ML20063G053
| ML20063G053 | |
| Person / Time | |
|---|---|
| Site: | Indian Point  |
| Issue date: | 10/26/1981 |
| From: | Anderson G, Kearst D, Towers D BOLT, BERANCK & NEWMAN, INC. |
| To: | |
| Shared Package | |
| ML20063G050 | List: |
| References | |
| FOIA-82-236 NUDOCS 8207280069 | |
| Download: ML20063G053 (517) | |
Text
- - - - - - -
l
.s..
- s. .
Bolt Beranek and Newman Inc. '.
I "'"85 r
~
o EVALUATION OF Tile PROMPT ALERTING SYSTEM FOR T11E INDIAN POINT NUCLEAR POWER STATION A Memorandum Report
_ 26 October 1981 D. A . Towe rs G.S. Anderson D. N . Keast
,r.
Submitted under Subcontract No.
B-A2740-A-V Submitted by:
~ Bolt Beranek and Newman Inc.
10 Moulton Street Camb ridge , Massachusetts 02238
~ Submitted to:
Ba ttelle
' Pacific Northwest Laboratories
. P.O. Box 999 Richland, Washington 99352
~
8207280069 820618 g h h-236 PDR
i
- a L_
l EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE
"- ~ INDIAN POINT NUCLEAR POWER STATION i
A Memorandum Report 26 October 1981 i
.m_
D.A. Towe rs G.S. Anderson
- D.N. Keast i
L Submitted under Subcontract No.
t B-A2740-A-V L
Submitted by:
L Bolt Beranek and Newman Inc.
L 10 Moulton Street Cambridge , Massachusetts 02238 i
Submitted to:
Ba ttelle Pacific Northwest Laboratories
! P.O. Box 999
~~
Richland , Washing ton 99352 9
Sum.
m
I
.y *
. o EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE INDIAN POINT NUCLEAR POWER STATION This memo report summarizes the evaluation of the siren
~
alerting system for the Indian Point Nuclear Power Station. The
, procedure that was' used consists of a detailed analysis of siren alerting capability at each of 50 randomly chosen listener locations, under four different " sample scenario" conditions.
L Both the random selection process for listener sites and the four test cases (sample scenarios ) are described in material submitted
[ by BBN to Battelle on 9 September 1981, and are included here as Ap pe ndix A. The analysis is based on existing and proposed siren locations as of 25 August 1981. Maps which show the siren locations are provided in Appendix B.
f L. The results of the evaluation for Indian Point are summarized in Table 1 and indicate that 'the chance of alert
[ ~
is estimated to vary between 57% and 95% depending on the sample scenario under consideration. The remainder of this report
- l. describes the procedure used to arrive at this conclusion.
Le l Estimating Siren Sound Levels Out of Doors at Listener Sites t_.
The first step in the procedure is to determine that siren
_ in the vicinity of each selected listener site which is expected to produce the highest sound level at that site for each sample
- ,' scenario. This choice is not always obvious, because the sound level caused by a particular siren at a given listener site depends not only on the sound output of the siren and its
"~
distance from the listener, but also on shielding and atmospheric
.. ef fects (particularly wind direction) . Therefore, it is generally necessary to evaluate several sirens in the vicinity of i
9 w
W.
TABLE 1.
SUMMARY
OF INDIAN POINT SIREN SYSTEM EVALUATION RESULTS.
w.
Chance of Alert
_. Scenario Po pul ation-Urban Rural Weighted Average
- No. Description (%) (%)- (%)
1 Warm Summer Weekday After-L_ noon (clear to partly cloudy) 98 93 95
- [, 2 Summer Weekday Night (clear to partly cloudy) 80 70 74 3 Winter Weekday Evening (cold and overcast) '
91 78 83 i
'.L 4 Winter Night (during snowfall) 63 53 57 i,
'-
- Based on a total urban population ofl10,928 and a total rural population of 146,454.
ll
. m.
I
't -
,L. -
!i L. ~
l:
1 i I l- .
each listener site in order to determine the dominant one. As a general rule, the closest, highest-rated, nonshielded sirens, are
_ selected for evaluation at each site. Fur the rmo re , sirens are chosen such that they are distributed north, south, eas t, and I
i- west of the site (or in 'any other four mutually perpendicular directions) where possible to account for different wind d irect ions . For the Indian Point analysis, 4 or 5 sirens were evaluated at each of the 50 listener sites.
[,
i The next step in the procedure is to establish the outdoor sound level produced by the selected sirens at each listener
- l. location. This is accomplished by applying adjustments to the rated sound level of the siren as follows:
L(listener) = L(siren) -Ad-A s -Amol - Aa tm '
cI where L(listener) is the outdoor siren sound pressure level at the listener site (dB), L(siren) is the rated sound pressure level of the siren at 100 ft (dB), Ad is the distance attenuation
~
(dB), A s is the shielding attenuation (dB), A is the air mol j
absorption (dB), and A atm is the atmospheric attenuation caused by wind and tempe rature gradients (dB) .
The rated sound pressure level for all of the Indian Point sirens was taken to be 125 dB at a distance of 100 feet, based j upon the siren manufacturer's rating; all sirens are rotating type un its .
L- The firs t two adjus tments - ( for distance and shielding) are the same for all four test cases and are based on information j ,
obtained from USGS maps. Distance attenuation beyond 100 feet is calculated by assuming sound propagation from an acoustic point m
v.
(.
source with a reduction of 6 dB per distance doubled. It is -
calculated as follows:
dh Ad = 20 log 10 TUU) '
shere d is the siren-to-listener distance (ft).
Shielding attenuation (As) is estimated using the following _.
formula for the attenuation of a rigid straight barrier for sound incident from a point source [1] : _,
e {2nN As =
20 log tanh Y2n N + 5 dB for N 1 -0.2
~
=[ 0 othe rwise N is the Fresnel number (dimensionless): -
N=
A ( A + B - d) s-Where A = wavelength of sound, f t (1.79 f t for 630 Hz siren tone) d = straight-line distance between source .and receiver, ft
~
A+B = shortest path length of wave travel over the i barrier between source and receiver, ft --
+ s ig n = receiver in the shadow zone (i.e. , barrier obs tructs line-o f-s igh t )
- s ign = receiver in the bright zone (i.e. , barrier doesn' t obstruct line-o f-s igh t)
When N is nega tive , the above equation for A s is evaluated by replacing N with lNl, and by replacing tanh with tan. .
Shielding attenuation is limited to a maximum of 24 dB based l upon a large body of experimental data. For the Indian Point l
i w
w WW L
h
_, analysis, sirens are assumed to be at a height of 52 ft above terrain level, listener sites are assumed to be at a height of 5 1
- i. f t above terrain level, and barrier heights are obtained from ground contour information on USGS maps. ,
__ The adjustments for air absorption and atmospheric ef fects depend on the meteorological conditions for the particular scenario. The assumed conditions for the Indian Point site are provided in Table 2 for the four test cases, based on local weather information [2] . In terms of air absorption, these conditions indicate the following attenuation rates based upon
- j. temperature and relative humidity [33:
L.
Scenario Amol (dB per 1000 f t. )
1 0.85 h 2 0.81 3 0.49 4 0.46
)
The adjustment for atmospheric gradient ef fects ( Aa tm) is based on siren-to-listener azimuth with respect to wind direction and on wind and temperature gradient characteristics. Table 3 l summarizes the calculation procedure for determining Aatm ff each scenario at the Indian Point site. A more de tailed
- description of the estimation procedure for A a tm can be found in
-D Ap pe nd ix C . .
i Application of the above calculations yields the estimated
~~
outdoor sound pressure level for various sirens at each sample listener site, for each of the four scenarios. For the balance of the analysis, only the highest siren level at each listener site is used.
MW m
2
' TABLE 2. METEOROLOGICAL CONDITIONS FOR THE FOUR SAMPLE SCENARIOS USED TO EVALUATE THE. INDIAN POINT SIREN SYSTEM.
4 Relative -
Scenario Temperature Humidity Temperature No. Wind Conditions
- Gradient (%) (OF)
. 1 10 mph from the SSE; from -1.00F/100 ft. 65 80 '
the south in the river' Class A valley 2 6 mph from the NNE; from +0.50F/100 ft. 80 70 the north in the river Class E valley 3 10 mph from the northwest -0.50F/100 ft. 70 30 Class D _.
4 15 mph from the southeaet -0.50F/100 ft. 90 30 Class D i
~
- At 100 ft. above ground level.
e e
GmEmW
- s. ,
e
'-- TABLE 3. - CALCULATION OF ATMOSPHERIC ATTENUATION, A a tm' CAUSED BY WIND AND TEMPERATURE GRADIENTS (See Aooendix C for details)
-- Siren-to-Listener I Distance, D (Ft)
Relative to Xn (Ft) Aatm (dB)
D 1 1.2 X o o 1.2'X, < D 1 1.7 Xo 5 i 1.7 X, < D i 2.4 X, 10 -
L-2.4 K, < D 1 3.4 X, 15 3.4 X, < D 20 t
L l
u_
- Computation of Xn Xc e .f 1057/6Scos$-a L _-
Scenario 1 2 3 4 1
Wind Direction, O General 157.5 22.5 315 135 h Valley 0 180 - -
w AT F (95'-7') -0.9 +0.44 -0.44 -0.44 a = as = AT/in 95'-in 7') -0.35 +0.17 -0.17 -0.17 Wind Speed, V ft/sec@l00ft 14.7 8.8 14.7 22 2 ,
a8 - V2 / (in 100' - in 2') g_
3.75 2.25 3.75 5.62
/
V R/S 0.1 f(R/S) Op5
~~ \
X (nin) @ $ = 0 I522' 633' 534' 439 950 850 93 0 920
&c = Cos (f) h h
we e i i
J Estimating Indoor Sound Levels of Sirens ,
The result of the above calculations is a single outdoor siren sound pressure level at each of the 50 sample listener ,
locations for each of the four t'est cases. Corresponding indoor levels are then obtained by subtracting typical values for _
commercial or residential building sound attenuation. Sor test cases 1 and 2 ( s umme r) , residential windows were assume'd to be partly open; for test cases 3 and 4 (winter) residential windows were assumed to be closed (with storm windows ) . For the frequency region around 500.'z, H the sound attenuation into buildings is estimated to be 16 dB for test cases 1 and 2 and 31 dB for test cases 3 and 4 [4]. For commercial buildings, the _
ou tdoor-to-indoor noise reduction is es tima ted to be ,31 dB ,
assuming closed and sealed windows for all four scenarios. ,
Assumptions about Chance of Alert ,
The outdoor and indoor siren levels calculated by the above procedure provide some of the information required for the 't analysis of the chance of alert. In addition, it is necessary to know the level of interfering background noise at the listener locations.
Figure 1 is a flow chart of the analysis computations. ' The _
analysis is divided into components (rows') that correspond to the i
possible activities of people for the 'various scenarios. The
~
major components relate to people '(1) at home (outside or ins ide ) , (2) a t wo rk, or ( 3 ) in motor vehicles. The chance of l alert is estimated for each activity . component and is then multiplied by the fraction of people likely to be engaged in that activity (activity fraction). The results are summed to obtain . - .
the overall chance of alert for each ' listener location and for l
l M
m
e 0 0 0 0
i e
W.
I w
t- 1 3, v
Ll E s, h:m b b $k
- 1. 2_
h n ,],!
2
- z. -
7 i
il [ T]@ 5 [i , [ [
_ r . .
1 l .S -- m 11 11 v @ r l
'-. I ! $ 11 1 J .
, r j
i} ,
}I} ,
- I t,
,l 2 I i I!9 85 2.
] .
85 l q- 't
- 2. __
& gi ,.
) -
u 1 ' il 55 I I
; yT ,
" 5_ .
11: hy _a Ig
-i -
It i
!_ t 11 Ill 4 ..
l l . s
'yll la 18 5 i , j:- w x
- + s=
il lu 3
/.
ii 111 , 1)111@' 1 .
. 1. )_.
- lh a m _
ll
!! d o i s"
== l essee un GG t ou t I-M So')
a i i d L 1 i
-l i s r3 si s
a es ' L. 4
.e.m e.
( each test case. Overall chances of alert for the various scenario (test case) conditions are then obtained by averaging the chances for all rural and/or urban sample lister.er sites. Note that all estimates assume siren signal duration of 4 minu tes : an average of the "3 to 5 minutes" called for in Appendix 3 of NUREG-0654. The ef fects of dif ferent siren signal -- durations are discussed in Appendix D. Siren detectability is a function of the siren signal level and of the background noise level in a " critical- frequency band" centered at the signal frequency. For this analysis, outdoor and - indoor de tectability is estimated based on the signal-to-noise (S/N) difference in the 630-Hz 1/3-octave frequsncy band. The _ chosen criterion for alerting is that the given signal level must be 9 dB or more above the minimum background noise level at any time during a 4-minute period for people who are not sleeping (i.e., a S/N difference of 9 dB). The chance of alert while sleeping is based on the indoor siren Single Event Level (SEL) - - a measure of total acoustic energy - and the sleep awakening l model, developed by the U.S. Environmental Protection Agency [53. . l The graph used for estimating the chance of alert during sleep is shown in Fig . 2 ; for the Indian Point analysis, -the curve for the , chance of awakening one out of two sleepers was used. Alerting People Out of Doors F';r the analysis of the ability of sirens to Llert people out of doors, background noise levels are based on previous noise j measurements conducted by BBN in the vicinity of the Trojan Nuclear Plant in Oregon, near the Indian Point Nuclear Power Sta. tion in New York, and upon the body of data in BBN files. -The data typically consisted of statistical summaries of background noise at various types of locations. The summaries provide the s_ l 4 aump
. , _ _ ,,__,__._._,,_m. . , _ , _ . , _ . , _ . , . , . - _ _ _ _ . _ -
1M . i l 1
. I
'.- l 99.9
-/
99.8 99.5 l i
- 9.
l , l l
- , ,/ ,
't- 1 Out of 3 Sleepers, /
a y" /
/1 Out o'f /
/
g r /^ 2 Sleepers p w 70 N ff f' N1 Siesper 6 [ [ [
// /
l Y" 8 'o
// /
7,- 7 lN 10
)?/ SEL = L + 24 For Stationary Sirere SEL = L + 18 For Rotating Sirera.
