Least effect" of three methods.

                     Averaging over the three methods will also provide
higher NIPTS predictions than  some one method alone.  Similiar to the
worst case discussed previously,  the maximum difference  between a single
method and the average  is small.   In fact this difference is < 2 dB for the
speech frequencies (either 1/3 (0. 5,  1, 2 kHz) or 1/4 (0. 5, 1,  2, 4 kHz)
and  < 6 dB for 4000 Hz.

                 (2) Bias introduced in manipulation of the basic data.
                     Figure 21 shows how Passchier-Vermeer used the
data available to her for NIPTS at 4000 Hz.  On this  figure a curved line is
used to connect the data points represented.  One criticism of her work is
that a linear least squares regression line could have been used just as well.
As can be seen in Figure 21, a linear regression line will predict that the
median NIPTS threshold is at 80 dBA, not 7 or 8 dB  lower  as would be ex-
pected by extrapolating Passchier-Vermeer's existing curve.   It can only
be left up to individual judgement as to which approach is correct.  Using
a linear regression line, the NIPTS {. 9 percentile) would be expected to be
0 dB for 75 dBA (8 hour) exposure and 8 dB.for an 80 dBA (8 hour) expo sure.
This compares to a NIPTS (. 9) of 10 dB for  75 dBA and  13. 8 for 80 dBA.
At 85 dBA either approach predicts the same amount of NIPTS.  Therefore
the greatest possibility of error at the 4000 Hz audiometric frequency  is
below 85 dBA.   The average of the three methods produced 6 dB for 75 dBA,
                                   56

-------
      MEDIAN  AND MEAN HEARING LOSS  CAUSED  BY EXPOSURE TO NOISE
      FOR AT  LEAST  10 YEARS,  AS  A FUNCTION  OF SOUND LEVEL.
                     ( From Passchier-Vermeer )
05
a
A\
A\
r
 0
O
a
o
                 80
             90                   100
	*» SOUND- LEVEL  IN dB  (A)
104
                            Figure 21.


                               57

-------
 so the maximum error at 75 dBA is 6 dB.   Likewise,  it can be shown that at
 80 dBA this possible error is 3 dB.  Note that the magnitude of these errors
 is the same as was obtained by looking at the "least effect" of  the three
 methods.
              c)  In  summary,  the  4 kHz and (. 9 percentile) data presented
 in Table 17 can reasonable be considered accurate within a range of +4 dB
 and -6 dB (or more  simply Hh 5 dB) of the values given as long  as the L
 range under consideration is between 70 and 90 dBA.

      B.  Requirement for "Quiet"

          Recent work by Ward (15) has shown that the quiet intervals bet-
 ween high intensity noise-bursts must be below 60  dB SPL for the octave
 band  centered at 4000 Hz if recovery from Temporary Threshold Shift (TTS)
 produced is to be independent of the quiet period SPL.  Ward suggests
 55 dB SPL as the point where the "effective quiet"  might be.  Assuming then
 that (1) TTS recovery from a 90  dBA (8 hour) occupational exposure also
 requires this  same level of effective quiet for some part of the 16 hours
 between the exposure the following day, and (2) total TTS recovery  is impor-
 tant in order to prevent TTS from becoming NIPTS, noise exposure should
 be controlled in order to reasonably insure an effective  quiet of 55 dB SPL
 at the 4000 Hz octave band (approximately 62-65 dBA).  The population
 exposed to TTS producing  sources  (both occupational and non-occupational)
 will be guaranteed by such control  the availability of a quiet period  of less
 than 60 dBA.  That such a quiet period is really required is not absolutely
 proven, of course,  but there is enough evidence to suggest at this time that
 this approach is  advisable.


 in.   SUMMARY
      Selection of a  permissible 24 hour exposure will be 5 dB below the
 permissible 8 hour exposure SPL if equal energy is to be used.  Table 19
 summarizes the effects, as baaed on the 8 hour exposure, of exposures of
 either 8 or 24 hours for different SPLs. The expected absolute error is
 estimated to be well within 5 dB  for the NIPTS values predicted.   For Hear-
 ing Risk,  a fence of 25 dB (1964 ISO) is used.  Baughn's and Robinson's
 Hearing Risk values are averaged.  For the 85  and 90 dBA (8 hour) exposure
 conditions,  the resulting average is within jh 3 percentage points of Hear-
 ing Risk predicted by either method.   For an 80 dBA condition, Robinson's
 estimate (10 percent) and Bsughn's estimate (0  percent) were averaged to
'obtain 5 percent.  While these values might seem rather divergent, it is
 noteworthy that NIOSH predicted 3  percent for this level.  The Hearing Risk
 at 60 years of age was used.  Hearing Risks at younger  ages are less than
 these values (see Tables 15 and  16).
                                   58

-------
Table .19-    Summary of  effects  expected  for continuous noise
            exposure of 8 hours to  the levels  stated.
Max NIPTS  (.9)
NIPTS at 10 yr  (.9)
Average NIPTS
Max Hearing Risk*
                                    75 dBA (70  dBA for 24  hrs)
Max NIPTS  (.9)
NIPTS at 10 yr  (.9)
Average NIPTS
Max Hearing Risk*
Max NIPTS  (.9)
NIPTS at 10 yr  (.9)
Average NIPTS
Max Hearing Risk*
Max NIPTS  (.9)
NIPTS at 10 yr  (.9)
Average NIPTS
Max Hearing Risk*
    25 dB ISO Fence
Speech (.5,
1 dB
) 0
0
N/A
1, 2) Speech (.5, 1, 2, 4)
2 dB
1
0
N/A
4K
6 dB
5
1
N/A


