ML20154L454
| ML20154L454 | |
| Person / Time | |
|---|---|
| Site: | Seabrook  |
| Issue date: | 09/17/1988 |
| From: | Stusnick E PUBLIC SERVICE CO. OF NEW HAMPSHIRE, WYLE LABORATORIES |
| To: | |
| Shared Package | |
| ML20154K393 | List: |
| References | |
| OL-1, NUDOCS 8809260193 | |
| Download: ML20154L454 (44) | |
Text
_.
r September 17, 1988 UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION before the ATOMIC SAFETY AND LICENSING BOARD
)
In the Matter of )
)
PUBLIC SERVICE COMPANY OF ) Docket Nos. 50-443-OL-1 NEW HAMPSHIRE, et al. ) 5 0-4 4 4 - OL-1
) (On-Site Emergency (Seabrook Station, Units 1 and 2) ) Planning and Safety
) Issues)
. )
AFFIDAVIT OF ERIC STUSNICK I, Eric Stunnick, being on oath, depose and say as followas
- 1. I am the Manager of Arlington Operations of Wyle Research, a division of 'Ayle Laboratories, El Segundo, California. I have participatad in designing or evaluating the siren alert systems for approximately nine nuclear power plants including Seabrook Statio. A statement of my professional qualifications is attached hereto and marked "A".
2, This affidavit addresses the allegations in contention Basis A.1, which states in pertinent part:
"The VANS and the New Hampshire fixed airens because of their locations, height, acoustic range and number, do 8809260193 8809D -
hDR ADOCK 05000 V,3 f PDR
not provide tone . . . coverago for ecsentially 100 parcent of the population in the Massachusetts plume exposure pathway EPZ at the sound pressure levels required in NUREG-06S4 and FEMA-REP-10."
- 3. The objective of the Seabrook Station Public Alert and Notification System is to provide coverage to essentially 100 percent of th'a population within the Seabrook Station EPZ in accordance with 10 CFR 50, Appendix E and the guidance in FEMA-REP-10 and Appendix 3 of NUREG-0654/ FEMA-REP-1, Revision 1. FEMA-REP-10 presents, on page E-8, the acceptance criteria for those geographical areas covered by sirens positioned at fixed locations as:
"The expected siren sound pressure level generally excesds 70 dBC where the population exceeds 2,000 persons per square mile and 60 dBC in other inhabited areast or The expected siren sound pressure level generally exceeds the average measured summer daytime ambient sound pressure levels by 10dB (geographical areas with less than 2,000 persons per square mile)."
- 4. Public alerting within the Massachusetts portion or' the Seabrook Station plume exposure pathway EPZ (hereinafter referred to as the "Massachusetts EPZ") will be accomplishad through the activation of the VANS sirens positioned throughout the Massachusetts EPZ. The VANS acoustic coverage nests or exceeds the regulatory guidelines quoted in pdragraph 3 above. Figure 2-2 of the Seabrook etation FEMA-REP-10 Design Raport depicts the alert systsm coverage for the Massachusetts EPZ (copy of Figurc 2-2 attached and marked "B"}.
L
l l
l l 5. As discussed in the FEMA-REP-10 Design Report for Seabrook Station in Section E.6.2.1.d (copy attached and marked "C") the sound level coverage for each siren in the alerting system was determined by means of a computer model developed by Wyle Laboratories. Figure 2-2 of the FEMA-REP-10 Design Report depicts 60 dBC and 70 dBC sound level contours calculated by the model and then graphically combined into envelopes depicting the total system coverage.
- 6. Appendix B to the FEMA-REP-10 Design Report contains a Wyle Research Report, WR 88-9, "Siren Ranging Model" dated April, 1988 (copy of WR 88-9 attached and marked "D") which describes the computerized siren ranging model for use in designing public alert and notification systens for nuclear power plants.
- 7. As provided in Wyle's report WR 88-9 in Section 4.1 and as documented in Wyle Research Test Report WR 88-4 (copy of tect report attached and marked "E"), the acoustic output of the dual siren model employed in VANS was determined by direct measurement (i.e., field measurement as permitted by FEMA-REP-10, paragraph E.6.2.1, page E-8, copy attached and marked "F"] of the C-weighted sound level at a distance of 100 feet on the siren's axis. The acoustic output of a dual siren was measured to be 134 dBC at tha tone frequency of 550 Hz.
- 8. The siren input parameters for the computer model calculati6ns discussed in paragraph 5 above reflected the h
dual measured output of 134 dBC at 100 feet from the siren and siren activation at a height of 45 feet above ground level. Since there is a possibility that some VANS sirens may be activated at a height of 25 feet, during the process of being elevated te maximum height, the computer model was also used to calculate the sound level coverage for each VANS siren location at that lower height. An analysis of the calculated 70 dBC and 60 dBC contours for both activation heights for each VANS location indicates that, on the average, the sound levels at the predicted contours would vary by less than one dB for activation at the lower height and would return to the full predicted level within one minute as the siren was raised to full height.
- 9. Although a height of 45 feet for siren activation was used as the basic for the sound coverage analysis, the actual siren height achioved by the VANS vehicle is approximately 51 fact. This add 8_tional six feet of siren hoight will result in greater coverage than that calculated for a siren height of 45 feet because the sound will encountar lowei barriers along the projected path.
- 10. I nave also performed calculations to estimate the serisitivity of predicted sound level coverages to changes in acot.stic location. My calculations indicate that generally, if a VANS vehicle is parked within 400 feet of the assigned acoustic location the calculated 60 dBC contour will vary by j less than one db.
- 11. From a review of the information provided on Figure 2-2 of the FEMA-REP-10 Design Report for Seabrook Station it can be seen that all the geographical areas within the Massachusetts EPZ where the population density exceeds 2,000 persons per square mile will be subjected to a sound level of at least 70 dBC. With the exception of four small areas discussed separately below, the remaining area of the Massachusetts EPZ is covered by a sound level of at least 60 dBC.
- 12. As depicted on Figure 2-2 of the Design Report, four small geographical areas in the Massachusetts EPZ are not subjected to at least 60 dBC of siren coverage. An ambient noise survey was conducted in each of these four areas to determine the ambient background noise levels for an average summer day. This survey is described in Wyle Test Report TR 88-11 (copy attached and marked "G").
- 13. The four areas are briefly described below.
Parker River National Wild _ life Refuae. Newbury This area of approximately 350 feet length along Plum Island Road at the Newbury/Rowley corporate boundary is located approximately 9.8 miles south-southwest of Seabrook Station and compriaes an area of approximately 0.08 square miles.
South Face of Crane Neck Hill in West Newbury This area is located approximately 11 miles southwest of Seabrook Station, comprises an area of approximately 0.32
square miles, and is bounded on the north by Crane Neck Hill, on the south by the EPZ boundary, on the east by a dirt road, and on the west by Georgetown Road.