5 p ,4, g;no,,3;,,, op.,,,; ,, f The Mkkfle Curve is Used For 2 This Analysis l 1.0 0.5 l y 0.2 0.1 l,, 0.05 'L ,y 40 50 60 70 30 90 300 110 120 130 140 l lNDOOR SIREN SEL { db ) w
} FIG. 2. CHANCE OF ALERT FOR AWAKENING PEOPLE ASLEEP.
l L.
\
l
L (sound level exceeded 90% of the time) for 1-minute samples of 90 _ data in the 1/3-octave frequency band centered at 630 Hz.* .These data were used to estimate the range of background noise levels that are likely to exist during any 4-minute period (1-minute for rotating sirens) for a variety of outdoor environments. The results are summarized in Table 4, which specifies the background - noise environment for urban and rural areas. Only daytime noise levels are presented since the nighttime scenarios assume that _ essentially no people are outdoors at night.
~
The siren sound level necessary to alert 9 dB greater than the minimum background noise level that could exist in any 4-minute period (1 minute for rotating sirens) , adjusted for the probability - distribution of such minima. The chance of alert for people outdoors was determined for each scenario at each listener site using _ Figure 3. I outdoor background noise in urban areas and along rural -- roadways is caused predominantly by motor vehicle traffic. It is generally insensitive to seasons of the year, but varies markedly _ with time of day. Minor traffic variations (i.e., less than a factor of 2 in traffic volume) have little effect on the - background noise. In rural areas remote from roadways, outdoor background noisc - can be seasonal (brids, insects, etc.) and can vary with the weather (wind, rain, waterflow, surf). Few people live or work in suc.h " natural" acoust1c environments.
*The L was used as a conservative est1 mate of tha minimum sound level.90 v
= -
emme
i t t._ TABLE 4. MINIMUM BACKGROUND NOISE LEVELS FOR GENERALIZED f."- CATEGORIES,OF OUTDOOR ENVIRONMENTS. (See Figare 3 for Distributions) g Generalized Background Range of Minimum Background Noise Environment Noise Levels for a 1-Minute Period l ,2 (dB) i- 3 L I. URBAN-DAY (Includes Rural g locations within 21-57 1000 ft. of major { roadwa's) y
- I t_ II. RURAL-DAY 4 *
(Except Rural
;. locations.within 17-48 1000 ft. of major '
- roadways)
( i , NOTES: l i- -- Refers to the range of the minimum. (L90) sound pressure levels 1. in the 630 Hz one-third octave band during the specified time period.
" 2. Apr,licable for analysis of rotating sirens operated for 4 minutes.
i ' b- 3. Urban locations are defined as the p'nk " building exclusion" areas of topographic maps, or as those communities with a population density exceeding 2000 people per square mile. Major roadways are definad as roadways with more than one lane in each direction. l 4. Rural locations are taken to be all sites not classified as I u_ urban (above) .
=m==
l ! 5-I
, - , - y . , - , - - , - , - -
k 9 0 S - t 1 100 99.9 99.8 99.5 . gg ' 98
/
RURAL-DAY
$ [
g s - / J '-__ a0 / / E. 0 { / N 50 8 / / \ _ s"' -/ / \ l 20
/ p tuRoN_oAY 1
10
/I
S
/ / .
1.0 I I m2 l 0., 0.05 OT / 0 10 20 30 40 50 80 70 80
, OUTDOOR SIREN LEVEL (d8)
FIG. 3. CHANCE- 0F ALERT FOR PEOPLE INDOORS (4-MINUTE ROTATING SIREN). m. W
I.
'- Note that rotating sirens would actually produce their
, estimated sound level during about one quarter of the presumed 4-b_ minute operating time at any particular listener location. Thus, the results for rotating sirens are based on 1-minute statistics rather than on' 4-minu te statistics .
j In s umma ry , informa tion regarding. siren type , estimated siren
-- sound level, background noise category at the listener site, and tes t-case conditions were u' sed in conjunction with Fig . 3 to
[ estimate the chance of " siren detection outdoors at the Indian Point Site. L Alerting People Indoors -1 L For the analysis of alerting people indoors at home, three types of activities are considered. These are (1) listening to I radio or TV, (2) sleeping , or ( 3) other activities that range trom l quiet to noisy situations. Table 5 provides the percentages l assumed for various activities for each scenario. l For people listening to radio or TV, the chance of alert is
, 100%. For people sleeping , the chance of alert is calculated from the indoor siren SEL using the relationship shown in Fig . 2 for
!g the chance of awakening one out of two sleepers. For all other
~
indoor activities, the chance of alert is based on classifications of actual indoor background noise measurements under a wide
-- variety of conditions.
- [ Results for test cases 1 and 3 are provided in Fig . 4 for 4-
~
- minute rotating sirens. Thus, given the indoor siren level and test case condition, this figure was used to estima te the chance
'- of alerting for indoor activities other than sleeping or listening to radio or TV.
'f I w
TABLE 5. ASSUMED ACTIVITIES AND BACKGROUND NOISE ENVIRONMENTS FOR PEOPLE IND0 ORS. i Percentages of People Engaged in Various Activities Indoors (%) Indoor Noise Environment At Place Listening to Obviously Busy and )bviously Scenario of Business TV/ Radio Sleeping Noisy" Active Isolated 8 Qui,et"
- 1. Warm Summer Weekday 41 27 5 - - - 8 5 14 Afternoon (clear to partly cloudy).
~
- 2. Summer Weekday Night 4 ' ' - - 96 -- -- -- --
(clear to partly cloudy)
- 3. Winter Weekday During --
20 -- 5 50 20 5 Evening Commuting llours (cold and overcast) ,
- 4. Winter Night During 5 --
95 -- -- -- -- Snowfall NOTES:
- 1. Vacuum cleaning, dishwasher, shower, vent fan on, etc.
- 2. Dinner conversation, kitchen work, playing music, children at play, etc.
- 3. Noise-producing activity in adjacent room, soft background music, etc.
- 4. Reading, study, eating alone.
( f ^ i . i i : i , t , i . ,
O $ 9 l 100 U l~ . 99.9 99.8 1
- 99.5 1 /
i ,, If j./ N L ,
/ /
f /
, 2a
= 70 s- @ l' l' l g
" s, s' /
y ,, J / g o 20
.y j w. w
..f
, . f
/
t 10 ra - ic- .h- - i. 2 { i.0 7 0.5 1
; 0.2 0.1 0.05 1
0 20 30 40 50 60 70 80 90 100 0 10 INDOOR SIREN LEVEL ( db ) FIG. 4. CHANCE OF ALERT FOR PEOPLE INDOORS AT HOME (4-MINUTE ROTATING SIREN). 1 enese y - - - , . .
i O ,
, '~
( For the analysis of alerting at work, two activity categories ~ are considered: (1) commerical/ institutional, and (2) industrial e nvi ronme nts . For the Indian Point analysis, it was assumed that 75% of the working population are in commercial establishments while the remaining 25% are in industrial locations. For commercial locations, the chance of alert is based on the _ statistics of background noise measured in a typical of fice environment, using Figure 5. For industrial locations, it has , been assumed that 100% of the people are likely to be alerted by some means of communication other than sirens. _ Alerting People in Motor Vehicles The analysis for the alerting of motorists ~is based on the assumption of an average siren signal strength and spacing throughout the EPZ. The chance that a motorist will pass. within the alert range of a siren. during its 4-minute operation is s-estimated as follows: C = 2 R+d x 100 (not to exceed 100%) where C is the chance of alert (%), R is the maximum alert - distance ( f t), d is distance traveled in 4 minutes ( f t) , and L is the average siren spacing (ft). Separate analyses were carried - out for urban and rural areas of the Indian Point EPZ. The average urban or rural siren produces a sound level of 1.25 dB at 100 ft. Alerting ability was evaluating by using the results of a study for the U.S. Department of Transportation ~ ( DOT) [6]. Siren alerting levels for speeds of 55 mph and 30 mph with windows shut or open were first determined from the DOT study ' results. The average siren source levels for rural and urban he tumumf
. - - , . , , - - - - - * - - - --,- --t-- v - -r
3 8 h # s i L 100
- S I
L* 99.9 99.8 4 MINUTE STATIONARY SIRENN f4 MINUTE ROTATING SIREN 99.5 99 _.._ i c i 95 i y 80 1 t s. : w i .0 m 1 M 'I*7 L N 40 '
,' .i 30 i
s "
- 4, L 10 i
5
~
2 1.0 g L 0.5 0.2 5 0.1 0.05 O - 0 10 20 30 40 50 60 70 80 90 100 INDOOR SIREN LEVEL (dB) FIG. 5. CHANCE OF ALERT FOR PEOPLE INDOORS AT WORK IN COMMERCIAL / INSTITUTIONAL ESTABLISHEMENTS. h I t_
,- ~ - , . , ~, - - - , - - - - - ,n-, - , - - , - - . - , - ,, , , , , - .--
( areas were then reduced to alerting levels in accordance with the _ propagation models from current NRC guidelines ( i .e . , 10 dB/ double distance).[7]. In this manner, the maximum alert distance (R) was calculated for .each driving condition. The distance' traveled in 4 minu tes (d) was calculated based on speed for each case, and the . average siren spacing (L) was estimated to be 4,890 ft for urban - areas and 12,530 f t for rural areas. s The calculations of alerting ' ability for motorists are summarized in Table 6. The results indicate that the chance of
~
alert is expected to be 100% for all conditions applicable to the Indian Point analysis. , Input and output data for the analysis are included in Appendix E. Comment and suggestions regarding the procedure are
~
mos t welcome . t
, me m
W
=
b I l
, ~ , - e- ,,,n -------w-,---w- , ,m- - ~ ~ = -, ,s--w+. , , , , - >--,---s g,,-,,,o , -- -,e-m-, ee- w e -em
,g- (-- _p (. 7 .. ( ... .. .
\.
TABLE 6. SIREN ALERTING FOR MOTORISTS. Vehicle Vehicle Reqd. Signal Max. Alert 4-min Avg. Siren Chance i Speed Window for Alert Dist. R Travel Dist. , d Spacing L of Alert Area (mph) Condition (dB) (ft) (ft) . (ft) (%) URBAN 30 Closed 89 980 10,560 4,890 100 , Open 86 1,210 10,560 4,890 100 RURAL 55 Closed 96 650 19,360 12,530 100 Open 90 980 19,360 12,530 100 l
l . REFERENCES ,
- 1. L.L. Beranek, ed., " Noise and Vibration Control", Chapter 7, Section 7.6, McGraw-Hill Book Company (19 71) .
- 2. Final Facility Description and Safety Analysis Report for Indian Point No. 3 Nuclear Power Plant, Section 2.6 (1973). ..
l 3. C.M. Harris, ed., "Hardbook of Noise Control", Chapter 3, _, l l McGraw-Hill Book Company (1979).
- 4. D.A. Driscoll, J.P. Dulin, Jr. , and D.N. Ke as t , " Attenuation of Northern Dwellings to a Linear Source of Noise." Presented at the 95th Meeting of the Acoustical Society of America, -
Providence, R.I. (May 1978). t ( 5. U.S. Environmental Protection Agency, " National Roadway Noise i Exposure Model," Draf t Report (1980).
- 6. R. C . Potter et al., "Ef fectiveness of Audible Warning Devices i
on Emergency Vehicles," U.S. Depar tment of Transportation, Report DOT-TSC-OST-77-3 8, Washington, DC ( August 19 77) .
- 7. " Criteria for Preparation and Evaluation of Radiological Emergency Response Plans and Preparedness in Support of _
Nuclear Power Plants," Nuclear Regulatory Commission, NUREG-0654, Revision 1, Appendix 3, Part C.3.e, Washington, DC (November 1980). (The Federal Emergency Management Agency also publishes this document as FEMA-REP-l.) M hmW
'M
j ,
- 9 O
\
APPENDIX A i C e v .- e e + e d fo e n go /1G /8 s Lt- . kcac+ fo . D e s ,. , , ; g , , h
/
A-1. RAN DOM SELECTION OF POPULATION-NEIGHTED LISTENING POINTS AT THE INDIAN POINT SITE The objective of the listener-site-selection process was to identify.50 randomly selected building locations within the EPZ s ur rounding the Indian Point Nuclear Plant. These locations are assumed to be residential locations and are called " listener s ites" . The various ' steps used in the site selection procedure are described below:
- 1. A population-distribution map (see Fig. 1) consisting of a 10-mile-radius circle divided into annular sectors defined by interior circles and radii, was superimposed on topographical. maps of the EPZ. Population distribution informa tion
- consisted of the number of people within each annular sector. These data were used in order to popula tion-we ight the randem selection. process j described below.
l-
- 2. Each annular sector was firs t assigned a ,
des ig na tor , ranging be tween A-1 and R-10 (see Fig. l 1). A range of numbers was then assigned to each i
, , . . - _ . . _w.. ,_m - , , . , . _ . , _ , , .,..w -_.,y .., ,_ _ ,, .-,v.,
sector according to the popula tion in that sector. For example, Sector A-1, just north of the site, and sectors B-1 and C-1 (moving clockwise) have zero population and thus were not assigned any n umb e rs . Sector D-1 has a population of 35 and was assigned numbers 1 to 35. Sector E-1 has a population of. 60 and was assigned numbers 36 through 95. This process was continued until each number between.1 and 256,015 ( the total estima ted population) was assigned to a particular sector. A random number generator (available on a Texas Instruments Model TI-59 hand calculator, for example) was then used to select 50 numbers at random between 1 and 256,015.' Each number selected represented one site ( to be chosen la ter) within the sector containing that number. Thus, sectors with larger popula tions had a greater possibility
, of including chosen listener-sites.
