Method For Calculating and Using The Noise Reduction Rating - NRR
Method For Calculating and Using The Noise Reduction Rating - NRR
The NR R is a single-number ra ting which is required by law to be sho wn on the label of each hea ring protecto r sold in the United S tates. The N RR is specified by 4 0 CFR (Code of Federal Regulations) P art 211, P roduct Noise Labeling, Subpart B - Hearing Protective Devices. It is independen t of the noise spectrum in which it is applied. The following description of the method of measurement and calculation of the NRR is taken from Method 2 of the NIOSH Comp endia (Kroes et al., 1975) and a similar table can be fo und in 44 FR (Federal Register) page 56142 (1979). Calculating the NRR: The values of sound attenuation used for calculation of the NRR are determined in accordance with ANSI S3.191974, "American Na tional Standard for the Measurement of Real-Ear Hearing Protector Attenuation and Physical Attenuation of Earmuffs." The experimenter-fit method must be used; that is, the expe rimenter (not the test subject) must fit the hearing protector onto the head or into the ear of each test subject for each occluded test. Mean attenuations and standard deviations are calculated in accordance with the stand ard. T he N RR is then computed from the mean attenuations and standa rd deviations according to the following equation:
8000
Af
-APVf98)
- 3 dB.
where L Af is the A-weighted octave band level at freq uency f of a pink no ise spectrum with an o verall level of 10 7.9 dBC, and APV f98 is the mean attenuation value minus 2 standard deviations at frequency f (2 standard deviations accounts for 98% of the variance in a normal distribution). The equation can be broken down into the steps shown in the table on the following page. The N RR assumes a pink noise with octave-band levels of 100 dB (line 1). The corrections for the C-weighting scale (line 2) are then subtracted to compute unprotected C-weighted octave-band levels at the ear (line 3). These octave-band levels are logarithmically summed to obtain the overall sound level in dBC at the unprotected ear; this value is the first term of the equation (i.e., 107.9 dBC ). The corrections for the A-weighting scale (line 4) are then subtracted from the pink-noise octave-band levels (line 1) to comp ute the A -weighted octave-band levels at the ear (line 5). The average attenuations (line 6) minus twice the standard deviations (line 7) are subtracted from the A-weighted octave-band levels to compu te the protected A-weighted octave-band level at the ear. Note that the attenuation data for 3000 and 4000 Hz and for 6000 and 8000 Hz are averaged to derive the attenuations at 4000 Hz and 8000 Hz. Note also that the standard deviations for 3000 and 4000 Hz and 6000 and 8000 Hz are summed (in lieu of multiplying by 2) to calculate the total standard deviation for 4000 Hz and 8000 Hz. The protected A-weighted octave-band levels at the ear are then logarithmically summed to calculate the overall protected A level. This is the second term of the equation. The N RR is computed b y subtracting 3 dB from the difference between the unprotected C-weighted and the protected A-weighted levels at the ear.
Using the NRR: The H earing Conservation Amendment to the Occupational Noise Standard (OSHA, 1983) describes six methods for using the NR R to d etermine a wo rker's pro tected A-weighted noise exposure. These methods vary according to the instrume ntation a nd param eters use d to d etermine the unpro tected noise levels. However, they can be summarized into two basic form ulas, depend ing on whether unprotected exposure levels were measured on a Cweighted or an A-weighted scale. For C -weighted measuremen ts: protected dBA = unprotected dBC - NRR where the protected dBA and the unprotected dBC are 8-hour time-weighted averages determined according to the Occupational No ise Standard . This m ethod is how the NRR was designed to be used. For example, if a protector has an N RR of 17 dB and it is used in an environm ental no ise level o f 95 d BC , the noise level entering the ear co uld be expected to be 78 dBA [95 - 1 7 = 7 8] or lower in 98% o f the cases if the pro tector is worn acco rding to manufacturers specification. For A-weighted measuremen ts:
Method for Calculating and Using the Noise Reduction Rating -NRR
protected dBA = unprotected dBA - [ NRR - 7 ] where, again, the protected and unpro tected dBA are 8-hour time-weighted averages determined according to the Occupational Noise S tandard. T his method is an adaptation for those whose instrumentation does not have C-weighting capab ilities. The 7 dB correction factor is used to account for the de -emphasis of low-frequency energy inherent to the A-weighting scale. So, for example, if a protector has an NRR of 17 dB and it is used in an environmental noise level of 95 dBA, the noise level entering the ear could be expected to be 85 dBA [95 - (17 - 7) = 85] or less in 98% of the cases. NIOSH recommends derating the NRR by a multiplicative factor of 75% for earmuffs, 50% for slow-recovery foam earplugs and custom earplugs, and 30% for all other earplugs. T his variable derating schem e con siders the realworld performance of most different types of hearing protector (NIOSH, 1998). Also, the NIOSH derating scheme doe s not affec t the 7-decibe l dBC to d BA corre ction as it is applied to the NR R only, dera ted or not. T his compend ium uses the NIOSH d erating of the NRR when searching for hearing protectors based on the compendium users input of noise exp osure levels in dB A or octave band levels.