West Newbury, West of Route 113 and South of Pleasant Street This area, located approxinately 11.2 miles southwest of ,
Seabrook Station, comprises an area of approximately 0.10 square miles. The area lies to the south of Pleasant Street in West Newbury, on the west side of a hill north of Pentucket Regional Junior High School, and extends to the Merrimack River. A small, noncontiguous area approximately 600 feet southeast of Route 113 is also part of this area.
Egrish Road. Newbury This area is a small triangle to the east of Interstate 9S located approximately 11 miles from Seabrook Station, and comprises an area of approximately 0.02 square miles. The area is bounded on the west by Larkin Street and on the south by the EPZ boundary.
- 14. The ambient noise surveys, described in Wyle Test Report TR 88-11, were conducted on July 17 through July 23, 1988. The purpose of the surveys was to determine the average summer daytime (7:00 a.m. to 10:00 p.m.) amnient background noise level for each area in the 500 Hz one-third octave band containing the dominant siren frequency of 550 Hz. This is in accordance with the ASLB Memorandum and order (Denying Massachusetts' Motion of March 3, 1987), page 15, I March 25, 1987. The noise level exceeded 50 percent of the time, i.e., L50, was used to represent the average ambient background noise level. The L50 level is more conservative (i.e., results in higher ambient sound levels) than the commonly used L90 level, i.e., the noise level exceeded 90 percent of the time.
- 15. For each of these four areas, a measurement site was selected where the highest ambient noise level in each area was expected, referred to as the primary site.
Additional measurement sites were chosen in each area to provide an estimate of the spatial variation of ambient levels in the area, referred to as secondary sites.
- 16. Data were recorded at each primary site during the entire daytime period (i.e., 7:00 a.m. to 10:00 p.m.).
Sample measurements over a short period of time were takon at the secondary sites. At the primary sites both continuous A-weighted L50 measurements and measurements of the 500 Hz one-third octave band were obtained. At the secondary sites, A-weighted measurements were obtained. A correction factor was developed (i.e., A-weighted to 500 Hz one-third octave band) based on the two sets of measurements at the primary site.
This correction was then applied to the A-weighted measurements at the secondary sites to obtain 500 Hz one-third octave band values.
J
- 17. As provided in paragraph 3 above, the average measured summer daytime ambient sound pressure levels should oe compared with the "expected siren sound pressure level."
The expected siren sound pressure levels for these areas were determined by means of the computer analysis (discussed in paragraphs 5 - 8 above) for each siren location which could produce sound levels in these areas. The lowest sound level from the siren predominantly influencing the area was used as the expected siren level for that area.
- 18. The following tabulates the highest average daytime ambient noise level recorded from either the primary or secondary sites, the lowest e qected siren level and the resultant difference for each area.
Average Ambient (L50) Level in Expected 500Hz One Third Siren Area Oftave Band. dB Level. dBC Difference. d3 i
Parker River 39 58 19 Crane Neck Hill 32 56 24 Vest Newbury 39 56 17 Parish Road 43 59 16
- 19. B6 sed on the results of this survey, the expected siren sound level for each area will be greater than 10 dB above the average ambient background level.
- 20. In summation, based on the foregoing, the following factual conclusions can be reached regarding the siren system within thase portions of the EPZ in Massachusetts:
(a) Those geographical areas within the Massachusetts EPZ where the population density exceeds 2,000
n persons per square mile will be subjected to sound levels of at least 70 dBC; *
(b) Except for four small areas, the remaining area is covered by sound levels of at least 60 dBC; (c) These four small areas, whose population density 19 less than 2,000 persons per square mile, will be subjected to sound levels which excied the average ,
measured summer daytime ambient sound levels by at least 10 dBC; (d) The sound coverage provided by the VANS sirens meets or exceeds the sound coverage guidelines provided in FEMA-REP-10, Appendix 3.
i
..~. -_ . .-. _ - _ . _ _ _ - _ _ _ _ . . . _ _
i f
a Eric Stuanick COMMONWEALTH OF VIRGINIA Arlington County, ss. September d , 1988 The above-subscribed Eric Stusnick appeared before me i
and made oath that he had read the foregoing affidavit and that the statements set forth therein are true to the best of his knowledge.
Before me,
- 71 .
Notary Public My Commission Expires:
, IfrDemelmhm Eqdm Nmmeeris, test Stusnick Attachment A, 1 of 1 ERIC STUSN!CK POSITION: Manager, Arlington Operations
?OINED WYLE: 1977 PRINCIPAL DUTIES AND RESPONSIBILITIES:
Management of research and consulting staff in Arlington office. Management of experimental and theoretical programs in transportation and environmental noise, and underwater acoustics.
BACKGROUND:
Wyle Laboratories, Arlington, VA - Program management and applied physics support for studies in acoustic signal analysis and sound and vibration measurement and control. Project managw for the design of soundproofing modifications for home' in the vicinity of commercial airports; the development of statistical energ analysts software for estimating sound and vibration levels in the space stations an experimental study of the detectability in locomotive esbs of railroad track torpedo detonations; and the measurement and control of low-level floor vibrations in a semiconductor manufacturing facility. Principalinvestigator for the develop-ment of computerized underwater acoustic intensity measurement systems; an analytic study of short-range acoustic propagation in a turbulent atmosphere near the ground; and the design of emergency community alerting systems for nuclear power generating facilities. Principal author of a handbook for the measurement, analysis, and abatement of railroad noise. .
Calspan Corporation, Buff alo, NY (5 years) - Senior Physicist. Provided physics support and program management in signal analysis, acoustics, and ballistic missile defense.
EDUCATION:
B.S., Physics, Carnegie Institute of Technology,1940.
M.S., Physics, New York University,1962.
Ph.D., Physics, State University of New York at Buff alo,1971.
CERTIFICATION:
Professional Engineer, State of New York,1976 Membw, institute of Noise Control Engineering PROFESSIONAL MEMBERSHIP 5:
Acoustical Society of America American Physical Society American Association for the Advancement of Science The Society of the $lgma XI PUBLICATIONS:
Forty-eight technical reports or pub!1 cations.
M-e
Stunnick Attachmant C, 1 of E.6.2.1.d Siren Range Calculations The sound level coverage (tone) for each siren in the alerting system was determined utilising a computer model developed by Wyle Laboratories. This codel determines the range of specified s tren signal levels based on atten-ustions along the siren signal path. Field measurements have been made and the measured siren sound levels have been compared with those predicted by the model. This comparison illustrates that the predicted levels are con-servative and are, thus, appropriate for the systes design.
The 60 dBC and 70 d5C stren tone coverage for the siren alerting system is shown on Figures 2-1 and 2-2 as sound level contours. To develop these contours. the model calculates the contours for each siren. The 60 dBC and 70 dBC contours for all sirens are then graphically combined into envelopes depicting the total systes coverage.
The range for voice alerting messages broadcast by the sirens was based on speech intelligibility tests on the sirens employed in the system. This intelligibility test data was then used in conjunction with the sound propagation model to prsdict the voice alerting range for each siren.
Appendix 5 contains Wyle Research Report 88-9 which presents the siren ranging calculation proceduree utill ed in the system design.