- 3. Having de termined the sector locations of each potential listener site, the next step in the procedure involved selecting the actual-sites within the respective sectors. This was accomplished by firs t ove r-laying a rectangula r coordinate grid on each sector of interest on the topographic map. The grid was composed of boxes O
. _ , . , ,-, - , , - - --r v.- .,.--e-.--,,--.m. - ,c ,-,i---- ,,--,----------.--,--,-v -- w a-
l with dimensions of approximately 1000 feet square, and each box was assigned an X and a Y coordinate
~
according to its location on the grid. The grid was positioned such that the X-axis was oriented in the east-west direction an'd the Y-axis was oriented in the north-south direction, and such that all parts of " the sector of 'interes t we re cove red by a positive (X,Y) coordinate pair box. A random number generator was then used to select random pairs of numbers within the X and Y ranges covering the sector of interest. Each X,Y pair was used to select a particular 1000 feet square box on the map. If there were buildings within the box, one of them was arbitrarily chosen as a listener site. If there were no buildings ins ide the box or if the box fell outside of the sector of interest, that coordinate pair was disregarded and another pair was chosen at random. For urban sites in the pink " building-extension" area of the topographic map a residential building was always assumed to exist, and was sglected at the center of tne pink area in the 1000 feet square box. G-S
, ,a
', ~
\ . s-i
- 4. The above process was repeated until 50 listener sites were randomly chosen. It was found, however, that some major urban communities did not include l , any listener sites, and thus the chosen sites did '
not properly reflect the population distribution in the EPZ. Therefore, th'e selection process was
?
l continued until this condition was rectified. Four new urban sites were randomly chosen to replace the j four mos t recently chosen rural sites. This I replacement only affected the balance between urban and rural listener sites. Since .the subsequent I analysis treats urban and rural areas separately, this replacement will not bias the results. It l
~w ill merely ensure that no major population concentrations are ignored . The above procedure ] resulted in a pseudo-random sample of 50 specific l
listener locations, distributed throughou t the EPZ as shown roughly on Fig. 1. (The actual locations of each specific house are on topographic maps not included with this memo.) In subsequent steps of the siren system
, evaluation procedure, siren audibility will be caref ully l estimated indoors and outdoors at each' of the 50 sample , sites. Based on these estimates, statistical j conclusions will then be drawn'regarding overall siren cove rage .
_. - L
- :n
. n .7
- '. ; A
. R -
- i
. i j i * .
r,- q .. I, . I a
-1 --. s
- 0 P ,e ,
t; - i
! '- W # ..
A.
- - h- 5 _6 .J .8e _1 LO, E a N --.- - - <-- -- - --. -. --
e
'I
~ I l'I '
w ,
.
- 97- .
e s' I e
; - el '
- e l C t
.e -. ,
i
- l' e. l e P K H J -
Population Density / Block {' l J below 2,000/mi 2
; , 2,000-4,000/mi 2 "
E 4,000-10,000/mi 2 8 5 t Schematic Sector / Zone Diagram for Inflian Point I Figure A.1 EPZ with 1980 Population Density Distribution L eium.
o SAMPLE SCENARIOS FOR THE EVALUATION OF SIREN ALERTING AT INDIAN POINT ( . Scenario Summer Summer Winter Winter
~
Season Time of Day Weekday Late Weekday Evening Late - Mternoon Night (Rush Hour) Night Stormy
~
General Warm, Clear to Warm, Clear to Cold,
. Partly Cloudy Weather Partly Cloudy overcast Home/ Vehicle Open Open Closed Closed Windows (& Storms) (& Storms)
Temperature, F 80 70 30 30 Relative Humidity, % 65% 80% 70% 90% Temperature Gradient, O F/100 feet -1 +0.5 -0.5 -0.5 (meas. heights =95' &17') ~ Wind Direction: i . General SSE NNE NW , SE Val 1 eys . Up-Valley Down-Valley - - Wind Speed, mph 10 mph 6 mph 10 mph 15 mph (meas. height =100') . Percent of People Located Outdoors 20% 55 - In Motor Vehicles 6 15 25 - Indoors at Work-Commercial 23 3 - k% Industrial 7 1 - 1 _ In Home Sleeping k 95 - 95 In Home Radio /TV 20 - 14 - in Home Noisy - - 3 - In Home Active 6 - 35 - In Home Isolated 4 - ik - 10 h In Home Quiet - -
-A t
i ... __ . . _ _ _ _ . ~ , _ _ . . . _ . . _ _ _ . . . ~ . - . _
0 9 0 s
- I I' .
L f 6
*e G n..
l t i! - ) I I l APPENDIX B h 5 4 duumma a f 8 9 h ) w I t l t 4 t . t l . l
+m t
m,
'el S.E m I
i 4 I e . w' W t w
/
1 SIREN LOCATIONS FOR THE INDIANT POINT EPZ ._. This appendix provides existing and proposed siren locations
~
for the Indian Point EPZ as of 25 August 1981. Siren locations are provided on a set of topographical maps (Figures B-2 through B-12). Figure B-1 shows the relationship of the individual maps -- to the Indian Point EPZ.
~
A total of 88 125 dBC sirens are employed of which 12 are in
~
Orange County (#1-412), 24 are in Rockland County (#13-#36), 43 are in Westchester County (#37-#79) and 9 are in Putnam County - (#80-#88). Table B-1 provides a guide for locating the sirens on the topographical maps. _
=
M h w w w Giles emmm8 w mea w
,y -n-.- - , , . - , - , , , , . .g -
-__-,y, , . - - , - - _ _ _ . - . , _ . . . . . _ _ -
TABLE B-1: SIREN LOCATIO': RY M s.o
- Siren ~m Map = Siren =
- ta n =
Stron = Mae = B-11
- 1 B- 3 j 31 ,61 n- 8 l- 2 B- 2 32 B-11 62 n- 3 3 B- 6 I 33 D-11 .63 B- ?
4 B- 6 34 B-11 64 H-2 B- 6 35 0-11' 65 n7 6 B- -6 - 36 B-11 66' B- 7 7 B- 6 37 B-12 f~ 67 B- 7 8 B- 6 38 B-12 68 B- ' 9 B- 6 39 B-12 '69 B- 7 10 B- 5 40 B-12 70 B- 7 l -. B-12 71 B- 7 .l 11 B- 7 ^ 41 i 12 B- 7 42 B-12 72 B- 7
! '13 B- 7 43. B-12 73 B- 7
,- 14 B- 7 '44 B-12 74 B- 8
! 15 B-10 45 B-12 75 B- 8
.i 16 B-10 46 B-12 76 B- 8
, l- 17 B-10 47 B-12 77 B- 8
' 18 -
B-10 48 B-11 78 B- 8
'19 B-10 49 B-11 79 B- 8 i 20. B-10 50 B-11 80 B- 8 l [t . - -
21 B-10 51 B-11 : 81 B- 8 i 22 - B-10 52 B-11 j 82 B- 7 23 B-10 53 B-11 - 83 B7 l- 24 B-11 54. B-11 ! 84 B- 7 25 B-11 55 B-11 85 B- 3 [ 86 B-
.j _. '26 B-11 56 B-12 3
- . ' 27 . B-11 57 B-12 87 B- 7 ji 28 B-11 58 B- 8 88 B- 4 g ~ 29 B-11 59 B- 8
- ' 30 B-11 60 B- 8 1
t is e 0 w- { s' emmme 6 h l I.~.------._ _ _ . , , , _ _ _ _ _ , _ _ _ _ . _ , _ _ _ . , _ _ _ _ , . , , _ _ . _ _ _ . _ _ _ , , _ _ _ _ . , , , . , , _ _ _ _ ,, _ _ , _ _
FIGtJRE 9-1 : SIREN LAYOUT MAP LOCATOR j Figure B-2 Figure B-3 Figure B-4 Corn Wall West Point Os,cwana Lake
~ /-
% o.4 Figure B-S T qure B-6 Figure B-7
\D Figt e B-8 Poplopen kill Mohegan ake -
Lake IN DI A N \e ' PolNT
+. lhe t J .-
Figure B- Figure - 10 Figure B-11 [FigureB.2 Sloatsburg Thiell Haverstraw .Ossinin r A S p. 9@ l e e er--w----e *,w------*+-mww-*wa- o w w-at-mei---e--er-----r= -r-w- --ae-e- ,w .e -we. r-- .*-,..me----+c--- ,ww-b -- w ,.e-e----w- e-- w-----wv -e,
i
*?
t',...'.,.,",/,,.?.',,,,..<.
<w . FIGURE 8 c ...-.-"t"*"*
. . - /
~ ~A *""
#Grsm,~g g iryg
~
jf
; - ; rc s.v.a. ', , . .;.;,'1 . .., . %.*. .*uT". * ./ Ajjf~MWfs.4;};;"'.'r_ :.. - ,"w.mv - MT. ~
g y .Y.' .:; g' , w
~
-[: ?.c i 4 ~\ ~ t' ; .gs m-1 7%~e -
p .;'c N ~ .7*p -
- t. . ' ' 5.~ .-
i d .. O.'.; M; t\,*q g
* ^ * ' ' . . , , T.' S'*j2 ,
~t/ 3
[. s' k; '.3
&f.o'>j lg, , f m .
c;Mia & . q.( s 1
,. k p ;,f,,,7.s: .p ^w W i h- %ger it i,. .
. ~
sv g(9_.f . \, 4 x ,\ -1,*d . .c. :
?
y .g e
~
-g. . -
l
. v l. .. o g-.. pnj -
t' { t_ 1
'? " '
N ! ff,f, ._. .?. a;; g' g4a.Wh'f.. jsf i L i m' r..M:. c. 2. .4 s .
.M.,
.ls
%m%. 4. s h_
n s g x
$w%e t .F ~ gw.
s,
.g y# .
2.-,_s 1_ 2 . . a.[f i;.. f'v
- 3. 4i, ah'. . Y J ; '<. n.
'j, p j Wi -
' SQ
. e M(e{. 1 .s. 0-T }$. $ @s-% P..o7 Qg' j
P 9 P 6 4 %.. A %G6 '9f.1 a9 P i f x .O . M 6 M F Y. , M-n# f %,f.1 %a j.g.%s y&ag$f
~ ... c
; y ... u. y n .,.s >. p g.. .
.1 . . n
, .n u . .:, ~ , ..
.. W 2 '.
s4"* 1hn.;~sk'
^*
f MyLPia Q.U@ l . l f kfg}fhWQ;,} &y$- - e , y. was ~ we..a:/v
,q g e . . 7 ,,
f,., n, ,. J. a
.~.y
~
p ,p;,y.,. m .m
.i
+
, py'~iz ;.., 9 ,
u-3.y ,7.,M.a. E b,y D.u,7$ N'y
~
p,D. p[, -/.$,,.7 x. . m. a., ' . c, gip' %a ~ f v8 ,, w. w .s.. f*?(
*\ l
~I*
**
- t j .:
f .
]. . AY$$ bh".'N'?h: - Ea 'l
.4 T .L.
h.$h,
**T' r** e5*
'? '
*bNY:,, .* .'. . ***
. :, . =,,
* ~"~. *' Y Ii T.EU$ E . *~'
~A$ . . . . . .
2 = e. . ._ f 2h~f.T.:.G .. ;. i~ .'"; :::- =:-
. :_-- ; - ~ ~ l
..- - . - -~
. ~~.....~.u-~..
b*b d ~ '.
; ,-- - -- ,~
.= J.:t.: - *:- .. 2?_:.. .
.v- -- - -
t
. , . . . ~' :: :.:~~-- .,.
<*. . =;. ;,;;;.J,..*.. . .~ ._.
- 2. -- "-:: -"" . r ..-.; '
. FIGURE B-3_, ,-.;zy.3 .. . ' 4~;-m ;, ' r ~.c 21' a ,. ;.- ' ;. .
3 u r ; .s, .. .. w w.-: . ,rct -
~
' ' Y s ,
'?
, j .Q $ ( y "l"l'f.h' g v ew . t ._. r i
l_ . j
.. i , . .
. #g. . ,
- =x 4 ia w% _
~Y
; . ]f. ~ -
$$ g?. $_ Q]Op;VST .
. . j; $. . JM ;
ls. . $f
~l ' -
.; f
.=. : .
t'
-l ~j n,A w,x as..; : - g x K n P. I.'
i M%M': W 9' q?; c
, L., .
- )m.,. fj
,',w.
- 4 ffe y ;
N. j'g [ [ i ,' -
+ , s
; %e , u, A .
.~ .+ .
P pc- 7 x_ (. s .,
.,.j..
(y \ V . - .-
.e,
,F .,qx -
\wg ., b.6 % ,,.
,e.. .T.
y t t ., ,a.
?y, . . .... 42
.j
.r' .- .. n. .
c
\ _) , sf g
^
V'
.s fy .
( di 'T/, ,7 , p . , .
\k .
, y.
c ; F.
. < e4M. , , .c -d %: ~
..c
- e }.
. . e. _
, y rv ,- s.
. 3,. ,3. , .e c.
.y.m ~.
- y. -
. ~, f , _
4 tr .- . .
> t. . .
y y , 4 a .a.c.Ae..
,a I * #[.
9 '"
, ,t . 6 .j ('
; > pr .
j, .> ./ '.,
. [$. 3 y . .