Speech ( . 5 ,
1 dB
) 1
0
5%
80 dBA (75 dBA for 24 hrs)
1, 2) Speech (.5, 1, 2, 4)
4 dB
3
1
N/A

4K
11 dB
9
4
N/A


Speech ( . 5 ,
4 dB
) 2
1
12%
85 dBA (80 dBA for 24 hrs)
1. 2) Speech (.5, 1, 2, 4)
7 dB
6
3
N/A

4K
19 dB
16
9
N/A


Speech ( . 5 ,
7 dB
) 4
3
22.3%
90 dBA (85 dBA for 24 hrs)
1, 2) Speech (.5, 1, 2, 4)
12 dB
9
6
N/A

4K
28 dB
24
15
N/A

                                59

-------
IV.   CONCLUSIONS

      The main purposes for preparing this report were twofold.
      (1)  The first purpose was to resolve the question of what and/or whose
data should be used to depict the relationship between loss of hearing sensi-
tivity and noise.  The question was resolved by using three leading predictive
methodologies and averaging the results.  This averaging has been criticized
by some as unscientific.  The argument is that one should pick the most
scientifically sound method and use it alone.  But the problem then remains
of how to select the single best method.  Averaging the three methods avoid*
such a selection.  But even more important, averaging the three methods
prevents the possibility of selecting the worst method.  Therefore,  the
averaging technique was considered as the best way to handle the problem
of data selection.

      (2)  The second purpose of this  supplement was to discuss the method-
ology of Kryter (16).  Criticism of Kryter's paper is provided by several
reviewers in the same issue of the Journal of the Acoustical Society of
America.  At this time there are too  many basic inconsistencies in Kryter's
method for his  results to be included  in this report.
                                   60

-------
                              REFERENCES
 1.   Passchier-Vermeer, W. Hearing Loss Due to Exposure to Steady-State
      Broadband Noise.  Rept. No.  35,  Institute for Public Health Eng.,  The
      Netherlands,  1968.

 2.   Robinson, D.W.  The Relationships Between Hearing Loss and Noise
      Exposure.   National Physical Laboratory Aero Report Ae32, England,
      1968.

 3.   Baughn, W. L. Relation Between Daily Noise Exposure and Hearing
      Loss as Based on the Evaluation of 6835 Industrial Noise Exposure
      Cases.  In Press as AMRL-TR-73-53,  Aerospace Medical Research
      Laboratory, Wright-Patterson AFB,  Ohio.

 4.   Passchier-Vermeer, W.  "Steady-State and Fluctuating  Noise:  Its
      Effects on the Hearing of People" in Occupational Hearing Loss,
      D.W.  Robinson,  Ed.  Academic Press,  N. Y., 1971.

 5.   Robinson, D.W.  "Estimating the Risk  of Hearing Loss due to Continu-
      ous Noise!1.  In Occupational Hearing Loss, D. W. Robinson, Ed, Aca-
      demic Press, N. Y.,  1971.

 6.   Kryter K. D. W. ,  J. D. Miller,  and D. H. Eldridge.  Hazardous Expo-
      sure to Intermittent and Steady-State Noise.  J.  Ac oust.  Soc. Am.,
      Vol. 39, No.  3, p. 451, 1966.

 7.   Spieth, W. and W. J.  Trittipoe. Intensity and Deviation  of Noise Ex-
      posure and Temporary Threshold Shifts.  J.  Acoust. Soc. Am. , 30,
      710, 1958.

 8.   Ward, W. D. and D. A. Nelson.  On the Equal-Energy Hypothesis
      Relative to Damage-Risk Criteria in the Chinchilla.  In Occupational
      Hearing Loss, D.W.  Robinson, Ed.  Acad. Press, N. Y., 1971.

 9.   Smith,  P. F., M. S. Harris, J. S.  Russotti and C. K. Myers, Effects
      of Exposure to Intense Low Frequency Tones on Hearing and Perform-
      ance.  Submarine Medical Research Laboratory, Naval Submarine
      Medical Center Report No. 610,  1970.

10.   Yuganov, Ye, M.  et_al.  Standards for Noise Levels in Cabins of Space-
      craft During Long Duration Flights.  (Tech.  Transl.  F-529,  National
      Aeronautics and Space Administration,  Washington,  D. C.,  1969. )
                                  61

-------
11.   Mills, John H. ,  Roy W. Gengel, Charles S. Watson and James D.  Miller,
      "Temporary Changes of the Auditory System Due to Exposure to Noiae
      For One or Two Days". . J.  Acoust.  Soc.  Am. , Vol.  48, pp.  524-530,
      1970.

12.   Melnick,  W. Investigation of Human Temporary Threshold Shift (TTS)
      from Noise Exposure of 16 Hours Duration.  (Presented at the Acous-
      tical Society of America, December 1972 meeting.)

13.   Mills, J. H. and S. A. Talo. Temporary Threshold Shifts Produced by
      Exposure to High-Frequency Noise.  Journal of Speech and Hearing
      Research, Vol.  15,  September 1972, pp.  624-631.

14.   Carder,  H. M. and J. D. Miller,  "Temporary Threshold Shifts From
      Prolonged Exposure  to Noise, "  Journal of Speech and Hearing Research,
      Vol. 15, September 1972, pp.  603-623.

15.   Ward, W.  Dixon. The concept of "Effective Quiet,  " presented at the
      85th Meeting of the Acoustical Society of America,  April 1973.

16.   Kryter, K. W., "Impairment to Hearing from Exposure to Noise".
      J. Acoust. Soc.  Am.,  Vol.  53, No.  5, May 1973,  pp. 1211-1234.
                                      *U.S.Government Printing Office: 1973 - 758-426/78

                                   62

-------