E.6.2.1.e Maintenance of Siren Systes A regularly scheduled, preventive maintenance program will be initiated for the sirens and VANS vehicles in the system. Maintenance will also be perforted if any of the regularly scheduled tests (see Section E.6.2.1.f) indicate malf unctions. In addition. repairs will be made if it is known that something has happened to disable one of the sirene or VANS vehicles
( v a nd a li se , itghtning s trikes, accidents, etc. ).
2-20
__ _ . _ - . .. - - - - _ .. -- _ - _ - . . - - - - - - = -.
I Stusnich Attachmc D, 1 of 30 i
1 i
) d WYLE RESEARCH REPORT TR 88-9 O SIREN RANGING MODEL "
9
\ 1 m
WYLE RESEARCH C 9 l
)
i l
)
)
4 f
l 3Q.~i.f3 L A80R ATORIE S ca;,s
{
i l
Stusnick Attachment D, 2 of 30 M_ - - .. ,. ~
WYLE RESEARCH REPORT TR 88-9 SIREN RANGING MODEL ,
(
Prepared Fx:
l PUBLIC SERVICE OF NEW HAMPSHIRE _
U.S. Route 1. Laf ayette Road Seabrook, New Hampshire 03874 "
=
Purchase Order No. 38619-03 l
Prepared By:
i WYLE RESEARCH q ,
2001 Jefferson Davis Highway -
Ar!!ngton, Virginia 22202 l
April 1988 l
Stusnick Attcchment D, 3 of 30 TABLE OF CONTENT 5 P*At
1.0 INTRODUCTION
I 2.0 MATHEMATICAL MODELS ..................... 2 2.1 Spherical Spreading . . . . . . . . . . . . . . . . . . . . . . .
3 2.2 A ir A bsor p tion . . . . . . . . . . . . . . . . . . . . . . . . . 3 2.3 Scattering Attenuation ..................... 4 2.4 Excess Ground Attenuation ................... 4 2.4.1 R ural / Suburban A re as . . . . . . . . . . . . . . . . . . . 6 2.4.2 U r ban A reas . . . . . . . . . . . . . . . . . . . . . . . 6 2.4.3 Heavily Forested Areas . . . . . . . . . . . . . . . . . . 8 2.4.4 W a ter A reas . . . . . . . . . . . . . . . . . . .. . . . . 8 2.3 Barrier A ttenuation . . . . . . . . . . . . . . . . . . . . . . . 10 2.6 Attenuation Due to Temperature and Wind Gradients ... . .. .. 12 3.0 COMPUTER IMPLEMENTATION ................... 13 4.0 IN P UT P AR AMETER5 . . . . . . . . . . . . . . . . . . . . . . . . 18 4.1 S tr en P ar a m e t er s . . . . . . . . . . . . . . . . . . . . . . . . 18 4.2 M eteorological Conditions . . . . . . . . . . . . . . . . . . . . 19 1
3.0 VALd ATION OF MODEL . . . . . . . . . . . . . . . . . . . . . . 23 REFERENCE 5 .. . . .. . . . . . . . . . .. .. . . . . .. . ... . R1 i
N I
l l
Stusnick Attachmsnt D, 4 of 30 l l
LIST OF FIGURES Fig.
No. Page 2-1 Ettimated Excess Attenuation Due to Scattering Near Earth's Surface . . . 5 2-2 Empirical Estimates of Excess (Ground) Attenuation in Rural / Suburban Areas (Source Above Roof tops) and the Additional Excess Attenuation Due to Shielding in Urban /High Rise Areas (Source Below Roof tops) . . . . 7 2-3 Excess (Ground) Attenuation in Heavily Forested Areas . . . . . . . . . . 9 2-4 Barrier Attenuation as a Function of Fresnel Number . . . . . . . . . . . 11 2-3 Effects of Temperature and Wind Gradients on Sound Propagation . . . . . 13 3-1 Comparison of Siren Sound Level Contours With and Without the Ef f ects of Barrier Attenuation . . . . . . . . . . . . . . . . . . . 17 bl Eff ect of Wind on Sound Level Contour . . . . . . . . . . . . . . . . . 21 3-1 Comparison of Measured and Predicted 5iren Sourd Levels for Indian Point Data .......................... 24 3-2 Comparison of Measured and Predicted Siren Sound Levels for Seabrook Data ........................... 25 Table LIST Off TAB'.E3 No.
4-1 Acoustic output of Sirens Used in Seabrook System ... . ... . . . . 19 b2 Averar,e Seasonal Values of Relative Humidity and Temperature in Seabrook Area .......................... 20
Stusnick Attachmant D, 5 of 30 l
1.0 INTRODUCTION
This report describes the computerized siren ranging model which has been developed by Wyle Laboratories for use in designing public alert and notification systems for nuclear power plants.
The model is capable of taking into account acoustic energy losses due to spherical spreading of the waveicont, air absorption, scattering by turbulence, excess ground attenuation, barrier attenuation, and wind and temperature effects. In order to account for barrier attenuation, the model relles on digital ground elevation data obtained from the National Cartographic Information Center of the United States Department of the Interior.
This data is processed on a mainframe computer system to provide estimates of siren sound level as a function of distance along a sequence of equally spaced radials originating at each siren. The resultant sound levels are transferred to,a microcomputer on which 60 and 70 dB contours are computed, scaled, and plotted. The plots fet each siren in the system are then transferred to a base map.
The model is exercised in an iterative f ashion with the location, height, power level, and frequency of each siren being continually adjusted until acceptable coverage is obtained in all portions of the Emergency Planning Zone.
As will be seen in the discussion to follow, the design of the model and the choice of values for input parameters are such that the siren sound levels are generally under-predicted. This was done by conscious decision since some of the algorithms used in the model are, of necessity, engineering approximations. By using reasonable conservatism in the choice of algorithms and input parameters, a buffer is automati: ally built into the model to correct for any adverse effects of such approw. Nations.
This report consists of five sections. Section 1, this Introduction, summarizes the ,
report, Section 2 describes the mathematical algorithms used to model the various mechanisms by which sound level decreases with distance from the siren. Section 3 out11oes the computer implementation of these algorithms. Section 4 describes the choice of input parameters for the Seabrook system. Section 5 presents a validation of the model based on siren sound level measurements made at Seabrook and other nuclear power plants.
i WY.lf.
Stusnick Attachmont D, 6 of 30 2.0 MATHEMAT1 CAL MODEL5 This section describes the series of mathematical mo 21s which have been developed to prd!ct the attenuation that occurs as sound propagates '.om a siren to a receiver.