Computation of the Noise Reduction Rating Octave Band Center Frequency (Hz)
Frequency in Hz 1. Assumed pink noise (dB) 2. C weighting corrections (dB) 3. Unprotected ear C-weighted level 4. A weighting corrections (dB) 5. Unprotected ear A-weighted level 6.Average attenuation in dB at each frequency 7. Standard deviation in dB at each frequency 125 100 -.2 99.8 -16.1 83.9 21 21 3.7 x2 7.4 250 100 0 100 -8.6 91.4 22 22 3.3 x2 6.6 15.4 500 100 0 100 -3.2 96.8 23 23 3.8 x2 7.6 15.4 1000 100 100 0 100 29 29 4.7 x2 9.4 19.6 2000 100 -.2 99.8 1.2 101.2 41 41 3.3 x2 6.6 34.4 43 3000 4000 100 -.8 99.2 1.0 101 47 41 45.0* 3.4 6.7** 38.3 6.1 6000 8000 100 -3 97 -1.1 98.9 36 38.5* 6.5 12.6** 25.9 Log Sum
107.9
3.3
8. Compute APV9 8 in dB at each frequency (line 6 - line 7) 13.6 9. Protected ear A-weighted level (average attenuation minus 2 standard deviations develops the ear A-weighted Levels (line 5 - line 8) 70.3
76.0
81.4
80.4
66.8
62.7
73.0
85.1 ___
10. NRR is unprotected ear "C" level (line 3) minus protected ear "A" level (line 9) minus 3 dB 19.8 * average attenuation at 3000 and 4000 Hz and at 6000 and 8000 Hz. ** summed standard deviation for 3000 and 4000 Hz and for 6000 and 8000 Hz.
Method for Calculating and Using the Subject-Fit Noise Reduction Rating - NRR(SF)
The Subject-Fit Noise Reduction Rating, NRR(SF), is a single number rating method which describes real-ear attenuation at thre shold (RE AT ) measurem ents ob tained acco rding to AN SI S1 2.6-1 997 , Methods for Measureing the Rea l-Ear A ttenuation of Hea ring Pro tectors , method B. The N RR(SF) was de veloped by the National Hearing Conse rvation Associations Task F orce on Hea ring Pro tector Effectivene ss to address labeling related issues (Ro yster, 19 95). The NR R(SF) method is derive d from the Noise R eduction R ating (N RR) presented in Calculating NRR and the Single-Number Rating (SNR) method presented in Calculating SNR for a protection performance of 84% (EP A, 19 79; IS O, 1992 ). In principle, the rating method s exam ine the amou nt of noise reduction achieved by a hearing pro tector, but minor difference s separate the metho ds. Like the NRR, the NRR(SF) uses a p ink noise at a level of 100 dB SPL in the octave bands. Like the SNR, the NRR(SF) includes the octave-band frequencies between 63 and 80 00 Hz in its calculation whereas the NRR includes only frequencies between 125 and 8000 Hz and includes the half-octave frequencies 3150 and 6300 Hz. Calculating the N RR (SF): The N RR(SF) is calculated with the following equation:
8000
Af
-APVf84)
- 5 dB.