% }' f ,f r L lm
'Q_3g'u>5L/.c- .
5" 'g..
'bl 5 ,-
r'. ..
~
g%p'
.,,, .a
" .2_4,. I a}s$.p6:,.a
.6 y ' ~., A &gQ u -,.
- n.
g._, _ . -..- 1, -- _
, . . _ . . ~ -
-q _-j _ pn:.
T =_. s :;
=.:. - * '
_ " " ~""'~~~~ L .
~-
g
._.___-- .. s : =, . .. .- =___= .s 5_~ :5Ed?-- 2 ,,~ - _=dhik_='.w . - ..
'*"'. , ~.'.,
. _ _ _ . _--.__ i - -- ... ..
~ " " "
E__ __ - ;_ ._:EEEEs=. .'.T. __ _ _ .: . .
4 g...==....- ricuat e-4 ~.:"J:.@-;':. t "N5fh..Y'.!_)...' Y [j. u.f.dM,l guf.F .4>7. d.. !(/, i- . .. i
.f h
;z
[ g . . . q - j ELY s. \'
'..f ll:f.R
')', ~
-E,[
,f,'- .
?* {. ;"
,, a.m x a. ,p.y' m s A.r.t 4r 2 7 ,.. ,
t 4 . y$ f.l#((_jijj ;)? 7:-
.!l tu + . < , <-
wl y m .
- s. s'. ,
. c.g5
.e t
-y , pq f
'S . 1?r,,,,f g >. ,{ '; ;
,;s
.,,hi.. n' '
3 .- . ,
't - c , .jo FN. ,f'Yv,8N Q i
l l" _ kh$$_$e$_ $_${ J E = _=~-_ ~, . _
- r
-. g_= %.
-- is~F"" 5~~' "*~j... -- = = =- ..;. 3,
; - EZ ""
._-ss.=. s 5se._=. E .-._
- - ' ~
.'_'.'s . _".
L i i
r r
. "";-",'f, ;,*,,",','"" FIGURE B-$
-.-. ,. - 2.p , ,-g +,, ~%. , ,-. , . ; . ,./.; ; .
,,* gf' ~ y* .- -' l '
, ;f..
',' '.. a.". ,'L L. *$~~ .#
,h, . .cjQ, WG ' * ;! c I..,9y7
. , ff ~~, .. , A. , . f , ' *,g'~ , ,,.i,Q-
;, 'y'h
,9'. .
is ***f's . Jf
$,l .
- fJ' y i '.)h,,W,"' r',I , Qp.k~,'5x /, '*
y' , . 6 t . .', a , .L. , < ,r/,.-
,r ; *I s s l5 ~ &' . . : . hi I;'s_*
s ?5, . .c ' vi h 'kh.:p. $*h.- s. yr a Afy - I'
- m. .v Mm'$;;hhh.,
- g. L.
- h. h rs " ' ;'.qj1,g\, ,, f.5
...gG gA - fir ^
,/ ~ . %g$,, lofit k..M+i.h .4j ,.!_y
- g. ,,,e .y4 f b g p n.
v j y \e (. 7p4([;.'A
.1, , .- ... _ . . , .
- t .. .
w: .v N.~,.. ; 3 j - , A,r '. ~w a"e r t- [ . . . f, i ~ 7. / ~:g- Q -&p i
, ,.,p/. g f- ., .. , .
a c 1-- .' 4;, r-
- r. / i5 u . .#. 4h:T,
.. w,g- ~n.cm- dn. nri f.' . ,;.
YI y>a~.a,p.6
't s*',.
f )$' h),g%~y f. ,trA'en,kS W -w&, . .y i s ; Y , f.n* ik)~ $,. ..f-
~
, .dll, ., q . .a f 1 rd { ' gi . '
?
,s; + >. -W+. \'
s
~
l .s. ~
.it,..sy
; e .._ . p
. ?.hfth'Y dk.h;,.%akb$~% ., ,e '. W' N .
< Al.
l"
- h '
Y
%q, Q gw. k . % T M p e g gk e% M O V [(F4 p u ;.u m
jg, y,m. 5>awe
. e h %..y. .sw.3 L~<.~a-- -. .
. w..s
.[) .- ji! .a k' ng A. ! '" u, . n
' L k thik ,@ 4 3 h S .
~
kn 2 wyes,n. / s.I ' ,
,A kk b.h { l?
w e:b
.s.
i
, -y g.j}l,,gg ~ny W_hdg. #
I h.m " _ . .3#b$_c.Vu. Dr. 'q u.. d. e 1. r
~n
..f71 s, m.
g k_NM,,, ,e s,- a
-. '.. s? .M 2 s.
1 M . ' ' .'. { ' , ,. p(q N>; m. s, # ff
~ ,
y ~ $.. Q.:}c}4ffis' . k',' - l l
,.4 - - . . - . - g ,. ,.. T ?.a
~~__~..
i'.~.--.'".~. - ,w. --:
"'~ "'
_._.-_._..,mt,__ _-. ._ 1: .
.2,________.__ _ -.~.-
ie
......m..
. . . . .. .. .. .u,.,... . . .. - FIGURE B_6 -.... _ . ~ '
~. ..= , .-. .
)f , " ~~~
l ~ . . p. 3. -- . g.:r -=y .. ,. ,
. r.l. y,"'s",'.;i ', , i ./ J /:
.1 l- % ' '.,, n l. ;v % ; .t "V i / :/, .-
-- ,J 4
,: g'. ] ;i, t,s*
u e b ., y . . Q,' U.. M, /F ,.a,t'. .M 2. , %[2 "II a[,r. . [<.. ; .- 1;.j ' / e#' , ,. 3 L .g l
> -.,.7p <r 7 . e ,, l
;.g.g .*. 4, h ,mu . :w M .> j', g- 3,.f~.f* ./
x ,,r/.?
- u ,-t - . . n.I -; ,c.1 - ,i t w.
. ,o ,.,. .
/;
.4, .i: .. , p ~ n . K,' n
,.- t
,..(. %: :a.~ .A
.Q ',> } r,
.w .
- f. v <
y ;y.J*.* : ,e Q~,s p.
,..-q s *j r._ ; ,: u. p i n, . n' <'
==
,r'N
? ;;6, cig/ ' . <h - -
, [sQ . /' y y' ", , , 'f
. ,[? ?
/ /--
+-
" . / . u-l J-]:;s "' .,p[; ' y:.%. . .',Q".VJ' /
g . , :.. . A
. */ - , f~ ., :
, ' ,,o :;
yd, w . 'Yu!< ;, , . - '
/. . .
l~ , l'
- ' , ..' .. , .;l~ ;
n .. Q ; ?;
- i ~y;s V *
. ,i , ,i i}r,.. $. *1 < f. ,,,. x u.... s 4 - )
,(,'
o s- , ,, ,- , (l , r; , R< . , Y, .sra;.l
; v.. #>V s , i .
i y
, l.f. ,,- t f,
**s-?
n., '. i<
- i /.
.a
?.' ; , . . m
~ :
.n,
~sy
. \'. L.n . *x .- - N.
<;. a,: - . ..
,f. . ,7.
, v.- . . ; ;9.., .,;
.. .. .w.. ,w
;. -p
, r. . ,, . ,.s/, . ,j r
.)
h .k h. # , ei ' f ,j lr'.I,.J', # .
. , . * < ' I'/
M g &> . ix
^^ ; ; &yl'"'...M ..v ' . ;h, 7,.
*i G. c a ,- , ,s n f;:. . i z. ., .L .;q
;.. . . , r - * .,.
Y , , ', '
, g: n.~ ' % f .m n .', n &. 'J.y,,; l
' ' ' s
< ./
- m jfe L. 4.y! .r <- .- .
6h . ;u L;! f ,
..A
.p C ( =-' < l}.
p.tk.a p pp%e.f[:;
; n x s
. ., Q .
.,.cadFe p. 1. pg Y.w,,.q\.7.
e
- 9.n.&n et - . n, , ,c <
< ~ v.,sg. ,.. s.
I. . a.
.g. .:g-l . 2:.
-- r.s f e ~ ,,. n - . ,. .. . .* - h- L3 i p'.,,e
+r 5 -\~
[ , y .: - - pr e re a ~
.;A . . lb. D
&l. .
V8 : . !, {
. Q, ,&y{e. .
2}. ?. A.. ;F '$4);~{.",*:vY-f Y,Y. k ,
. - 'y i
n n . ., n x,f,4'..x.k. i $E
. < ..,g:t.
3.sm.,, ,. .; .x
- .t-
. c:'
a yw s. p W ti )? .y Ar
~ ' p ,?: . . S u, Q.' ,\ W .
. 9 lM..-Q .L. .; %-a. .-
~
'w' -m-A Q * & s ,ts
^ q. $-1 .
~\t.,\'.!g %
v.-r& ' . , .? *
..J.Q&'C*T .;t-
, ~*~ '- vs , *':.- r j g & 2~A G.,x . s l
L.
~
. , . l. K'.yi}: s.. . l m* , ., .us...
~
. A G &I uN > ~ W.<. ' ! ~'~n=
,l w..&,aM~ 4. 8<
T ( r. v v i. v *. - . ../. %yg6v. 'in '1
..- ~ f. km,L.,v. e.L, . n,. Wh. e.fMV>,Nd> / ,S,,..
.,s . - .
s ,.
? $
*A # . .%
&.% , r.
. r.- 3
.'Ol-t r .1 #
1 /: 4 3 - ' I (
- w&W,; ..p %,?.
;.N; :*;.~& f;f,.'&': .g.?.. . f* g.r,.A ; y . , ;y L- t
,- e ago.c
,~. .. ,...-p....
sn. - G.. r . % ., ,
, , ~. - n.2, :
s._ . '.
' ,,I e..p,+,. . a/ + :..r.
b b;;f- ,.
- 2-i s
m..n y ' ',92 Jgp f- [ ' n. u.~ 4 -
~' y ~ 3, n' M..&y:l.ye/. 3. 7 8; 9 ' r.::. _ ':
x' .. n y 9 ;.. ; A.<.y,,j ; p g-
'*w
**- %* . . . .o,.,
< ~s.
t )*. .. . ,,, c( ~ e.a .',. s f .- . 4 ,. N' . D G,.)K a , s/ ' t.# k
.a -t? <.
;;- l' f ,. .I".) W l .3 *q -
Y , U . m6 v V N;p
%klp.;..'f ^.:l+m > m .. f. , % .:. .'*yjg.f ;
$.. .s.,. c ..
^ , . '
,?+.2,\ -
.~
W' .
,..ey, , f :.o . N -,, ' . P4 _.3 '.=f4 -..s"r,
x .\ _ . c.e. w ;. 1, b <, o. y s.. - . . w .s . . '
,, 'pA .. W _ ' .1< h,,3.MED. ..a g'ld ( "..' ~i. . k. g'ekA ~'.',;f .
- [ f. *~;j.
. w . A* *--.
==**-
T
.. y
,' a..-
.-..~~_x._..-
- n. _m -. - a -1 ~.g- - _ . . .
.x
. // .- K~
~ ~* ".E. .;.- ;. " * * '..' . ~ ~ ~ ~ " .
4.h g;. I~ .,:7: .".'0* -
- j --
. .: _ . - - - - ~ _- . ~ _ __ y.m.
-- - . ~.,.;. :- ,
v = - - - - - . - - ...,,..,_ _ _. ,
.e.
I 1
, .=
4
...........,a e
., . u . . . t.
FlCURE b- 7
* ' " :, . '. . 't . ,' " *
. . ~ - .: . -. ._ . .! .:.. - .. -.
i
,j .',' ;
-{.--.
[ M p'v gdW ' e ac y
+-
,, j p
.,'1
,/. G . ' i ' *,.r
,t l - '., +
a j.* _ .?hsl:::<D y
! ' . '-.,i .fig. %.
. -~ *
' s:. i,."~
' ' '[. , , g : %' .-
' i*
- ,; -q ' ," l~.{87.'-
.c :
~ ~
m<!/ e,..i? w g . . ' 7I ; ,. =, G U ,.4.. .. ~: . . . :, ,p . (, "o k
%,y ,
j('t "Wr .h
, i' .
.rh p&r
4
. .Y.,g.Q y A' L-
. ,s
- ~}fc- g- .
t,' g
. A ., ,l-lm(
}
- L.j'
- s. ).Y htJ' . ),
- r. .. ,
- Y %,
,'1' d
* ' ' .R;ff d z,, l.
klf) '; . ,
,Jk.a
.a , ' ' (h' . M.
n=~j!,}.G ,,-lp- a ,n ~.
,',, ;r . .
4's.
. - '?( , ss ) * ~
- v.
l f y W3t; a ,_ . n . . . .w. a..a [/ , ;*j';' ,;p~ . .',-
..v ' *
. . . s. , ~
' :s s v.e e- ~ . .e
(, w,. g f.
~
y[c k#' i y , P ' : k. 1 -'V X'/ b $ f./S
/
,t,qhf" . il Y k' .
N.XI)i . k-(i.: e i
~'
I Y
- r. \,'y,y ,. . t. !
.N. & ? . , . .&
W .I QY /$h id
' :- w . E"S*-- h[k M*
- p. b.,.. . h%1'.'2 E /,' .
.g.-
,)%%
hp2i IIn:7
~
h [ ,f ., 1
.. k w'mvs.
. .m --'l w .F i s,' .;))p - ' i g' 9
. ,j.$$'-{$,f:~.i.-;' i .
~'
,W .
. x .
.,..t e %y' X f N. ' J1,;,; ,
e ,f D .
' [. N .. a , - .
, . . g, - 8 i Y[;d y - ] <t) ,-< [
= v .t.