The sound level, L(R), at a distance R fron 'he siren can be expressed ast L(R) = L, + Aspread
- Aabs
- Ascatt
- Agrnd
- Abarr
- Atemp
- Awind I2*I where L is the source sound pressure level on the siren's axis at a predefined n
reference distance of 100 f eet, A spread is the attenuation that occurs due to the spherical spreading of the sound, A abs is the attem. ' ion that occurs due to absorption of acoustic energy by the air, ,
A scatt is the attenuation that occurs due to scattering of acoustic energy out of the directional beam of a rctating siren by atmospheric turbulence, A is the attenuation that occurs due to absorption of acoustic energy at grnd the ground surf ace as the sound wave propagates in a nearly hori-zontal path, A
barr is the attenuation that occurs as a remit of the reflection and diffraction of acoustic energy by barriers formed by hills between the source and receiver, A is the attenuation (or ampilfication) that occurs as a result of temp refraction by the ternperature gradient that exists near the surf ace of the Ground, and A,g is the attenuation (or amplification) that occurs as a result of refraction by the wind speed gradient that extsts near the surface of the ground with a wind present.
in general, each of the attenuation terms is a negative numb *r so that the siren sound level diminishes as the distance from the siren increases. In certain cases, however, some of the it'.enuation terms can have positive values, indicating sound ampilfication.
M,-
2
Stusnick Attachmont D, 7 of 30 The following section' briefly describe the models that were used to estimate the value of each of the attenuation terms in Equation (2-1).
2.1 Scherical Screadina At any given distance, R, from a point sound source close to the ground, the total acoustic power output of the source is spread ory a hemispherical surf ace having an area proportional to R .2 Thus the sound energy per unit area reaching the receiver decreases with distance at a rate proportional to 1/R 2. This so-called spherical spreading effect causes an attenuation between one distance R, and a recond distance R of:
A spread
= -10 log 10 (R /R ) (2-2)
It is this effect that produces the well-known 6 dB per doubling of distance attenuation as one travels away from a point source.
2.2 Air Absorption ,
in a still, uniform atmosphere, sound waves lose energy as they pass through the air due to minute heating and viscous effects (classical losses) and due to molecular energy exchange processes (molecular losses) which are influenced by the amount of moisture in ;
the air. An ANSI standard,I well supported by experimental data.2 is available which maka. predictions of this (crm of sound attenuation relatively straightforward. The loss is defined in terms of a frequency- and weather-dependent atmospheric absorption coefficient in e per 1,000 feet. The model requires that temperature, relative humidity, i and utmospheric pressure be defined.I (Atmospheric pressure has only a minor effect on <
atmrpheric absorption, and a standard sea level atmospheric pressure is generally assumed without any loss in accuracy.)
For example, based on the extrema of the seasonal average values of the ;
temperaturs and relative humidity !w the Seabrook area, as discussed in Section 4.2, below, a 0.99 e/1,000 f t air absorption coefficient results for a single-tone siren f frequency of $30 Hz. ;
i l
t 3
M.m.
m t
Stusnick Attachmant D, 8 of 30 l
l 2.3 Scatterina Attenuation Scattering of sound waves occurs in turbulent air and can result in an additional
' propagation loss for a very directional source such as a rotating siren.3 Measurements of scattering attenuation in horizontal propagation are severely limited. The available data have been obtained under uncontrolled experimental conditions making it difficult to ,
separate out any scattering attenuation from other effects.3 Furthermore, no evalcation of the effects of directivity of the source appears to have been considered. There is, however, one unique set of data which has provided direct and rather convincing evidence of scattering attenuation of low-f requency sound over a long horizontal path."
l The data were obtained from measurements made over a path length of 30,000 feet of the directional sound field radiated by two nominally identical rocket test stands located back to back with the exhaust and resultant directional sound fields from each stand oriented 180 degrees apart. The propagation loss ovee the same path from the two separate rocket engine sound sources, fired one right af ter the other, 'was not the same for the two tests. The diffef ence in propagation loss was an apparent additional excess attenuation for the source whose primary directional sound field was oriented along the measurement ratn. This phenomena could only be explained as scattering attenuation, and an analysis of the data produced the estimates of scattering loss coefficients as a function of frequency shown in Figure 2-1 by the circle data points. Estimates of scattering attenuation are also plotted for comparison and show at least an order of
.nagnhude agreement with the Indicated data points.U For example, for a 330 Hz tone, an additional propagation loss of 0.20 dB/1,000 f t must be included for directional sirens to account for scattering attenuatior, of the directional beam.
1 2.4 Excess Ground Attermaation When sound waves travel from a source to a receiver over a ground surface, two different ray paths are possible - the first directly from the source to the receiver, the second by reflection from the ground surf ace. These two waves interact to produce either l
i attenuation or amplification. The exact nature of the interaction is a complex function of the source and receiver heights, the source-to-receiver distance, and the impedance properties of the ground suriace.
M,
Stusnick Attcchm3nt D, 9 of 30 l
10 i i * .i I i i ii . , , ,-
- Delony Esiimot.6 Brown & Clliford g - Estimes3 _
- 1 *0 ,, '
h . -
o b ..
e o 9 -
5 t v', == =. . f' ~ ~. , ;
e
.e
- ~
.. / .. is ' ']L ,% .
4
.. ,f> "
/
,{ o. l ._ , ~
~ ..
o/ / o a
~ s ..
t > /
k ~
/ ** " ]o' Meesuremenn EstimetedonFrom -
- l e'
I f = = $ tven fest:3 ..
I ~ 2(f /, M)I' I/3
-~e=3 dulooo fr-
.I + (f / M P .
i i i ii i e i i! , , , ,
0.01- too 1000 i 10
- Frequency, Ha Figure 2-1. Estimated Excess Attenuation Due to Scattering Near Eerth's Suriace.
i l
I l
b --
Stusnick Attachment D, 10 of 30 In addition, when considering long-range propagation of sound, shieleiD5 and scattering by small buildings and other small surf ace irregularities can be considered as an sdditional distance-dependent attenuation f actor.
Because the exa:t nature of ground surface properties and irregularities cuanot be determined a priori, estimates of such ground attenuation terms can only be made by modeling phenomenological data. Four general types of ground cover are included in the models rural / suburban, urban, heavily forested, and water.
l 2.4.1 Rural /$uburban Areas Figura 2-2 shows excess ground attenuation as a function of 61stence for several naral and suburban areas. Thk Ilgure La based on published data from which spherical spreading and air attenuation factors have been removed.' The best-fit design curve to this data, as shown in Figure 2-2, has the form: ,
A " " I 3 I'810 (R/100) dB , R < !700 f t ynd (2-3)
- n. 16 dB ,
R 2 1700 f t 2.4.2 Urban Areas i
b to shielding ~oy buildings, an addit onal excess attenuation, over and above that ,
def!ned above, must be included when predicting siren range la urban araas where the stren is mounted below re.of tops. Sound propagation data for this condition is quite meager, but Reference 6 has provided a reasonable summary of the limited information which c:.. be used to predict total excess attenuation for such areas.
1 1 The average additionalincrement in excess attenurtion over what is necessary for '
rural and suburban areas (see Figure 23 in Reference 6) is used here as th4 basis for predicting thh added excess attenuat%. (The amount of this additional excess attenuation due to shleiding by buildings is also roughly consistent with mors recent studies on sound propagation in urban areas.7) The resulting litte for the total excess attenuation in such areas is shown in Figure 2 2 by the light das%f line. A siniptification of this trend is in the model so as to remove anomalous peaks in the predicted values of excess attenuation between 300 and 3,300 feet. Ths resulting design curve, shown by the 4
VP.U.