where LAf is the A-weighted octave ban d level at frequency f of a pink noise spectrum with an overall leve l of 108.5 dBC, and APV f84 is the mean attenuation value m inus 1 standard deviation at frequency f (1 standard deviation accounts for 84% of the variance in a normal distribution). The C-weighted noise levels are determined by subtracting the C-weighted corrections from 100 dB for each of the octave band frequencies (line1). The sum of the C-weighted noise levels is 108.5 dBC. Similarly, the L Af (line 3) are determined by correcting the 100 dB octave band noise levels with the A-weight corrections at each frequency (line 4). The APV f84 are determined by subtracting one standard deviation (line 7) from the mean REAT attenuation measured for the no ise ban d cen tered at frequency, f (line 6). The half-octave data at 3150 and 6300 H z are not used in the calculation of the NRR(SF). The 5 d B which is subtracted at the end of the calculation is representative of the approximate difference between C-weighted and A-weighted industrial noise (Miller, 1995). By compensating for the difference in the calculation, the NRR (SF) may be subtracted directly from A-weighted noise to estimate the noise level under the protector. When data for 6 3 H z are lacking, o ne sho uld use the 12 5 H z data in its place; i.e., use the 125 -Hz da ta twice with the appro priate A and C weightings.
Using the N RR (SF): The NRR(SF) is intende d to b e used with A-weighted noise exposures. H ence, the NRR(SF) is subtracted d irectly from the A-weighted noise exposure to provide the protected exposure: protected dBA = unprotected dB A - NRR (SF). If the C-weighted noise exposure is used, a 5-dB adjustment is made to account for predicted differences in the A and C levels (which effectively undoes the 5-dB correction in line 12 of the following table): protected dBA = unprotected dB C -5 dB - NRR(SF). If the octave-band noise exposures are used, the best approach is to apply the A-weighting corrections and log sum the levels to provide the A-weighted noise exposure level for use with the NRR(SF ). Ava ilability of the NR R(SF ): The ado ption of ANSI S12 .6-1997 as the standard that replaced ANSI S3.19-19 74 did not change the regulatory requirements that all protectors sold in the United States be labeled with the NRR as described in Appendix A1 and obtained by testing according to the experimenter-fit method of ANSI S3.19-1974. After ANSI S12.6-1997 was adopted, many hearing protector manufacturers began testing their products in accordance with Method B , the subject-fit method. As manufacturers begin to submit their Method B data, the compendium will show both sets of
Method for Calculating and Using the Subject-Fit Noise Reduction Rating - NRR(SF)
attenuation data and both the subject fit NRR and original NRR. Ho wever, the Method B data will be used when a search for hearing p rotectors is pe rformed. If and when the U.S. Environmental Protection Agency is successful in promulgating a revised Hearing Protector Labeling regulation that requires the use of the NRR(SF)or som e other subject-fit based rating in lieu of the present NR R, then manu facturers will start providing data that they already have. In the absence of Method B data, NIOSH recommends derating the NRR using NIOSH derating schemes as described in the NIOSH re vised noise criteria do cument (C riteria For a Recomm ended Standard : Occup ational No ise Exposure Revised Criteria 1998. DH HS (NIOS H) P ub. N o. 98 -126 ). The derating schemes which are specific to hearing protector type and based o n real-world data are explained in Calculating NRR . NIO SH suggests that consumers contact manufacturers to request the NRR(SF ) for any protectors they are considering purchasing or recommending.
The NRR(SF) Calculation Frequency in Hz 1. Assumed Pink Noise (dB) 2. C-weighting correction (dB) 3. Unprotected ear C-weighted level (dB) line1 -line 2 63 100.0 -0.8 99.2 125 100.0 -0.2 99.8 250 100.0 0.0 100.0 500 100.0 0.0 100.0 1000 100.0 0.0 100.0 2000 100.0 -0.2 99.8 4000 100.0 -0.8 99.2 8000 Log Sum 100.0 -3.0 97.0
108.5
4. A-weighting correction (dB) . . . -26.2 5. Unprotected ear A-weighted . . . . 73.8 level (dB) line 1 - line 4 6. Average REAT attenuation at each frequency (dB) 7. REAT standard deviation at each frequency (dB) 8. Assumed Protection Value for 84% performance (dB) line 6 - line 7 9. Protected ear A-weighted level (dB) line 5 - line 8 10. Log sum of line 3 (dB) 11. Log sum of line 9 (dB) 12. NRR(SF) (dB) line 10 - line 11 - 5 dB 19.7*
-16.1 83.9
-8.6 91.4
-3.2 96.8
0 100.0
1.2 101.2
1.0 101.0
-1.1 98.9
19.7
19.7
20.7
21.2
28.3
36.6
37.8
8.7
8.7
9.7
9.4
8.4
7.4
9.1
10.4
11.0
11.0
10.0
11.3
12.8
20.9
27.5
27.4
62.8
72.9
81.4
85.5
87.2
80.3
73.5
71.5
90.7
* the data for 125 are used at 63 Hz without alteration beca use 63 H z is optional.