~~
3 f.dg] . -. . Vp f -
-( i ' *
,4. i . l,('g, f -.Q '
- y"1 .
i ; kt " 1-
]h' :14. s- .. , , . -
1 - g s rg,y - g p, 4 T. f',; M.t.-N. q;)f y .,g., ea -
. g= --y . .. - ,,
..'$ )
L y .Ag. s,
.s ur: :- .
*~
(. '
-s '
@=f _- p
, ~ ~
M
#~~
'ig i:jfg, i
' ' " .~
^^'
,^^ ' '
/~ x;;
d
,,4, 4 p,
.y .e L . N 1 ( .. .
e
' v.
.. r
.*a y
. .L .r a ",
~ ',c. '
r <- . (%,. .u l L.
. ,' .if- - -%..
r .J$ ..A r.
/, ~m *' 1;, . ! ,,,,L - - ,sa,',,,,; ;
'o'h' ~yh ,Mjs.y
_/ // /
^
' " * * *?-
'. .v.ejw)w*', ei -.,.Y s'
~ /l -
r ., e ( ;
;*4 :., ,
. 't prM. 1 * .h,). ' ~
'pO. ' ,
s lC'"7 {WQ 's L u. ? r '
! ~
.vs~ ' ,'v. r js;r*h
" -r
[O h , g , ,,
-U ; *hy n,. ag n,- v
&en, \ q ' ./
j .N g ,y 9,;'}y.
._ . y g,.t'.'ai j -
q.':.> , , yh.'.,'yr, q[, i'tta
* .' N
/,,jf4 - .
P v1~
.n
., #&% I
- t.a . s w,-, L N
.. u g ,,g. t
,&V Y [. ,pQ b'. T', ., J, '
- n- %.s3 Mj, -y*. (? $'
- M,5 f.
.x , ,
~ ~ *
a i
~. ?
R,&,-w
, xcy , ,_.;.. . : - - - .- .
&, ;*.~~.~~.,_,,_.-
".- ' l:tt .-- '
;-- a ~ - ~ ~ ~ 'k, 1
G ~ ::... . . --- ~ ...._.
.. - m,.
~ ~
** ~ n-::--- - * " ' " - -- . . .
- y,eaeans f.,i,
. - se .,Nr. .r,- ,
~1
*5O. bE..-5:"d-1.
~~**~";*~~"*
t
- i. .
v
......... --.w g a.,y- f
..% **1"[C.,*/- FIGURE 8-8 . :s f.- . .:.: - <
- ~ -
- - q - = .. . :. .-
y*':r; . .c r,7 . g;w v;;;.5 %-:1.a f n:W :., -'s % 3--L3 ,.r p ..
-..:;y){W.:..".f.f@f,/'{'frG'qqy}'Ig?pQip'rSjJ'!
.:/ !
i., n' .t@.,]:.. 4 .g '
- b. .:n* i j .
y..
,.v.p; t \ :, },w n:<., ,q.
,. ;s j :. .::,
': Shp
;, bx . +Jr . A .... . . f 1 o - q ". g .r g
'i, ]d 7u,'s+g (* ' [/ .y(r}.b, ' t.wa' .
- f *y j[f..A-. . A W . N .h h h ' (,-
.I i . h h @Q,hff " k ,3 h' ,
~
k 'f'].iba;rp,g{k;..k.. h. h{.&j.!W) .. , i
~ '
~t ', -
.. 2 ', ,.3 ?.;. -
ft
-hQ-)ff'n Q 1. ' "$'[,.hN. 4,.1%gk3:.h_&,h
% .1, j. r .. .l .Y M 9 '
-?h%
,i. (h b" -
kw}. n& ,~; .f ;f C. jp >;k: . -]. . . gs <>. c.; e=- p W,'Y ' ,,J$, Q. . N'aj..
.: ::: A l-
_ y.Jf'. &f. T' g> p w
,y %.2 b
*?
. yql.
,' J yv f,.s ., ,. , .. }.
. . g t?:.V p{Q. ... 3 .,g w.A..
, '-:.T ' r,>-- .
w i, :'n',. :. u - _. m. %. .. .
. , . . - a .. P. \
~.% g jziu .n ,c,g.n;.* i . % .:--( f. %a ,% -r
+.' ' . %
g:P f. ' M\.V.. k1'j
,p " / ' ff!~ y.ya'.* .ss vifyN@x:
- h. q 7.,L. <,1,1'.,'
i,
,,W.
e.c.
'n' $ -
$. 4
-Qp a y. v. gi-
.g eO .
-< -at kW;.w,.e
, :s;-
n
- s<
vs,2 ' , {g,/%y,;..
;y ,,
g A A b, ,,;'F t . 4 q. y . , . .s . , 4- ;. y' x 1L l;i . W 1 ? gi n y<< B ~~ - t, . J q
- . .c
[ h'. V . f. % ,)_ % ' p;
.n i ., s . .- r. : .. ; -f :
L '
% %g .,s d F -
3 d, . ~ ' , 5 (. aMj. g s .~Q M+ . r ; l, ~. . .
.- .:n ;n y[ t..
- g. .
- ,+ h,, . . . .. a 3 ,a ih ff _f,mn., ?h}3 k. .'b L '
- j. 3 -l
.. er,.-r-
-- $; > Q.ws %
c ,.. s,.. T., au 6.'/-
+ -
~ .
7 4 .h. t. .. > go - . (d - [.y [ f ;. L:^ .' ~ a.%G r - 'r p ,d
.e t = ;. ; q.m-9w
~'. -- iQ. rf. L;.[p,P4 g:o My.f :.g y.
s,.
.p. - s },y},.ng .,h} . -
,g -
.,-h , ,M 4, .,
. . ' i.l g ,
- y. 3-g e$,-i-om..u.f t -
1
~h 4 $ tthe . .:, yA, -l}u . 2 ,.y; ~ . .' -
s-m ~
,9- y ,$g,.
p f m:.,el r Ai,p) gW ,y:.9 3.m ,, , 1.f "W,.f..#,,
,, , .'f 4 i p. 'h, .- 4 L
-.l.y-& I
,,c.?.I Y l.,k.,' nh. gf. y ' k, X : fef .-'
- 5,$:.f ., -
p c ..,. n ... +
. f a:i. vi , . " .*1.,&s[4 < 49 .
. . l' Y..k~ ,. y!
. Q' "- ;~:'3 l l~'
j , i_ ;' 3
' ' i"' ' Q d! .i' (-
*?&,MjQE .f.f5
' f.f)I
^
~
- h, ,f f. X e
-s ** Q.... .
,- , _ ~ -. - -
\ . . l ,.. ..
~
y;. 2'.-3.~~: T.l ;; .;._;_,.n-- = = = . --- - 1 si - --- - u . n -- I __. -:=====- .. 2,m _ - _ .
~=...:--
, .,. _- ; .--_-- ; ; 2.my
^
. .E.2?L:.E-E5= ~:1 - - - - -
. . . . . . . ~ n'.: r---
. . . . , , . . . , , FIGURE B-9 -- --
y -j l- 4 AC" is ~n'%!!"*iV*'hf Y. 'p!A:~~ ?$ , Y'~.lT2:*,- ~.i' 'i.
- r. .pM .+ "X
.i-: t.:.?. ~
~)c w-t ' .~ :
U'4. tD, s. V:L ; A r-Om'il%{&p+6;Af t **
> ,* -}
afiy}.= Mi . d. t
' % :B'k2% I' .
'I *-
- ,/ s
' ,'5'.)(w~&Qj ~. ;' W v *- ij MX,y , *.T.
N V ,,, y;'bo :
- j &y. # * % :: % nf ,A Ifr
(!11['p\;ry8. i 'i .I: y & , 3 p.s &...' g4.,', 9 -- VMa l
- e;p .o 1.j?,:.?
~.s - ) i . *: . 4 '
x:.ea .. i
~
.'. : 5. $bh ? o
$s/l wk v.1, f$J' t *fi.' #6km w 1m$fl[$('5h.
7 " at ge. /t hgir
& & f;h
.7; Y pay.~[j D
j f w- -
.* 4,,'%
MQn GM 6 t 3,3 * - d
/ .;~,tcpy . n y r:
i-k /.,h.)4
- [ -h ; . 4, /,lL~
kgk7h.,h,(h.,Pg.x',Y..'/""7'N-- i
,.h ht'1 'h.e;l,:j -*
'd 4 C >h 5, 1: ,A,.%g ys.&
v,.c.
..a.
, pc- "
[
?
Q .',ulQ i s ': 9
, ./ l 4. .
w y .
- r. . ,. . ,;'oe ',s * ;:~%' (?! !
a. f - i fmuj .&fif, fq ', .f k I.$%lf%, iy Sfpd4
",Y / *5t 6 i
~ M 4 % %0
.f}f'Y
~5.")
,5flf,R
, Y? .] h~ ' k ?
yn4,.PWWW
- , n;ea%m m#f .
~ Qs
') f,I i, 1 j *i[ y f
8 qt9 w. %, -n.ng pp
! . , . - 1. ,le 4,1 ..
- 1. .
=- 4 ag .
9 . Qi ei r.Jf ; v p
- q;:t.p'fL y,,3.. 4 , . .d.M .:-u . ; W .1 d i
.V ' Y+.
s.
@Y . ' .:. o 1 ' -
.j,.
h, t ,ffi :k
,3p.
) nr #?f. . K".. f.,f . s,.ch, y..A't r
r:;- 6.fY. .1. y. FI,-
+....
.4 Q,%.:p
-~f . ,, f.
. -< b,
' f,c ,%*>4
-'* a .. A..';
6<-E"(f, s.d1.; e I;',.-~ r ce ,1- ".s e->,* a, .r + ;.* - .. . . ,<; w=. yr. s . y^ g ,t
. .}
yg.. e..b .,
;p';,0. ,,N T*C;)'j- '7p**;54
. - (k* ... ~
. ,.g, ., m' j '/ g / ? . t t
;.3,. " d
. NTWM )
! i-
-\**'A. 'l.$% -
]j'%'. *s. y-'$-?. ' i* 'I'..".*.*.
?2[.b
.~% a l
, .y -fIf f': .Q.. w- _f. . ')~~j~ g*Q ' ( -~ . ;h%h
'M. '. * , . , ,q.; *. r g s.
l' ,f ". jW ? %.* ,yy , % .*~l f $ f,,' f_ . , j ; , . . . ..e .- Q l
~~), 1. N ;(f , O*I;0$ D ' ~ D,!I$$l. v .,--
%%1 ,} /l,b, p.f. t}jA E*r;"3if '%tV ,
?.- . . ,
~' C t )$ h
. a -
-u.a: w .---
.~ --
.. . . . . . . - - - - = a. .
A'"".-":~."-~.'"'~~.".'""'
<' ---- .. .- : . ~
-- ** :_ '. , ; .L:~=- ~ ~L 5'= ~~
\
y __.~33 .=.~z. ...a... ~ . . .
;, ; - - ~ - a i ***"*"':~.*.
. . . . *'==l**= " **
~ ~ - - - - ~ = * ~ " -
~ ' ,
. - . . - - - . . ~ . . - . . . .
h
*6 -
1 l 9 - u .
. . ..= . :.~:.:....
-4
. .. . FIGURE 8-10 ~"**'"*".. .
' n .
w :? . . .
,% , -aq,c
. a. .
'g , -
,i N
* ~ ~~ L. ..
s
} ,---.% /, x .ti ~ "
' ,r
>~ '
- s
~.y, :
' :- Q'. ,
' x..*:. : * ? t> , ,,
wiy:.l_ K ~> ~ra #j.h
- Y[,,h .,. .'
.y
, h[/qu nf,,f .,, 39 f \ *
, -%- =; d' : g,
;~. 7 ~ J k ,_ '_ . ~ ~ ,_
~ .
-,,. K ' ' -- +? <a n.'.
6 i W,:c, . g , V)
.\
.,, g' . . . w,== ...-g.,,:~:.kc: ,
.- .-. ~ '
y . 7,4 . / .p . , ~ -.-w / v 3 f('.i , 3
, ~ '
l f . S QM [ ' _ .: g.,
$ q- t, Q
Jj- 16
*,,a ,,'
- - : ,, >fy p. ^4)r,jej,,W ' Q, p
j' a> C.. ** %
'y
's.'
4 .Qg. ' ,.*%fm " f.k'
.y
;l L.'1 l.. &l m -l .. .
. P, s
,A, " s
, L .-' \
,^, 's .f. - -
,. ' .9.f. ' , f m. -
- C'. :
n g
' . > (. .g..st,,,'f
.i .
~
$,J,* f '* h> e.,7 , if ll '.,
- s T-
-- < or ..:::::- , g. , t,, '
g ,, t ,y f/ ,, . .- .
' j g %C'i's
,s J. , - 4 /.'.d. ( "e ,
" # j,7 ;
' ' '~, I { .'
- j. . i . s.
- d.,
, 3 v *g-
-&g lw: bF *' 7h~g &;.; m." .~
'{ $,. h r d
- ' ,. GT .
,s s A . ' ,. . .
9 g o- , ,. p,, .s. .
~ ~
b- ,, ,
- f* . - .*
'a pg f 'M%
? g,{
.:f:$Q1f #4 - .
;'EG Q' ' w ,
' '% ~ r($f Ua.;,$w;~p&
??lc Wf f
.P Q,0.'p"'. .r v ~
t,* f gQ;%Q.'; nr Q:.,%,q
=dj'",*~f.,.;,&e,
,. . Yl;, _
l {!: .6 h.,5 - .- ,Y; l en, N
- h0
'",, r.
l h -
** == , * } b
[# .
\J
= =$b E - lI' } $ l' ,'
s l< Qe*:j ' &;Q }'j ,' ,h* , 4 4 ~ *' , 4.yg. i), ' V
,s e ' f:
e%, f { ;'. $}(%$, (f _ = a . ;3 Wq , .