- m
Stusnic9 AttacNnant D, 11 of 30 ;
I I
I l
.lo ,
O Rural Areas ,
Tyle Design Curve. Rursl/$uburban O $4 urban Areas
- - Wyle Design Curve. Urban /High Rise 0,
y _ .
N .
13 log (R/100) i 10 -
ww / 0 14iG o
.M log (R/ID0)
{ \ 12 m e
, ,i v
e o
=o "b - ~
__ 8 _
w -~
< *20 " N,m -
% ALitional Enceu Atienuation De Es.imate of additional ence:s % N 17 d6 to %id3D by Buildings in UtW/
attenuation for Urben/High Rise 's N l Hi6h Rise reas
- g g Arou based vi applying Delaney's 30 ,
regutt: to Upper Curve. ..,% t""*
N. ""* O
. .. "*"**~~~'"~~"""-
t )
100 1000 10,000 20,000 R, Olstence Fro,n $1ren in Feet Figure 2 2. Emp3rical Estimates of Excess (GrourvA Attenuation in Rural / Suburban Areas (Source Above Rooftope) end the Addidonal Excess Attenuation Due to 5hielding in Urban /High RLee Areas (Socree Below Rooftope).
(Based on Data From Delaney, Reference 6.)
l P
4
.. .n . ~ . --- ,. , , , . - - - - - - - --
y
n I
Stusnich Attachmont p, 12 of 30 l
l heavy dashed line in Figure 2-2, is exactly 12 dB greatstr than the upper curve for rural.
l suburban areas at distances greater than 300 feet, decreasing to an added increment of zero at ino feet, in summary, for sirens in urban areas which we mounted below roof tops, the total attenuation in excess of spherical spreading and u.c absorption is deflned by:
> -30 log (R/100) , dB R < 300 tt A
grnd = < 13 log f.R/100) - i2 , d6 300 s R s 1,700 f t (2 4)
, -28 dB , R > 1,700 f t 4
2.4.3 Heavily Foreited Areas
- Te investigate the magnitude of sound attunuattor. through hemvily forested ieas, a series of wand level measurernents have been made, the results of which are plotted in Figure 2-3. Spherical spreading and air al.oorption effets have been rer'noved so that the attenuations repreten; oNy the ground effect. The measurement sites included in the figure are in relatively flat areas, w that no barrier effects
- se present.
Although the amount of data at distances greater than 1,700 feet is sparse, there is indication that the 16 e attenuat!on cutoff that appears in Equation (2 3) does not occur.
] Thus, for heavily f arested areas, tM ground attenatie used in the modells:
J A " (2-3) gn d
- I3 I'810 (R/100) dB Other data show the nme it. crease in ground attenuation with distance up to about l 1,300 feet, at wt ich point tne spread in the data becoves so great that no further
- dependence on distance can be reasonably inferred. However n uden loca! data are 1
avalhble, Wy!) has adopted the more conservative attenuation figures as shown in Figure 2 3 for '*heav3y foretted' a eas (defined by Equation '2-3)) as repeesenting the worst case fot : Iron ranging studies, i
2.4.4 Wst," Arg j for propaganten entirely or mostly ovce water, 'here is little or nc, excen ground j attenuation, thur Agrnd = 6 O (2 4
'M t
m - i
Stt.tsnick Attcchm nt D, 13 of 30
+10 i .
e i . . . , , I I
- I I I g ,y o stren g o Siren 9 N Md '
' "Heevtly Forest,da Al,,,;,g 0 "" Iturel/suburber, Algorit%
l o 8
O
.h O O 10 - 8 j o -
o 0 o O c 0
O 1 === . . .
' 20 0,
= 80o .i 4
[ o o .
g j
~
l 4, ' .n - i n - i n i i 1000 1 , , , , , ,,
t
., 10,000 i A, dis m lm SluinFw i r
4 Figwe 2-3. Excess (Ground) Attenuation in H6 ..dy Fe ested Areas, i
i l
4
'N b
l l I S
M.-.
+na- -e n--- -
Stusnick Attcchmont D, 14 of 30 I
l 2.5 Earler Attenuatloit Reflection and diffraction of sound by barriers must be considereci when siting sirens in hilly areas to account for the ba:rier attenuation effects of tl.e hills. Well developed design methods are avAlable for predicting attenuation by thin barriers which esseMially
! I ignore ground reflection effects or which include grei.md rcilectir 1 effects c.) barrier attenuation.9 Since the sirens in the Seabrook area are installed at least l ground and since the treatment of thJ hills as thN barriers is an approximation, the refinement of includir's ground reflection effects on barrie attanuation is not warranted.
Figure 2-4 defines the sound attenuation provided by f, thin barrier based on the prediction model deflhed in Reference 8. This classical Fresnel diffractior, model is well supported by experimental data mes.sured under ideal ccaditions. Note that, although the fvem of the function in this figure Isolid line) is a straight line, the horizontal scale is non-linen, to reflect the fact thai barrier attenuation is a non-1Inen funct!on of ,
Fresnel number.
The barrier attenuation mo/el uwd in this study, shown as a dasheu line in Figure 2-L, employs a least-squares fit to the solid cu?ve in the figure arW has the form:
A u -10 log!O (20 N) bur , 12.6 > N > !.0 A = 8
- 2 AN - 8 bur ,
p A = 0.0377f , 0.3 s N s 1.0 8 = -0.02700 (2 7)
Ag = C YB' + 2 ACN -B g _
a A = 0.02461 8 = 0.00099 , - 0.3 *' N < 0.3
- C = *! ,
N30
= -l , N<0 A = 0 , N g -0.3 barr to W
- Y. .L E..
Stusnick Attechmsnt D, 15 of 30 30 i i .
ii g3 y a . . i i iig i i-N'= (A + 8 - 0) 23 -
F = frequency Cveoff -
. = sp..d of sound A \
Sour / I
" - ~ *
u ..I,., -
Ber,l.,
%
- N le positive if the ure ef. set f
- gp l,T O ,",'*e,';.,;. = a w. 6. , u. ,. . . n , .*
---t..s, Fi, j ,
10 -
s -
s ..,. ,A.. i %i.., -
r,oimi.n. i.,,i.,
of e Hill
/
al lli l e i l e i en i i t il i i I
-0.1 -0.1 -0410 no1 0.1 0.10.5 14 2.0 44 4D 84 to 20 30 40 50 F,.enel Non6.,, N Figure 2 4 Barrier Attenuation as a Function of Fresnel Number.
11 M,-
l Stusnick Attachmant D, 16 of 30 I
The n.odel represented by this set of equations does not include reduction of ground attenuation due to th$ pretence of the barrier or diffraction over the top of the barrier caused by foliage. The effect of these phenomena is generally approximated by imposing a cutoff on the barrier attenustion term. Highway noise barrier design guidelines usually suggest a cutoff of -12 to -13 dB, based on field measurements fro n previously constructed highway noise barriers. These measurements do not, however, in:lude data :
from extremely high barriers, such as hills or mountains, as is found in the case of siren I l
sound propagation. It is thus more conservative to ese a cutoff of -2k d8, which is l
suggested by Beranek for a thin barrier.10 To apply this model to hilly terrain, computer software is used to replace the actual ground elevations between the source and receiver with an equivalent thin barrier (see lower inset in Figure 2 4). A sequence of barrier attenuations is computed for all such equivalent barriers located at regular intervals between the source and the receiver.