Method for Calculating and Using the Single Number Rating -SNR
The SNR is a single-number rating which is calculated in accordance with ISO 4869 2.2 (19 92), "Estimation of Effective A-weighted Sound Pressure Levels When Hearing Protectors Are Worn." It is very much like the NRR in terms o f its calculation method with a couple of notable exceptions. First, the SNR can be calculated for various levels of protection performance - that is, percentage of the population which can be considered to receive no less than the calculated attenuation - while the NRR is fixed at 98% . Seco nd, there is not an additiona l 3 dB subtraction as in the N RR. Various protection performances levels (%) are shown in Tab le 1 along with the appropriate multiplier ( ") which is applied to the standard deviation. Table 1. Value of " for v arious protection p erforma nces, x W hen a protection perform ance level is selected, the SN R is so designated by a subscript. Thus for a protection performance of 80%, the designation would be SNR 80. For the SNR to be similar to the NRR, its designation would be SNR 98 - 3 dB. W hen the protection performance is yet to b e selected, the designation is SNR x . The ISO 48 69 2.2 test procedure calls for the use of 16 subjects who fit the protectors under the supervision of the experimenter. Each subject provides one unoccluded and one occluded threshold. Protection Performance x in % Value of " 75 0.67 80 0.84 84 1.00 85 1.04 90 1.28 95 1.64
Calculating the SNR: The calculation of the SNR x is based on an assumed background o f pink noise with an overall level of 100 dBC and Assumed Protection Values, APV x , of the hearing protector (see Calculating APV ). The SNR x is independent of the actual noise spectrum to which it is applied and is calculated using the equation:
8000
where:
L Af is the A-weighted octave-band level at frequency f as shown in Table A.3.2, and APV fx is the Assumed Pro tection Value for each frequency. If data are not available for 63 H z, the summation begins at 125 Hz.
Note 2: 100 dB represents the total C-weighted sound pressure level of the reference pink noise in Table 2. The values o f L Af are derived in Table 2. They are simply the octave band levels of a 100 dB pink no ise corrected to reflect the A-weighting scale.
Method for Calculating and Using the Single Number Rating -SNR
Using the SNR: The SNR x is used to estimate the noise level under the hearing protector for a specific protection performance level (L'Ax ) given a specific C-weighted noise level (L C). The effective A-weighted sound pressure level, L'Ax , under a hearing protector is calculated by subtracting the SNR x from the C-weighted sound pressure level of a specific noise. For exam ple, if L C = 103 dB, and the SNR 84 = 21 .6 dB , L A84 is calculated using the equation: L'Ax = L C - SNR x , so that L'A84 = 103 dB - 21.6 dB = 81.4 dB. It can be stated that the effective A-weighted sound pressure level, L'A84, will be less than or equal to 81.4 dB in 84% of the situations when the hearing p rotector is prope rly worn by various use rs in this noise environm ent.
Table 2. A-weighted octave-band sound pressures levels, LA f , of a pink noise which has a C-weighted sound pressure level of 100 dB Octave-band center freq. f in Hz Assumed pink noise band level Total pink noise A-weighting correction factors LA f LA (total A-weighted sound level) -26.2 65.3 -16.1 75.4 -8.6 82.9 -3.2 88.3 0 91.5 +1.2 92.7 +1.0 92.5 63 91.5 125 91.5 250 91.5 500 91.5 1000 91.5 2000 91.5 4000 91.5 8000 91.5 100 -1.1 90.4 98.5 dBA Log Sum 100.0 dBC
An example SN R calculation is shown in Table 3. The A PV values used are taken as derived in Calculating APV for 1 standard deviation - a protection performance level of 84%.