.u , '* .1
' * 'y .
' ' * ' b .,
f ,, ', 7 / ., N
*~' .* 'Y , t=ll '%
V%
. % f., '
[ M $$@$$2%d $js' jyVEsrhh.,i b d a$d,r M !.
- - ~
T,, *
* : ,, ,,,2.pib'.: 2 5 # .,
g3'.*.*.*~~.0".*'**.*J_"",.*.*.'--
- c .,
.- --.'.a...'
*%. 4.. v
.s -
-r.,~--
~; ,-- - ,,
....~ s.
's-
-Lo. .v.
T. :~~. 7:~r
~-._.,;;,,~-..-' * ..~
- .~~.~.~---.-.-
. : -s ,s re.:,- - - : 2, ,.
~ - - - ~ . - - .% **: .
um
.~:?.55=;~-M,_K L~',', _,, '....'
t e.ee
. , - . . _ _ . - _.._,_,_r_____, . . - _
f
- h. <
.. '"l.";,7 l'.'... ... M.. , i .' N , * '
.. ../. FIGURE B-i. . ". . . C....ff.'.' ". ' ; '
~ ' ' , n. ' , ,g ' } L ,
- K. , ..
-}4}
/ 7 ~-MkQ / u
~ g_ v y .\,n s/ :%. q
.g s 3-
.: .: .y
,, . ,.. g 3.. c ,. s
',, b,. ' , ,, - .,g,s ' i 4[, 5.' : a . ,g~ 4,-w;,. . . s -is's1 I r- . -
u, - L. . + - v.. o, .. \ 'r f.. .Qg ., - .- C \* .# l/ f.j 4
- .; ?-
. 4;..
6- s.. [jt. 1 j. i r;j,ij \ . n' x
\
\ =
r *h
,4. - ' y, -~
g s k V s.
>d
, - ~ <
j r, 4,
- i-.\ . u \
,e<4 / - ,
l k, .~ 7 % g ... N & b,fQ, y
- . ' .-'T-g
- s , ,_ .,
/ 4, j [.- s
[4 t, ,,
*'e
.. . L -
+ gy a V p -
u t ,
gq . ,
, 'b M . , .
1 n . , L, d
.- . utt.E -
. O s s w -
x .. . . . 4 r,r:e s a g, id + f: '. . . .
\
- f. ,
4 ,,, 9 . ,e .
- s. m ,
V 0- . ft, ~ ' ' -
'.? '
r,, ' ?.D :.y? __.'s '2 l- . y .-
- 0 .-
(
//
l4
/-
f'-
. , , j / 9 l %,
7 g,. - -
,. 4 .
7,'*9 s L l L.r
- g f s
.4 J
L+. -
' n. . . f,
- 4. , .
q *. - ,
- r -
k,- f ,,5f .
,[ ' _j: . F ,j . .
[ t - , . g. 8 a ei
, y~-
. 4,.
M> %', (l! " M ., v ,
,, l i
*
- M ,
' , e
, : c 4.y ' 't*. .
. [ fl g.Gt r- 3 h - .%
'I
_ . r, -y L L ,
'r v f q 14 1 --' ' - *
, t ,
[ , q,
- l n .e
< m -
? $, $ '
1.'f4 j 3C V'. r r, r-
),> .
o
,' , gl s yt i'
I- ,; ==.= } <-
....c ..
[
, ['" . O ~
.l y o.
L 7, <;. v . . r
~-?.W , &II W. ...l?;&. __.--.-~-;:-.--.w., -
p,= t . --~ _.
- ;_-t..z x c =_.:-.: ~ _ r',~ .%.s t - .r. . ,;--- .. .-. .. . . . - - ,
~' ~~
. .f~iE:'RQ"~_i.:. -_-
Z-.3---
._ :: - . .-.;;=;..' .
--.=.= : ~- ~ w,f w. . =.==: . =u -- -
,_ a . . . . , , ,
=:.: .,
==.-..=.~-
" " " ~ ~ ~ ~ ~ ~ ~ " " * ~ ~ - ' "
~ ~ - - " -
Z.'=-.. -. ---- O L*.""".~""* ~ ~ ~"~ 7. L .
. 1C *.~~"" 1 e
7= h L'.. h 6
...:n = . .g 3;
'
- e"2 =..,.mm . ,
,;.s. '
;; ~= c- (,[^"T~;..a::4,'
i 72 H g&C. jM'G w R- , . b .t h '" % g q. k.k.y ', , g'Afl[ghu1. 1 epj t f.e'.p b g, t 5 F + a g pg : 4 P. ; ;4. :r .
~ ~&, mi p r.Qr;.;g(,/
Mdf * . * ,,h * / 7., ,
~
c-y.fg': ppf ff7: ;f1 W y i .% e ;$i. 3:. ;u
.a,lg . p g.ge.mt 4.
.'$4.:r y 9 L
<u.
~ng* sf &s' . w ((
]::.Qf. -
l5
~
g
. :n ..
.n~
~f .
y a,s. he t
4 .^i.{
l l}3: .a ?
% 1 N $6>y ? n% *' l oll
~
-[,,,, i
- > ,.1 , 4 o
r r --
~ W Y.?[j. ~r ,
, , y l,..f ' '
1 , f ,A*9-L_ 7, , -
's J- i
< ' s..,5Q d m"'x .:m ? m ,
gg -
.J d..a_.sms-.,=x--
i L u<=.;.= s e d e. n .
. =-
7.".,_;-['O'*.*.. ** .
~ ~ -
. <2.n._ -
*~.~.7..-~,,,
a r. . : n.~ . -
==*
C "' L'*~%
, , k"
. : .:: c.-- - - .."'~
": *-. ' "t. *.::= '" ~Yd?.9I E.'.~.
~
E *"'**.7 2 ~ ~ ~ ~ * "O ,.,,.._,]_ _,
, g ;;.g.g,. .
hkkd- . M
* - - - ~ . . . .-..~,;--..
4
- 8 8 .
~e LY h..
bd e i I W m APPENDIX C I I I y. i e L i s t ll . I w e e i e I ~ l e 4 v f t 0
- - - - - . _ _ - . _ ---w,-y- _ _ _. , _ ,_ ___ _ _
~N'- m--w--m. , . _ , __ _
4 ( Estimation of Aatm ~ The speed of sound in air increases with the square root of _ the absolute temperature. When the atmosphere is in motion, the speed of sound is the vector sum of its speed in still air and
~
the wind speed. The temperature and wind in the atmosphere near the ground are almos t never uniform. Hence, atmospheric nonuniformity produces gradients of the speed of sound, and thus - refraction (bending) of sound wave paths. Near the ground, this refraction can have a major effect on the apparent attenuation of _ sound propagated through the atmosphere. For the purpose of this procedure we have assumed a - horizontally stratified atmosphere in which temperature and wind speed vary only with the logarithm of height above the ground. _ During the daytime, temperature normally decreases with height (lapse), so that sound waves from a source near the ground are .
'~
refracted upwards. In the absence of wind, an " acoustic shadow" forms around the source -(Fig. Cla) into which no direct sound - waves can penetrate. Marked 4ttenuations are observed at receiving points well into the shadow zone - it is just as if a solid wall had been built around the source. At nigh t a -- temperature increase with height is common near the ground , ( inve rs ion) and our " barrier" disappears as in Fig . Clb. j Near the ground, wind speed almost always increases with heigh t. Because the speed of sound is the vector sum of its - speed in still air and the wind vector, a shadow zone can form upwind of the source, but is suppressed downwind ( Fig . Clc ) . - _ The combined effects of wind and temperature are usually - such as . to create acoustic shadows upwind of a source, but not downwind. Only under' rare circumstances will a temperature lapse l . J
L_-
~
PATHS OF SOUND WAVES u SOURCE SH ADOW ZONE m
~
- a. TEMPERATURE DECREASING WITH HEIGHT '
Typical Daytime SOURCE
, [fy u
- b. TEMPERATURE INCREASING WITH HEIGHT Typical Nighttime c
WIND DIRECTION i : SOURCE [fr ZONE
- c. WIND SPEED INCREASING WITH HEIGHT ABOVE THE GROUND FIG. Cl. SKETCHES ILLUSTRATING THE EFFECTS OF VERTICAL TEMPERATURE AND WIND GRADIENTS IN FORMING ACOUSTIC SHADOW ZONES.
'l .- w
p a . i 4. be suf ficient to overpower wind ef fects and create a' shadow __ surrounding a source. It is less rare, but still encommon for a surface inversion to be suf ficiently strong to entirely overcome an upwind shadow.
~
The general situation is illustrated in plan view on Fig. _ C2. A shadow boundary; symmetrical about the wind vector, can exist in the upwind direction from a sound source when the ~ vertical wind gradient effect predominates o.ver any effect caused by a temperature inversion. It is likely that no shadow will exist downwind from the source, for the wind gradient will usually overcome the ef fect of any temperature lapse. Along a radius at an angle 4c fr m the wind vector, the shadow boundary ,_ (theoretically) approaches an infinite distance from the source. In the " upwind" sector of Fig. C2, the sound wave paths are
-generally concave upwards, as on the right side of Fig . C1c. In the " downwind" sector, they are generally concave downwards, as __
on- the lef t side of Fig . C1c. In the "crosswind" direction, the sound wave paths are approximately straight lines from the source , to the receiver. For the purposes of this propagation model, we have assumed -- that temperature in the atmosphere, T, is horizontally uniform i and varies with the logarithm of height above the ground, z.* _I T = a inz I T -T aT -
" " Inh, - Inh, inh, - Inh, BT and -- = az_i -
I l -
*This approxim'ation is generally valid close to the during strong surf ace-based temperature inversions.1,73 [ ground except -
u [ J
_. . r .
- l. '
NORTH WIND
; DIRECTION UPWINO N CTog
~
8, t RECEIVER ' SHADOW BOUNDARY i
/
l de O
- y. gg
- i P
- s'
" ECURCE
/ \ '
.- ,/
j
\ s 1:
N _ %No secTon i
~
FIG'. C2. PLAN VIEW OF SOUND PROPAGATION SECTORS, WITH PARAMETERS USED TO DESCRIBE THEM (See Text). I
- a e
1 e I - . _ . , _ _ _ _ . _ - _ _ - _ - _ _ _ _ - - - _ _ . . --. . - _ - -- _ _.-- _ _ - - - - - - - .------_ ----_ _ -- - ----- -
c The speed of sound, c, varies directly ,with the square root _ of the absolute temperature 7 q c=c 0 ## II"* - 1"*o ' =c 1+^ II"* - 1"*0I To 2To . where co is the speed of sound at some reference temperature, T,g - observed at a reference height of z o. Thus, the vertical gradient of the speed of sound due to temperature, a, is: _
-1 c 3,,e az -1 = 1.086az sec -1 in English units (2) _
o Note that a can be positive (inversion) or negative (lapse). - Likewise, we assume that the vertical profile of wind speed, B , varies only with the logarithm of height, z, so that:* V - V 6E Inh - Inh 2 (3) 2 1
. where V 2 is the speed of height h2 and v i is the speed of height --
at h l. Note that 8 is always assumed to be positive. The combined gradient of the speed of sound, C, resulting from both the temperature and wind gradients is thus C = z(8 cos & - a) (4) l
*This is a shakier simplification than that for the temperature prof ile, f1] and normally holds only for near-neutral conditions. [3]The actual shape of the wind profile is a f unction of surface roughness, qnd of vertical momentum transfer due to _.
thermal instability.
~.- .
o 1. u i L' where 4 is the angle between the direction from which the wind is I coming and the sound path (Fig. 2).
-. Each sound path can be classified as " upwind", or " downwind" for a given sample of meteorological data,.on the basis of the following steps.
! a. If a is positive and greater 'than 8 (a > S ; so that C 1
L- would be negative for all values of $), then no shadow zone can
, exist and all paths are classified as " downwind". This is the strong-inversion, low-wind condition. ,
- b. If a is negative and numerically larger than B
.- (i.e., l-al > S , so that C would be positive for all values of 4 ), then the shadow zone completely surrounds the source and all
- paths are classified as " upwind". This is the strong-lapse, low-wind condition.