The maximum value of this sequence is taken as the attenuation of the terrain.
The elevation data required for thh computation is obtained from planar standard digital terrain data tapes, available from the National Cartographic Information Center, 0.5. Geological Survey, Department of the Interior. These data were produced by the Defense Mapping Agency Topographic Center from 1:230,000-scale terrain r.ontour maps of the United States and provide a grid of terrain elevation values at 200-foot intervals.
2.6 Attenuation Due to Temperstwo and Wind Gradients Change over time in the structure of the vertical temperature and wind profile in the atmosphere produce temporal variations in sound propagation losses. As illustrated in Figure 2 3(a), a negative temperature or wind speed gradient (decreasing with height) causes sound emanating from a source near the ground to bend upwards, resulting in an increase in propagation loss and creation of acoustical shadow zones. On the other hand, as shown in Figure 2-3(b), a positive temperature or wind speed gradient causes sound to bend downward towards t5e ground. in some cases, with a combined negative and then positive gradient, sound is focused back to the ground at points distant from a receiver resulting in substantial increases in sound level beyond that normally experienced. For vertical gradients in wind velocity, as shown in Figure 2-3(c), a complex shadow zone forms around the source in a pattern dictated by the mean wind vector.it,12,13 Although -
wind speed also has an influence on refraction of sound by wind gradients, it is not as important as wind direction.I3'I" M
-'~
i2
Stusnick Attachmont D, 17 og jo a-
.. ~ %
[7A A Nk=%E (e) Roy Peik in Air Wn Vertical Wind (b) Rey Peths in Air Men Verticel Wind Velocity (la the Direction of Soved Velocity (In the Ofrectlen of Sound '
Propogetion) er Ternperotwee Gredient Propopetion) er Temperature Grodient is Nogettve. Is Positive.
I l l $hedewl t Zone -
l 1 m .ei e, j I
I
/ ! winf i sac.
t (c) Wind Genereted $bedw Zare.
Figure 2-3. Effects of Temperature and Wind Gradients on Sound Propagation.
!3 U...
Stusnick Attcchmont D, 18 of 30 Since negative temperature gradients as shown in Figure 2-3(a) are more common than positive temperature 5~8dlPnts (temperature inverslor.'s). grcuad attenuation data generally contain the eff acte si negatin temperature gradients. Since a temperature inversion will only tend to reduce the attenuation (i.e., increase the sound level at a given point), a conservative estimate of the attenuation due to temperature gradients is already included in the A ,,na ,? W(l. Thus the model assigns a zero value to Atemp' This to61 c cannot be applied to wind speed gradients since the data on which Agrnd is based was normally then dwing very low wind conditions. However, a conservative estimate of this offeu is that a 3 d8 increase in sound level downwind of the source and a 3 d5 decrease in rud level upwind of the source can be expected, more or less independently of :aa actual wind speed. Thus the wind attenulticn is modeled by:
A = 3 cos t (2-8) wind where $ is the angle between the source-receiver line and the direction the wind is bbwing toward.
M, -c 14
_ -S- ,
Stusnick Attechment D, 19 of 30
, 3.0 COMPUTER IMPLEMENTATION The mathematical e.lgorithms described in Section 2 have been implemented in two FORTRAN programs which are designed to run on a mainframe computer system. The use of a large-scale computer was necessitated by the quantity and format of the digital terrain data available fro 7n the National Cartographic Information Center.
The minimum block of data available encompasses one degree of longitude by one t degree of latitude. Since terrain elevation cata is provided every 200 feet, such an area, at latitudes within the continental United States, catains ;n exuss si two million data
~
points. This information is provided on 9-track, one-half-inch magnetic taH. Although j only a small subset of this data is required for any 8 ven 1 alrer., the computer system must be capable of inputting the larger amount of, data so that the desired subset can be abstracted.
< The output provided by the computer sof tware is a series el estimated sireo round levels at regular Intervals along a set of equally spaced radials radiating from the siren.
One program, which is used for siren ranging estimatet to distances of 10,000 feet, products levels et 200-foot intervals along 16 radlais, each separated by 22.5 degrees.
The other program, which is used for siren ranging estimates in excess of 10,000 feet, ph:, duces levels at 300-foot intervals along 72 radials, each separated by 5 degrees.
In addition to the printed output, these programs provide, in a digital file, the distances along each radial at which the 70 and 60 d5 sound levels occur. This file is transferred to a microcomputer in which smcothed 70 and 60 dB und level contours are i computed. A cubic spline fitting procedurs is employed to define eaui contocr at points between the calculated radlais. The resulting smeothed contours can be plotted using either a digital pen plotter or a est matrix pehter.
1 Since the digital terrsin elevation data sometimes di!!ars somewhat from the i I <!evation data provided on standard USGS 7.5-minute topographic quadrangle maps, whien are normally used as the base maps for plotting sound level contours, the resultant contours are overtald onto these topographic maps and manually examined. Any features L of the contours which do not appear to correspond to terrain features on the topographic maps are identified and corrected. 1 I
gy n aecm atoe4, i
jj
Stusnich Attachmont D, 20 of 30 in the absence of barriers, the sound level decreases uniformly with distance from ,
the siren along each radia.l. When topographic variations result in barriers that shield the receiver from the siren, the sound level will decrease sharply just beyond the barrier, effectively reducing the radius of any given contour point along that radial. An example of the effect of barriers on contour shape is given in Figure 3-1.
If the land should rise again beyond the barrier, it is possible to obtain a "hole" in the contour where the shleiding from the stren is localized to a small ran6e of distances along the radial. In such a case, the sound level estimates along the radial in question (and along adjacent radlais) are manually examined to determine 11 the hole should be ignored in general, the guidelines used in this judgment are
- a. If the sound level along any radial drops below the contour for 400 feet or less a
before rising above the value again, the "hole is igrwed and the contour value is assigned to the greater distance,
- b. If the sound level along any radial drops below the contour value for more than 400 f eet but less than 1,000 feet belore rising above the value again, then
- the "hole" is ignored !! the population is low, or
- the contour is pulled in to the distance where the level first orops below the contour value if the population is not low,
- c. If the sound level along any radial drops miow the contour for more than f
l 1,000 feet before rising above the value again, the contour is pulled in to the
- distance where the level first drops below the contour value.
)'
This procedure results in a conservative estim3te of the sound level contwrs.
4 e
i 4
- . . ~ .
LC b
$tusnick AttochmOnt D, 21 of 30
- ~ ,
.- ./
-N \x Withevt i
\
senter Art.nv ii.a ,
, with \ s rlw Attenvetle- N
/
,/ M ,
t !
i d
\ i
( )
/. ,
/ .