Table 3. Calculation difference between APVf 8 4 and LA f . All values are in decibels. Octave-band center freq. f in Hz LA f , from Table A.3.2 Example APVf 8 4 LA f - APV8 4 63 65.3 4.1 61.2 125 75.4 6.4 69.0 250 82.9 10.8 72.1 500 88.3 15.0 73.3 1000 91.5 18.8 72.7 2000 92.7 22.4 70.3 4000 92.5 33.4 61.1 8000 90.4 28.9 61.5 78.4 dBA Log Sum
Method for Calculating and Using the High, Middle, Low (HML) Rating
The HML is a rating which is calculated in accordance with ISO 4869 2.3 (1992), "Estimation of Effective A-weighted Sound Pressure Levels When Hearing Protectors Are Worn." Using the H, M, and L ratings requires both C-weighted (L C) and A-weighted (L A) sound p ressure levels of the noise. T he H and M values are used in the calculation of the protected expo sure level for no ises which have prima ry energy in the middle and high frequencies and for which the LC and L A levels differ by 2 dB or less. The M and L values are used in the calculation of the protected exposure level for noises which have appreciable low-frequency components and for which the LC and LA levels differ by more than 2 dB. Calculating the HM L: The calculation of the H , M, and L values is based on eight reference noise spectra with different (L C - L A) values as shown in Table 1 and the Assumed Protection Values (APV fx - see Calculating APV ) of the hearing protector. The eight reference noise spectra were derived from the NIO SH 100 noises (Johnson and N ixon, 1 974 ). The index, i, is used to refer to a pa rticular no ise spectrum. The con stant, d i is an empirically-determined value. The HM L values are independent of the actual noise situation to which they are applied. The HM L values are ca lculated using the following eq uations: 4 4 H x = 0.25 PNRxi - 0.48 (di * PNRxi)
i=1
i=1
8 8 Mx = 0.25 PNRxi - 0.16 (di * PNRxi) i=5 i=5 8 8 Lx = 0.25 PNRxi + 0.23 (di * PNRxi) i=5 i=5 where: 8000 PNRxi = 100 dBA - 10 log 10 0.1(L
x f=63 No te 1. PNR x i refers to the predicted noise reduction for a pro tection performance leve l x and a reference noise spectrum i. L Afi and di are given in Table 1. If data are not available for 63 Hz, then summ ation b egins at 1 25 H z. The value 100 dB represents the total A-weighted sound pressure level of each of the noises in Table 1. The resulting H x , M x , and L x values are to be rounded to the nearest integer.
A fi
- APVf )
No te 2. To calculate the H, M, and L values for a particular protector, it is necessary to obtain the Assumed Protection Value for the desired protector performance for each frequency (see Calcu lating APV ). For e xample, if the sample protector from Calculating APV were used, the APV f84 would be as in Table 2. Note 3. After the APV fx are obtained, the A-weighted octave-band noise levels under the protector are calculated for each of the reference noises (L Afi - APV f84). For the pro tector in this exam ple, these levels would be as shown in Table 3. The eight PNR i,84 values a re then calculated by substituting the differences fro m T able 3 into its equation as shown in Table 4. The H 84, M 84, and L 84 values are calculated for the example using the appropriate equations, the PNR i,84 values from Tab le 4, and the co nstants d i from Table 1. The calculations are shown in Table 5 for Calculating APV and the values are rounded to the nearest integer.