- c. If lal 5 S , then the " critical angle", &c, (where
, temperature, and wind effects cancel) is calculated by setting C = 0 t
L. in equation 4 C = z (B cos $c - a) = 0 l- (5)
&c = cos-1 a8
, where 0 1 4c 1 1800 It is now necessary to do some coordinate transformations of the azimuthal data, entered relative to true North, to bearings l rela tive to the direction from which the wind blows. Re f er to Fig .
ol l 2. The wind-sound angle,& , is: 1 -s 'l -
i 1 e *
& = lO p - Gy l, or if lO p - 0,l > 1800: _
& =.360 - lO p -Ol w (6)
Examine the difference &c ~ @ If $ < &c then the path'is an " upwind" path. If 4 > 4c then the path is a " downwind" path. _ It is clear that this simplified model does not take into __ consideration some common ef fects, such as changes of wind direction with height and location and upper level inve rs ions which can lead to significant sound propagation to distances quite remote from a source. Computing the Distance to the Shadow-Zone Boundary, X Nyborg and Mintzer [43 have derived an expression for the - distance, Xg ( See Fig . C2) , from a sound source to the boundary of its shadow zone at the height of the receiver, R,f t. above local ground, and in the presence of a vertical sound velocity gradient which varies with the logarithm of height. Their work has been _ adapted for this procedure in the following fo rm: Xg =S[3IL . f f feet - 47S . fffinEnglishunits (7)
/C w _
where S is the ef fective source height in feet above local ground, _. from Table CI. The distance Xg is and the function f fh)is obtained in feet and is assumed to be frequency-independent. M e-.J e-
, - -,,m -------w---- , -
i
.. 3 l
P I l l TABLE CI
, . f vs. f for computing X, in Eq. (7)
(after Nyberg and Mintzer4I) u R/S f(R/S)
< 0.05 0.4 0.1 0.45
'~
0.2 0.55 g 0.3 0.6 0.4 0.7 t 0.5 0.75 0.7 0.85 0.9 1.0 1 1.05
~
- 1. 5, 1.25 i
2 1.5 ! 3 1.9 4 2.3 m 5 2.65 t 6 3.0 y 7 3.3 8 3.65 I 9 3.95 (; ~ 10 4.2 i
> 10 Set I >D o
Interpolation is permitted, and f manual computations a graph _ of f(R/S) vs. R/S is most useful m w
. , . . _ . - _ . . - - - . . , , . _ , . , - - + - , . - - , , - - - , . . - -
- , - , , , , _ , .- - - - - - , , - - - - . - - - - ~ --
( Attenuation within the Shadow Zone, A,, _ Theoretically, the attenuation within a shadow zone can be arbitrarily large for large distances beyond the shadow boundary. - In practice, more than 25-30 dB is rarely observed because the loss of sound energy from the direct waves is partially replaced by the _ energy of indirect waves scattered from turbulence, ground surface roughness, etc. In this procedure, we have used representative values derived from the experimental work of Parkin and Scholes [6,7] and Weiner and - Keast.[8] The recommended values (Table 3 of the main text) have an upper limit of 20 dB. Attenuation because of a shadow zone has - occasionally been observed to decrease somewhat at extreme distances relative to~ closer-in distances. The conservative values in Table II allow for this possibility.
~
===
M tems M
% 4 mmE
.o
( t_ TABLE CII. ATTENUATION WITHIN THE SHADOW ZONE, A,g ,
._ VS. SIREN-TO-LISTENER DISTANCE, D, FT.
D i 1.2 Xo 0 dB l.2 X, < D 1 1.7 X, 5 L_ l.7 X, < D i 2.4 X, 10 2.4 X o
<D<3.4X o 15 w -
D > 3.4 X, 20 L
,+
.v w
w i 6 m w h l 5 h i
.J' I'
i REFERENCES ._.
~
- 1. Keast, D.N. and F.M. Wiener, "An Empirical Method for Estimating Wind Profiles over Open, Level Ground", Trans. Am. Geophys. ,
~
Union R , (5), pp. 858-864, October 1958.
- 2. Sutton, O.G. , Microme teorology , McGraw-Hil1 Book Co. , London, -
Sect. 6.4, p. 206, 1953.
- 3. Ibid , Sect. 7.2, p. 232 et seg.
- 4. Nyborg , W.L. and D. Mintze r , " Review of Sound Propagation in the icwer Atmosphere", WADC Technical Report 54-602, Sect. 1.5.5,
- p. 5 0 e t seg . , May 19 55.
- 5. Ibid, Fig. 16, p. 54. .
- 6. Parkin,.P.H. and W.E. Scholes, "The Horizontal Propagation of Sound from a Jet Engine Close to the Ground, at Radlett",J.
Sound and Vibration,1, pp. 1-15, 1964.
- 7. Parkin, P.H. and W.E. Scholes, "The Horizontal Propagation of ,
Sound from a Jet Engine Close to the Ground, at Hatfield",J. -i Sound and Vibration, 2 (4), pp. 353-374, 1965. 1
- 8. Wiener, F.M. and D.N. Keast, " Experimental Study of the Propagation of Sound Over Ground",J. Acoust. Soc. Am., 31 724-
~
733, 19 59. W W m
e o e
.., q h 4 l
l l w i meme I w e w W i. M APPENDIX D N I 6 ( .. t *
===
s om> o.-. 4 M O I e e I
l 1 { o APPENDIX D. DEPENDENCE OF ALERT UPON SIREN DURATION In the main body of this report, the chances of alert are - 1 predicted for a four-minute period of siren operation (here called siren duration). In this appendix, predictions are generalized for _. longer' and shorter siren durations. This appendix will allow readers to convert four-minute results .to results for other siren durations. This appendix begins with an overview of the relationship _, between siren level and siren duration, and how this relationship affects the chances of alert. It continues with development of the mathematics of this relationship, and then summarizes results for . the reader's use. D.1 Ove rview i Table 1 is a typical " chance-of-alert" table for a particular background-noise environment. Siren durations are listed across the top, .and siren levels down the lef t side. Within the table are the chances of alert -- from 100 down to zero percent. In the main body of this report, results are based upon the four-minu te columns of ~ tables such as this one.* Variations within the table are related l to fluctuating background noise in the listener's environment.** _. l I
- And upon the one-minute columns for rotating sirens. __
l i l ** Precision within Table 1 degrades for longer siren durations ( to the right) and for lower siren levels ( to the bottom) . For - longer siren durations, precision suffers from the limited amount of total data that underlie the table. These data include 250 l minutes of background noise, which is only about eight times the l longest siren duration. For lower siren levels , precision - l suffers from the very small percentage of time that these low I siren levels will alert the listener. Although the amount of data is large' compared to the siren durations, the background ~ noise is rarely low enough to contribute to the statistics at these low siren levels. For longer siren durations and lower - l siren levels combined , the precision is particularly bad. M
+- - ..w. -.____a
_ . . . e [~ TABLE 1. TYPICAL CHAMCE-0F-ALERT TABLE FOR A PARTICULAR BACKGROUNO-NOISE ENVIRONMENT, 5! Rett 31Etts LEVEL' DU9AT IDil (N IFPJ T E5)
- 1 2 3 4 5 4 7 S 9 to II .12 13 14 15 15 74 100 100 100 t ou 100 100 100 100 t ou 100 100 100 10H 100 100 100 73 99 100 100 100 ton ICQ 100 100 100 100 100 100 100 100 100 100 72 ?? 99 100 100 100 100 100 100 100 100 100 100 800 100 100 100 71 ?3 99 100 100 100 100 100 100 100 100 100 109 100 800 100 100 u- 70 93 ?? 100 100 100 100 100 100 100 100 100 100 100 100 100 100 69 99 99 100 100 100 100 100 100 100 100 100 100 100 100 100 100
. 69 97 94 100 100 100 100 100 100 100 100 100 100 100 100 100 100
; 67 97 ?9 100 100 100 100 100 100 100 100 100 100 800 100 100 100
~~
is 95 93 100 100 100 100 100 100 100 100 100 100 100 100 100 100
$5 ?2 96 93 ?? 93 100 100 100 100 100 100 100 100 100 100 100 64 92 95 99 ?$ 99 104 500 100 100 100 100 100 100 100 100 100 63 $9 92 95 97 96 100 100 100 100 100 100 100 100 100 100 100 62 97 91 94 95 ?S ?3 100 100 100 100 100 100 100 100 100 100
-- it ?i 90 ?4 95 ?s ?3 100 100 100 100 100 100 100 100 100 100 50 S5 90 93 95 96 ?? 100 100 100 100 100 100 100 100 100 100 59 13 37 90 ?2 94 95 100 97 100 100 100 100 100 100 100 100 59 St Si 90 92 ?4 95 100 97 100 100 100 100 100 100 100 100 f-- 57 ?? SD $3 39 90 93 94 ?4 ?S 9i ?6 95 100 94 500 93 55 73 S2 ST 96 S3 ?0 ?4 94 ?6 ?s 96 95 100 ?4 100 * ?3 55 74 ?!? ?2 33 34 36 89 ?? S? ?2 91 90 95 $9 94 37 54 70 TS St S3 34 Si ?? 37 29 92 ?! ?0 95 3? 94 G7 i 53 ss 7s $0 St 34 Ss ?? $7 39 ?2 ?! 90 95 S9 ?4 $7 i' 52 45 75 77 ?? S2 33 36 34 ?? S3 ?? 90 95 39 ?4 37
~~
51 40 70 74 75 73 ?! 33 31 36 34 33 34 29 33 33 30 50 55 65 70 71 74 79 91 St $2 30 $3 ?$ 34 33 S1 30 49 51 61 67 63 70 76 7S $1 79 30 33 SS ?4 33 31 S0 43 49 57 63 65 66 71 75 74 75 76 73 91 79 S3 75 30 47 42 54 60 43 64 71 75 71 75 76 74 St 79 33 75 30 l 46 37 47 54 57153 67 6? iS 71 72 70 76 74 7S 69 73 45 33 44 50 56 56 64 67 65 71 SS 70 76 iS 73 69 73 l' ' 44 33 43 50 56 Si '64 67 45 71 is 70 76 is 73 69 73 43 30 40 45 54 52 57 64 s5 63 63 70 71 6S 73 59 73 42 26 33 37 41 42 49 50 55 57 56 57 62 58 61 56 60 41 21 23 05 OS 40 45 50 55 57 56 57 62 SS 61 56 60 . 40 13 24 30 33 34 38 42 43 46 44 52 57 53 50 50 60
; 39 14 20 24 29 30 34 36 42 43 40, 48 52 47 44 50 60
,, 33 12 17 20 25 23 33 33 39 43 40 4S 48 47 44 50 53 37 to 15 19 22 24 31 38 35 39 36 43 43 47 44 50 53 36 3 12 14 16 13 21 22 26 29 24 30 29 3E 33 31 33 35 7 ? 10 11 12 14 14 16 12 16 17 19 21 22 I? 20 34 6 3 3 tu to 12 11 .13 14 16 17 19 16 1.' 13 13
-- 33 6 8 8 10 to 12 !! 13 14 16 17 19 16 17 13 11 32 6 3 S to 10 12 11 13 14 16 17 19 16 17 13 13 38 6 3 S to 10 12 11 13 14 16 17 !? 16 17 13 13 30 4 7 7 3 3 to 3 to 11 12 13 14 11 11 13 13
~~
-29 6 7 7 8 3 to 3 to 11 12 13 14 It it 13 13
. 23 5 6 6 6 6 7 6 6 7 3 9 to 5 6 6 7 27 4 6 5 6 6 7 6 6 7 ? 9 10 5 6 6 7 26 4 6 5 6 6 7 6 6 7' 3 9 to 5 6 4 7 25 3 5 5 6 6 7 6 6 7 8 9' 10 5 6 6 7
,_ 24 2 4 5 6 6 7 6 6 7 3 9 10 5 s 6 7 23 1 2 4 5 0 7 6 0 7 0 9 10 0 s 0 0
.m I.
mW I s ,
1 j In this table, the chance of alert is 100 percent when the _ siren level is much higher than the background noise could ever be at the listener. When the siren level is 74 dB, for example , the siren will definitely alert the listener even for siren durations as short as one minute. The chance of alert is zero percent when the siren level is low, say 20 dB or less, no matter how long the siren sounds. The _ background noise is always sufficient to mask (acoustically cover up) such low siren levels. . For siren signals of intermediate levels, ' the chance of alert falls between 100 and zero percent, in the detailed manner shown. ,, These intermediate de tails follow from the fluctuations of the background noise, from minute to minute. For these intermediate siren levels, the chance of alert For a increases with siren duration as indicated in the table. { . siren level of 50 dB, for example,- the chance of alert is 71 percent if the siren is sounded for four minutes. If this duration is doubled to eigh t minutes , the chance of alert increases to 81 percent. How can this increase with duration be unders tood mathematically? If such understanding results in a particula ma thematical pattern, then this pattern can be used to convert four-minute results to results for other siren durations. The search for I
~
this mathematical pattern is the subject of the next section. D.2 Developeent of the Mathesaatics The search for patterns within tables of numbers is necessarily an exploratory matter. First, some underlying mathematics must be M
?f-
- b.
.~
__ pos tulated , and then a numerical pattern must be sought with this mathematics as guidance. Once a. preliminary pattern is discovered, it must be simplified to be of use, and then must be generalized for other similar tables. Ideally, the pattern will emerge as a simple equation, with a small number of adjustable constants. The steps involved in developing such a pattern are: L. e preparation
,l e underlying mathematics and its simplification
<L e exploratory graphs, guided by the mathematics e simplification and generalization to all other tables
, These steps are discussed next.
~
D.2.1 Preparation
- Figure 1 shows typical background noise as it fluctuates over a one-minute period. The fluctuations are generally large, as shown here. In this background noise, a listener will be alerted by a siren whenever it is 9 decibels or more above the background noise leve l .
- The figure shows a siren that produces a steady 49 dB at
~
the listener. A dashed line 9 dB below the siren level denotes the i .I
*Throughout this appendix, background noise includes the noise in l_ a 1/3-octave frequency band centered at 630 Hz, a typical siren operating frequency. Dictated by the physiology of the ear, only j '
this 1/3-octave band is available to mask, or cove r up, the pure-tone signal of typical sirens. Siren levels are usually measured as ove rall sound levels , though the same values would be measured using only a 1/3-octave frequency band filter.
]- -
l FIGURE 1. BACKGROUND NOISE LEVEL AS A FUNCTION OF TIME. l 80 sa SIREN WILL ALERT - LISTENERS DURING THESE INTERVALS
~
70 -
.-. b l
~
% v i i
e ; 60 - Z3
$.E I oo i 1 F2 f -
}
8 l! _J s I w [ 50 4 - j " SIR EN '-
, cn ( I\
g 9dB m 40 --- ---- - - - - - - - -
" THRESHOLD MINIMUM 34 dB 30 '
O 0.5 1.0 TIME (minutes) m m W M m
m,7 . b.
~_ -
alerting threshold. During the shaded time intervals below this threshold, the siren will alert the listener. This siren level has succeeded in alerting the listener during
, its one-minute duration. However, a siren level some 7 decibels lower would not alert because the background noise would always be i above its lowered threshold line of 33 dB.