/ ,
Siren , I
] /' I t [
\
r :
\ '
i .
\
s a, e
\
~,
%.,/ \
f
/
N s -
70 e 60 m - . . - . .
Figure 3-1. Comparison of $1ren Sound I.evel Centours With and Withcntt tM Effecta of Barrier Attenuation.
17 M.
Stusnick Attochment D, 22 of 30 i 4.0 (NPUT PARAMETERS :
This section describes the values of the it:put parameters that have becq used in exercising the computerized model for the Seabrook area. These parameters can be organized into two groups:
- a. Those relating to individual sirens: '
1.ocation as defined by latitude and longitude,
$1ren helsht above ground level.
Acoustic output as defined by the reference sound level 100 feet from the siren on its axd, Freqvency of the tone emitted by the alren, and Characteristics of the ground (e.g., rural /sulsuroan, cr5an, heavily forested, or water)in the vicinity of the siren; -
- b. Those relating to area-vide meteorological concitions:
- Temperature.
fletative humidity, and Wind direction.
4.1 51ren Parameters Each siren location is defined by determining the siren location on a USGS 7.5 minute topographic quadrangle map and interpolating the corresponding latitude and longitude from the map coordinates. The siren coordinates ue determined to the nearest seccad of arc. These two input parameters relate the siren position to the grid of ground elevation values that are used by the model !or the barrier attenuation calculations.
The measured or proposed alren height above the ground level is also input into the model. This parameter also impact the barrier attenuation calculation.
The acoustic output of each of the siren models employed in the Seabrook system has been determined by direct measurement of the C weighted sound level at a distance ;
of 100 feet on the siren's axis. The values employed in the model are shown in Table 4-1.
l i, N..-
i
Stocnick Attcchm:nt D, 23 of 30 Table 4-1 Acoustic output of Sirens Used in Seabrook System i
C-Weighted Level 5W at 100 Fwt on Axis WS-3000 122 d5 WS-4000 129 dB Dual WS-4000 134 dB These measurements were all made at the tone frequency of 330 Hz, which is utilized in the system. t Finally, the ground characteristics assumed for the entire EPZ region are rurall suburban. No areas have sufficiently high structures to be classed as urban. No areas have sufficiently dense foliage to be classed "heavily forested".
4.2 Meteorolemical Conditions Attenuation resulting from absorption of acoustic energy by the air is a sensitive function of water content (as defined by relative humidity) and temperature, in order to model the worst-case dtuation, average seasonal values of early-morning and mid-af ternoon humidity / temperature combinations were examined to determine which situa-tion provided the largest as absorption coeificient.
Such historical data are not available for Seabr % directly, but can ba interpolated ,
from values at Socton, Massachusetts, and Portland, Maine.I3 Table 4-2 shows the results of tPIs interpolation. Also shown, for the Seabrook area, are calculated values of the air absorption coefficient corresponding to these humidity /teniperature pairs. l Clearly, the worst case (i.e., largest value of air absorptlon coefficient) occurs during a summer af ternoon. The corresoonding values of relative humidity and temperature (38 percent,21.6'C) were used as input values to the model.
[
l i,
f
Stusnick Attcchm:nt D, 24 of 30 l Table 4-2 Average Seasonel Values of Relative Humidity and Temperature in Seabrook Areal 3 7:00 A.M. 1:00 P.M.
1.ocation hth R.H. T Abs.* R.H. T Abs.*
(%) ('C) (4/1000 Ft) (%) (*C) (e/1000 Ft)
Jan 78 11.3 --- 63 - 3.7 ---
Ar P 74 0.2 --- 33 3.8 ---
Portland, ME Jul 80 13.7 --- 39 20.1 --
Oct 83 3.0 --- 60 9.2 ---
Jan 72 - 8.2 0.69 60 - 3.4 0.63 Sea ook Apr 71 2.3 0.33 34 7.3 . 0.61 (Inter. Jul 77 16.2 0.83 38 21.6 0.99 potated) Oct 81 3.7 0.38 45 11.0 0.6%
Jan 66 - 3.0 --- 37 - 1.2 ---
Goston, Apr 68 4.4 --- 34 8.8 --- ,
^ Jul 74 18.6 --- 36 23.2 ---
Oct 77 8.4 --- 37 12.8 l
- Air Absorption Coeffielent at 330 Hz as computed according to Reference 1.
As discussed in Section 2.6, the effect of the presence of wind speed gradients is to improve propagation downwind and impede propagation upwind. Thus attenuation in the direction the wind is blowing is decreased: attenuation opposite to that direction is increased: and attenuation at right angles to the wind direction is unaffected from the no-wind case. The not off ect is to distort the equal sound level contours, alongating them in the downwind direction and foreshorterlM them in the upwind direction.
For example, in the absence of barrier effects, equal sound level contours are circular if no wind is present, if a wind (and tesultant wind speed gradient) is present.
these contours become distorted as shown in Figure 4-1.
1 M.
Stusnick Attachment D, 25 of 30 Without Wind
,-y 4 4 " ' ,
f s
/ # eWith Wind
,/ , \ s
\ %
/
l/
i 4
g s s
/
N I ,' u \
, h \
- \
- e 1
- i $1ren o
i
\ ', l '
\ i / ,'
/
\ ,
/ ,'
\ ~
\ '
s
/ ,'
\ ' / -
N ', / ,
% ~
u mT_$s.- '
Wind Direction Figure 4-1. Eff ect of Wind on Sound Level Contour.
Stusnick Attachmont D, 26 of 30 I
lt is difficult to take this effect into account when designing a siren system, since the direction of the wind at a time that an emergency occurs cannot be known a priori.
Using a time-averaged or a most probable wind direction is not appropriate since there is no guarantee that the wind will be blowing in that direction when an eme mency occurs.
The most conservative procedure is to compute the individual siren contours assuming the no-wind case but to design the overall system so that adjacent sirens are sufficiently close that, with a wind, any "hole" created by the upwind foreshortening of the contour ior a given siren, la filled by the downwind elongation of the contour of the nearest upwind stren. This procedure requires that no-wind siren contt/ars at the edge of the EPZ extend f ar enough outside the EPZ that, if they are foresh.xtened by a wind blowing directly into the EPZ, the distorted contour still reaches the edge of the EPZ.
For Seabrook, the currently allowed siren locations along the western edge of the EPZ do not provide excess penetration beyond the EPZ at several locations. However, as will be demonstrated in Section 3, there is an inherent conservatism of 10 dB in the model. Since the inclusion of a wind blowins from west to east into the EPZ would have had the effect of reducing the predicted levels by 3 dB, coverep will extend past the edge of the EPZ.
l
1 Stusnick Attochmont D, 27 of 30 3.0 YALIDAT10N OF MODEL The computerized model described above has been validated by comparing its predictions with measurements carried out near sirens at several nuclear power plants.
Comparisons of measured sound levels with predicted sound levels from the model are shown in Figures 3 1 and 3-2.