Method for Calculating and Using the High, Middle, Low (HML) Rating
Table 1. A-w eighted o ctav e-ba nd so und pressure levels, L Afi, of eight reference noises normalized to an A-weighted soun d pressure level of 10 0 dB , L C - L A values a nd co nstants d i. All values a re in decibels. Index i 1 2 3 4 5 6 7 8 Octave-band Center Frequency ( f) 250 500 1000 2000 70.8 81.0 90.4 96.2 78.3 84.3 92.8 96.3 80.8 88.0 95.0 94.4 84.5 89.8 95.5 94.3 86.5 92.5 96.4 93.0 89.3 93.3 95.6 93.0 90.1 93.6 96.2 91.3 93.4 93.8 94.2 91.4 Difference (L C-L A) -1.2 -0.5 0.1 1.6 2.3 4.3 6.1 8.4 Constant di -1.20 -0.49 0.14 1.56 -2.98 -1.01 0.85 3.14
Table 2. Example APV f84 for the hearing protector in Calculating APV. Octave-band center freq. ( f) in Hz 63 Example APV f84 4.1 125 6.4 250 10.8 500 15.0 1000 18.8 2000 22.4 4000 21.4 8000 28.9
Table 3. Calculated difference spectrum for L Afi - APV f84 Octave-band center freq. f in Hz L Af1 L Af2 L Af3 L Af4 L Af5 L Af6 L Af7 L Af8 APV f84 APV f84 APV f84 APV f84 APV f84 APV f84 APV f84 APV f84 63 47.3 55.4 57.7 61.3 61.2 66.6 71.5 73.5 125 56.2 62.5 64.7 70.8 71.0 75.6 77.8 81.6 250 60.0 67.5 70.0 73.7 75.7 78.5 79.3 82.6 500 66.0 69.3 73.0 74.8 77.5 78.3 78.6 78.8 1000 71.6 74.0 76.2 76.7 77.6 76.8 77.4 75.4 2000 73.8 73.9 72.0 71.9 70.6 70.6 68.9 69.0 4000 73.3 72.6 72.7 71.1 69.0 68.7 66.5 66.5 8000 63.4 61.1 60.1 59.9 54.8 54.1 53.0 51.0
Table 4. PNR i,84 values for the sample hearing protector. PNR 1,84 = PNR 2,84 = PNR 3,84 = PNR 4,84 = PNR 5,84 = PNR 6,84 = PNR 7,84 = PNR 8,84 = 100 100 100 100 100 100 100 100 dBA dBA dBA dBA dBA dBA dBA dBA 10 10 10 10 10 10 10 10 log log log log log log log log (100.1*47.3 + (100.1*55.4 + (100.1*57.7 + (100.1*61.3 + (100.1*61.2 + (100.1*66.6 + (100.1*71.5 + (100.1*73.5 + . . . . . . . . .. .. .. .. .. .. .. .. + + + + + + + + 100.1*63.4) 100.1*61.1) 100.1*60.1) 100.1*59.9) 100.1*54.8) 100.1*54.1) 100.1*53.0) 100.1*51.0) = = = = = = = = 21.7 20.7 19.6 18.4 17.3 16.1 15.2 13.3 dB dB dB dB dB dB dB dB
Table 5. Calculated H 84, M 84, and L84 values H 84 = 0.25*(17.7 +...+14.4) - 0.48*(-1.20*17.7 +...+ 1.56*14.4) = 20.1 dB.
= dB. M 84 = 0.25*(13.2 +...+9.4) - 0.16*(-2.98*13.2 +...+ 3.14*9.4)15.5 L 84 = 0.25*(13.2 +...+9.4) + 0.23*(-2.98*13.2 + ...+ 3.14*9.4) = 15.5 dB.
Using the HM L: The values H 84, M 84, and L 84 may be used to estimate L'84 (total A-weighted noise level at the ear) for a particular protector in a specific noise situation. U sing the calculated values from Table 5, an example is shown below. 1. Calculate L C L A.
2. Calculate PNR x . The difference (LC - LA ) is calculated.
If this difference is # 2 dB, the predicted noise reduction level, PER, is calculated using the equation: Hx - Mx PNRx = Mx - ( )))))))) (LC - LA - 2 dB)).
4 If this difference is $ 2 dB, the PER is calculated using the equation: Mx - Lx PNRx = Mx - ( ))))))) (LC - LA - 2 dB)), 8
For example, if (LC - L A) = -1, then PER = 15.5 - ((20.1-15.5)/4 (-1-2)) = 17.3, using the HM L values from Table A.4.5. 3. Subtract PNR X from the Aweighted noise level.
The PNRx is then subtracted from the total A-weighted noise level to give the effective Aweighted level at the ear under the protector (LAx): LAx = LA - PNRx. For example if LA = 104 dB, then, LA84 = 104 dB - 17.3 dB = 86.7 dB.