Figure 3-1 shows a comparison between measured and predicted sound levels for ;
sirens near the Indian Point plant in New York. Most of this data was taken in very hilly, heavily forested terrain. Shown for reference on th!s figure is a 43-degree line indicating perfect agreement. The average difference between measured and predicted levels is 7.2 dB with a standard deviation of 3.3 dB.
The spread in the data is due to atmospheric variations during the measurements and terrain effects not accounted for in the propagation models. There is an offset such that measured levels are generally higher than predicted. This result is copslatent with the design goals of the siren siting model, which endeavors to be conservative in the predletion of the sound level so as to minimize overprediction of individual siren levels.
A reasonable design objective is that there be no more than a 10 percent probability that actuallents will be less than predicted. For the data set shown in Figure 3-1, !! of the 93 measurements are underpredicted, corresponding to an ll.6-percent probability of underprediction.
Figure 3-2 shows similar data for predictions and measurements for sirens tested at the Seabrook plant during the period from 3 March to 7 April 1988. The terrain in the test area was flat and rural. Again, as in the case of the indlan Point data, the model underpredicts the sound lent, as it was designed to do. The average difference between measured and predicted levels is 10.0 m with a standard deviation of 8.7 dB. Of the 123 measurements represented in this figure,10 are underpredicted, corresponding to a rate of 4.0 percent.
In summary, the siren range predletion model presented in this report is shown to provide a reasonable and conservative basis for siting siren positions. Based on the model validation measurements repor'ed herein, the model predicts a shorter range than actually observed about 90 percent of the time.
23 M.
Stusnick Attcchm nt D, 28 of 30 l
l
!10.00 , i , .
' a o,' C oCoo o
o o o o 100.00 - ,
,o I i
- 90.00 - g, l a Oo
- a oo oc go' 3 o %s .
! I 90.00 - ,* o o I o o o
og *,4 o o o % oco -
70.0C - o o o o
- o oo o
60.00 90.00 100.00 110.00 60.00 70.00 80.00 Predicted Level. d8 Figure bl. Comparison of Measured and Predicted Siren Sound Levels f or Indian Point Data.
l a
26 S..
Stucnick Attcchm:nt D, 29 of 30 R i . . . . i . , ,
O*
- a. . .<* -
9
- e4 g3 -
go .
1 . -r sg . ...e'
).
i
. . . n .s -
', s
- - = *
,' , se e #
l e e y e,
- 3 -
8
.e .
2 -
g -
t t t t t t i t i 30 40 M M 70 to to 100 110 120 130 Predicted Level, OS Figure 3 2, Comparison of Measured and Predicted 51ren Sound Levels for Seabrook Data.
t l
l
Stusnick Attachment D, 30 of 30 REFERENCE 5
- 1. American National Standards Institute, "American National Standard Method for the Calculation of the Absorption of Sound by the Atmosphere", ANSI 51.26-1978 (ASA 23-1978).
- 2. Sutherland, L.C., "Review of Experimental Data in Support of a Proposed New Method for Computing Atmospheric Absorption Losses", DOT-T5T-73-87, May 1973.
- 3. Brown, E.H., and Clif ford, 5.F., "On the Attenuation of Sound by Turbulence",
- 3. Aeoust. Soc. Am., M, pp. 748-794,1976.
4 Sutherland, L.C., "Scattering Attenuation of Sound in the Lower Atmosphere",
- 3. Acoust. Soc. Am., 49,,
9 p.129(A),1971,
- 3. Sutherland, L.C. (Ed.), "Sonic and Vibration Environments for Ground Facilities - A Design Manual, Chapter 7: Propagation Effects of Acoustic Waves", Wyle Research Report WR 64-2, March 1968.
- 6. Delany, M.E., menge Predletion for Stron Sources", National Physical Laboratory NPL Aero 5pecid Report 033, November 1969.
- 7. Lyon, R.H., "Rc!e of Multiple Reflections and Reverberation in Urban Noise Propagation", J. Acoust. Soc. Am., y, 493-303,1974
- 8. Maekawa, Z., "Noise Reduction by Screens", Memoirs of Faculty of Engineering, Kobe University, Japan, No, 11,1963.
- 9. Isel, T., Embleton, T.F.W., and Piercy, J.E., "Influence of Reflections at the Ground on Insertion Loss of Barriers",3. Acoust.5cc. Am., g, p. 339,1978.
- 10. Beranek, L.L., Noise and Vibration Control, McGraw-Hill Book Co.,1971, p.177.
!!. Pridmore-Brown, D.C., and Ingard, U.,"Sound Prepagation into the Shadow Zone in a Temperature Stratified Atmosphere Above a Plane Boundary", J. Acoust.5oc. Am.,
E,p.36,1933.
- 12. Pridmore-Brown, D.C., "Propagation of Sound into a Wind-Created Shadow Zone",
NACA Report RM 37523, 1937.
- 13. Tedrid, R.N., and Polly, R.C., "Measured Acoustic Propagation Parameters in the Misslasippi Test Operations Area" NASA TM X ll32, August 1963.
14 Jenkins, R.H., and Johnson, J.B., "The Assessment and Monitoring of the Contr!bu-tion From a Large Petrochemical Complex to Neighborhood Notse Levels", Noise Control Vibration and Insulation, November / December 1976.
- 13. "Statistical Abstrset of the United States", U.S. Department of Commerce,1978.
M Rt i
_. - -_ _. ._- . _ _ _ _ _ - _ - _. _ . - _ _ . _ _ _ _ _ = _
Stusnick Attcchmont F, 1 of 1 The reasonableness of the method for determining the siren sound l
cutout and the resulting siren sound pressure level contours should be documented in the design report. The validity of the sound pressure level contour calculation depends upon the validity of the determination of siren sound output at 100 feet ,
from the siren. There are. at least two ways to determine siren sound output
. Onsite field measurements around at least one of each i type of siren used within the EPt s or Anechoic, semi-anachoic, or reverberation chamber testa in a qualified laboratory on sirens that are repre-i sentative of each type of siren used within the EPt.
i Since consensus standards a e not available for field and j
chamber siren measurements, the rationale for the employed measurement procedures must be detailed in the design report.
The ' design report should provide a list of all sirens and should contain the following information for each siren: unique iden-tifier, siren type, sound output in dBC at 100 f eet, and mounting height.
i The design report demonstrates compliance with NUREG-0654/ !
FEMA-RIP-1, Revision l e criteria for those geographical areas covered by fixed sirens by showing that either:
Se espected siren sound pressure level generally '
exceeds 70 dBC where the population exceeds 2,000 persons per square mile and 60 dBC in other inhabited l areast at The expected siren sound pressure level generally exceeds the average measured summer daytime ambient ,
sound pressure levels by 10'dB (geographical areas with less than 2,000 persons per square mile) .
If the design report documents that the siren sound pressure levels exceed a measured ambient by 10 da, then the following inf ormation should te provided: l E-8 i
. _ _ - . - - _ _ _ _ _ _ . _ . _ _ . _ , , . . _ . . _ _ . . _ _ _ _ _ . _ _ . _ _ _ . . . _ - . _ _ _ _ _ _ _ .