Method for Calculating and Using the Assumed Protection Values (APVfx)
Assumed Protective Values (APV fx) are a m easure of hearing p rotector attenuation by frequency ( f). The APV fx is used in calculating other hearing-protector related values, such as the Single-Number Rating (SNR; see Calculating SNR ) and the High Medium Low rating (HML; see Calculating HML ). It may also be use d alone as a direct estimate of noise reduction at a particular frequency band when the octave-band levels of the noise are known. Calculating the APV fx: The APV fx is computed by subtracting a multiple of the standard deviation of the attenuations from the mean attenuation for each frequency band. The standard deviation multiplier ( ) is chosen based on the desired protection perfo rmance leve l ( x ). Table 1 gives the values for various protection performance levels. Once a protection performance level is selected, the APV is so designated by the subscripted x . For example, APV 500(84%) designates the AP V at the 500 H z octave ba nd for a pro tector performance leve l of 84%. Table 2 illustrates the calculation of the APV fx for a protector pe rformance level of 84% . Using the APV fx: The APV fx may be used as a direct estimate of the noise exposure level at a particular frequency band when the octave-band noise levels are known. T he AP V is app lied by subtracting it from the co rrespond ing octave-band noise level of the noise from which protection is being sought. For example, if the octave-band level for a noise is 92 dB SPL at 500 Hz and the APV 500(84%) is13 dB, then the protected level would be equal to the difference value of 79 dB SPL . The d ifferences may also be log summed acro ss frequency to provide a p rediction of overall effective noise exposure level under the hearing pro tector.
In the example in Table 2, the APVf84 values for a hearing protector are calculated; that is, the protection performance, ,
of 84% (1 standard deviation) is selected. These APVf84 values then can be used alone (as described above) or utilized in
calculations for the SNR (see Calculating SNR ) and HML (see Calculating HML ).
Table 2. Calculations of APVf84. All values a re in decibels. Octave-band center 63 125 250 500 1000 2000 4000
8000
7.4
10.0
14.4
19.6
22.8
28.6
38.8
34.1
3.3
3.6
3.6
4.6
4.0
6.2
7.4
5.2
3.3
3.6
6.4
3.6
10.8
4.6
15.0
4.0
18.8
6.2
22.4
7.4
31.4
5.2
28.9
10
References
American National Standards Institute. (1974) American National Standard for the Measurement of Real-Ear Hearing Protectors and Physical Attenuation of Earmuffs. ANSI S3.19-1974, American National Standards Institute, New York, N Y. American National Standard s Institute. (1997) M ethods for M easuring the real-ear attenuation of hearing pro tectors. AN SI S1 2.6-1 997 American National Standa rds Institute, New York, N Y. Environmental Protection Agency. (1979). 40 CFR Part 211 - Produc t noise labeling, Subpart B - Hearing protective devices. 44 Federal Register 56139-56147. International Standards Organization. (1992). Acoustics - Hearing protectors - Part 2: Estimation of effective Aweighted sound pressure levels when hearing protectors are worn. ISO/DIS 4869 2.2, International Organization for Standardization, Geneva, Switzerland. Johnson DL and Nixon CW (1974) Simplified methods for estimating hearing protector performance. Sound and Vib 8(6):20-27. Kroes P, Fleming R; Lempert B. (1975). List of personal hearing protectors and attenuation data, NIOSH Technical Report, HEW Publication No. (NIOSH) 76-120. Miller TW (1995). An analysis of octave band sound pressu re level data b ases an d various metrics based on the data bases studied. Unpublished masters thesis, North Carolina State University, Raleigh. NIOSH . (1998). Criteria for a Recommended Standard - Occupational Noise Exposure, Revised Criteria. National Institute for Occupational Safety and Health, Cincinnati, OH. DHHS(NIOSH ) Publication No. 98-126. Royster LH. (1995). In search of a meaningful measure of hearing protector effectiveness; Recomendations of the NHCA's task force on hearing protector effectiveness. Spectrum 12(2):1, 6-13. U.S. Department of Labor, Occupational Safety and Health Administration. (1983). Occupational Noise Exposure: Hearing Conservation Amendment; Final Rule. 48(46) Federal Register 9738-9785.
11