This figure suggests another way to phrase the alerting d
, question. Instead of asking if the siren is loud enough to cause l alert, one could ask: For a given siren level, is the background I 1, noise ever low enough to allow alert? Since the background noise is f'-
ii continually fluctuating, this question is inherently a statistical question. Its answer depends upon the statistics of the background 4' l
- t. noise fluctuations.
i The answer to the above question is: Yes, alert will occur during this one-minute period if !t . I' backgroundI minimum I Lsiren - 9dB
~
j Otherwise, the siren will fail to alert the listener. The only j statistic of interes t, therefore, is the minimum background noise {' level during this one-minute period.* 'i '
*Our analysis for this study actually utilized the 90-percentile background noise level, rather than the minimum level. The 90-percentile noise level is the level exceeded 90 percent of the - time; the remaining 10 percent of the noise f alls below this level. Use of the 90-percentile noise level adds a measure of conservatism to the results, since it requires slightly higher ._ siren levels before alert is predicted.
J w
f . i Figure 2 shows a series of one-minute minima .for forty ._ successive one-minute time periods. Every minute's minimum is different, as the figure shows. Tnese 40 minima were measured over a 40-minute time period, and are part of a much larger set _ (approximately 250) of total data. For the siren level shown, 35 - percent of the minima (14 out of 40) fall below the threshold - line. Therefore, this siren level in this background noise has a 35 percent chance of alert -- when sounded for a du' ration of one _ minute., This plot applies only to ~ sirens sounded for one minute, since the background-noise minima are one-minute minima. Stated another way, when a siren is ' sounded for one minute, it has a'n' equal chance a of encountering any of these forty one-minute time periods, which represdnt alb one-minuto periods. During 35 percent of these minutes it will alert the listener, since thE noise falls. below the ale rting threshold at least once during those minutes. t
~
Next, say that the siren is sounded for four minutes. Figure 3 shows' the four-minu te minima of interes t -- as circled dots. Each _ of the'so-is just the IcVest of four one-minute minima in each four-mi nu te ,grou ping , of these fo'ur-sinu te minima, 60 percent (6 out of i
- 10) fall below the threshold line. *Thereforef this siren level in this backg'round noi_se has a 60 percent 'ch'ance of alert when sounded for -a _ duration of 'four minutes. Note that~the~ chance of aiert has -'
increased -with the siren duration. ~
Needed is ma thema tic * ~
Z8 ; . . 9dB -
-w . . . .
t 2m . . . 3 L I w o ...
. . . o s g 40 THRESHOLD
- - 3 .
z-zo . . 23 . wo . .
.s 2 Q- '
00 * - lu: . ' y30 . m . I w 20 0' ' ' ' ' ' ' ' '10 i' ! i 20 30 40 TIME (minutes) I ( = l m m- - -,y -- - - , - - - - ,a m - - - - - - - - - , , , , , , ,-,
i FIGURE 3. MINIMUM BACKGROUND NOISE LEVELS OBSERVED IN FOUR-MINUTE INTERVALS FOR A 40-MINUTE TIME PERIOD (From Figure 1). O 60 , ,, , ,
@ FOUR MINUTE ~
MINIMA , m V 50 -
, SIREN
-<]w ,
3 E> . . - 5" . . . 9 dB Em * * *
- Wm ..
$ @ 40
- -@ -@ - - - . e- @
o
. THRESHOLD z
ggo is OE
.* ~
"Ou 30 -
f ,4 , 4 ,4 , 4 , 4 , 4 , 4 ,4 , 4 , 4 , , 20 ' ' ' ' O 10 20 30 40 ,
~
TIME (minutes) e <=e tumWD
. m "ee w
M
l_
'_ Chance of Probability Success of Success of Failure
'~
100% 1.0 0
, 80% 0.8 0.2 60% 0.6 0.4 40% 0.4 0.6
, 20% 0.2 0.8 4
0% 0 1.0 Note that I
.L Pf ailure = 1 - Psuccess i
-' and that failure occurs when minima points are above the threshold i line.
L D.2.2 Dnderlying Mathanatics and its Simplification l
~
Figure 2 above contains one-minute minima for a total time l : period of forty minutes. All the points in this figure are collapsed onto the vertical axis in Figure 4, at the left. They form a " cloud" of points denser at intermediate noise levels and & _ sparser for higher and lower levels. This is a probability " cloud", in which area is proportional to the probability (density) of one-l minute minima. For any cne-minu te period , the probability of failure is l __ proportional to the " cloud" area above the threshold line. This upper area, divided by the total cloud area, is the probability that
~
the background noise will exceed the threshold level throughout any a* a. ul .'~ .
FIGURE 4. PROBABILITY " CLOUDS" FOR ONE-MINUTE BACKGROUND-N0ISE MINIMA IN SUCCESSIVE MINUTES, ASSUMING EACH MINUTE IS STATISTICALLY - INDEPENDENT OF ALL OTHER MINUTES. 60 , , , g , g f ,'
- 50 '. f
* " SIREN 3 'i
< = 1 t 1,/ \ p g
- j ,
9dB - E ... h 40 *C'------ --- ---- ---- - - - - - - THR ESHOLD y *
.1 \
1
>l l
~
\h \t lh l )I E . .
o - 30 - h d' h j - 20 ' ' ' ~' 1 2 3 4 . TIME (minutes) ._; l l -
'm
x
-f* :
L one-minu te period -- that is, the probability that the siren will fail to alert the listener. This one-minute probability of failure is ( 1-0.3 5) = 0.65 for the example shown. To the right in the figure, this cloud is duplicated at each of four successive minutes. If we assume these four minutes to be independent of one another, this probability cloud would apply [_ equally to all of them, as shown. Let us assume this to be the case for a moment. Then, for the siren to fail after four minutes, if l must fail for each qf .the one-minute periods. Therefore, the probability of failure after four minutes is t 5 P(4) = (Py)(P 2)(P3 )(P4)
'l
=
(Py)4
, L In this equation, P( 4) . means the probability of f ailure af ter a total of four minutes have. gone by, while P 4 means the probability
.' of f ailure during the fourth minute only.*
i l _ This equation, however, is valid only if the one-minute periods are independent of one another. A glance at Figure 2 above ij indicates that they are not independen'. t For example, for a one-- minute period with a very low minimum, the following minute probably ~~~ j also has a low minimum. There is a regularity in the successive , minima; they are not independent. For this reason, the cloud
. picture must be modified to that of Figure 5.
1 l L. 1 1 -
*If we had worked with probabilities of success, combining four i
l minutes into one equation would be far more complicated. That is ,, why we choose to work with f ailure instead. As the very last '~ s tep, we shall convert from f ailure back to success. u !l__ l l
I FIGURE 5. PROBABILITY " CLOUDS" FOR ONE-MINUTE BACKGROUND-NOISE MINIMA IN SUCCESSIVE MINUTES, ASSUMING MINIMA IN _ SUCCESSIVE MINUTES ARE NOT INDEPENDENT. 60 , , , , 50 - j) ]
~ ~
g j n SIREN i - 9dB 5 o uJ 40 - - - - - - - ---- - - - - -
- THRESHOcc -
( a \t 3
- a. - -
2 0 30 - t 20 ' I I I l 1 2 3 4
~
TIME (minutes). - w m* w W
~~ .
h. In Figure 5', the firs t minu te 's cloud is un ch anged from Figure
- 4. However, the second minute's cloud represents the conditional probability of: " failure during minute two , give n that failure occurred during minute one." In other words, the cloud at minute j two. represents the probability that the second minute's minimum will
'~
be above the threshold, given that the first minute's was also above the threshold. Mathema tically, we write P2:l' for this conditional probability. Then i L_ P(4) = (Py)(P2:1)(P3:1,2)(P4:1,2,3)
\ Y L
conditional probabilities 3 L. Note that P 2:1 is greater than the independent P2*
, _- 2:1 1
~
This increase is due to the regularity between successive minutes -- technically to the correlation between successive minute's minima.
-- The higher the correlation between successive minima, the more this probability cloud will condense above the threshold line. The l ,_
remaining clouds condense even more above the line, since they are failure probabilities, given that several failures have preceeded. A short numerical example will be useful here. , For nct correlation, we have P(4) = ( 0. 6 5) ( 0.6 5) ( 0'.6 5) ( 0.6 5) ~ i __ P(4) = (0.65)4 = 0.18
; and therefore the probability of success is 0.82. For some i 4-i
/
l
\ __
t
( correlation, we have P(4) = ( 0.6 5 ) ( 0.8 ) ( 0.8 5) ( 0.9 ) P(4) = 0.40 for a probability of success of 0.60. And for full correlation we _ have P(4) = ( 0. 6 5 ) ( 1. 0 ) ( 1. 0 ) ( 1. 0 ) P(4) = 0.65 for a probability of success of 0.35. , In general, a P(n) = (Py)(P2:1)(P3:1,2)...(Pn:1,2,3,...,n-l I t (1) ~
=
(Py)" for no correlation
= Py for full correlation The upper. half of Figure 6 illustrates graphically how the
~
probability of failure thus decreases with increasing time -- that is, with increasing siren duration. The probability of success therefore increases with siren duration, as shown in the bottom half - of the figure. (This figure is an example only, not a general l result.), l Note for large correlation between successive minima, there is not as much benefit in sounding the siren longer. If the siren fails to alert during the firs t minute, it will mos t likely fail to , alert thereafter, because the first minute is nearly identical to _; all subs equent minu tes. l W unme
W
._ rIGURE 6. GRAPHIC ILLUSTRATION OF SIREN-ALERTING PROBABILITIES VS.
SIREN DURATION, FOR VARIOUS AMOUNTS OF CORRELATION BETWEEN BACKGROUND-N0ISE MINIMA IN SUCCESSIVE MINUTES (Example Only). t . 1.0 , , , , , , ,, .
~
, w -
. cc 3
FULL CORRELATION
~ ~
l I
!' O.5 -
*%Op _
~
{. . TION - u e . : +o '--------.-_ . {l -
.s. -
0 ' ' ' ' ' ' I t'- 1.0
- I i i t a e g 4-
- p. .
vs
$ +O
~
- 8 E" -
j S
~
SO,N haaan.AI - l o 0.5
.N
'/ --,
I 5 -
,g . . FU.LL COR.RELAT. ION. -
i 2 - l ,- n. , 0
' ' ' ' ' ' ' I i . 0 1 2 3 4 5 6 7 8 9 l SIREN DURATION (minutes)
- i l_
's s
- . ~ . . - , - -.
, ,y_ . . , . - , , , - . _ _ _ , _ _ _ _ , _ . , ., ,
( This underlying ma thematics res id es in Eq. 1 above. In Eq. 1, the notation Pn:1,2,3,...,n-1 reminds us that P n is a conditional probability, which assumes the siren failed during all previous minutes. We next simplify, so that this P n assumes failure only during the Lmmediately proceeding minute. Mathematically, . P:n 1,2,3,...,n-1 = Pn:n-1 Le t
~
Pn:n-1 = CPy where C contains all the conditional aspects of the probability. - The term P y is the unconditional probability for the first minute. Then _ t P(n) = (Py)(CPy)(CPy) ...(CPy) P(n) =P y n Cn-1 (2) Note that for no correlation, C=1 (3) __ and therefore - P(n) =P"y . as before. For full correlation, 1 ,, C= Py (4) ta=* w e M M
e.
~
to make P(n) =Pn( )n-1
=P y L'
as before. iI. '- Eq. 2 is the desired simplification. In the following section, we graph measured background data, to explore the nature of C, for
\
u correla tions typically present in measured background noise data. II t-
! D.2.3 Exploratory Graphs, Guided by the Mathematics i
,- To explore for C graphically, we first take the logarithm of Eq. 2.
b.
=P"C-1 n l
P(n) y j log P(n) = nlogPy + (n-1) log C log P(n) = -logC + n log CPy ] (5) I . L If log P(n) is then plotted against n, the resulting straight line should have a vertical intercept of -logC and a slope of log
; CP 1 After some curve-smoothing on linear paper, on Figure 7 we logarithmically plot part of the data in Table 1 above. Each line
; is for a different representative siren. level, labelled (})
through . of course, the linear curve-smoothing helped line up the points shown here. Even so, the regression fit to straight lines for each i siren level is very good. Note however, that the ve rtical
~
intercepts and the slopes vary from curve to curve. Therefore, C ,f must va ry with siren level. We then set each intercept equal to -logC and each slope equal I
'~
1 to logCP1, and solve for C and Py -- separately for each straight line. k
- __ _ ~ -
FIGURE 7. LOGdRITHM 0F THE PROBABILITY OF SIREN FAILURE-TO-ALERT VS. (' SIREN DURATION FOR FIVE DIFFERENT SIREN LEVELS, OERIVED FROM THE DATA IN TABLE 1. -- 1.0 ' -
- --- . h I l I e
- ~ . - - -
%. LOW -
-* SIREN .
0.5 \ -= LEVEL
~
ig-
\- *
' ~
ul . s . . . d - 2 o 0.1 - . , , _ c - E _ m O m a. 0.05 - . . _.
, h-h FOR DIFFERENT - -
SIREN LEVELS HIGH
' SIREN LEVEL l
Q 0.01 ' I - I e i , 0 10 20 30 40 SIREN DURATION, n (minutes) - l l l l 1
t' . .
- i. -
t s_ L_ - Line C P 1 Number (b) 1.073 0.925 (}) 1.426 0.678 f (3) 1.816 0.520 j_ ([) 3.062 0.293 l (h 4.064 0.199 g From Eq. 4 above, we suspect that C may be a power function of L P y, and so we plot logC against logPy in Figure 8. On this plot, the straight-line fit is also very good. It yields: i' , C= (Py)-0.87 ',}.i [ It seems to make sense, based upon this limited analysis, to
., generalize to ' l' C= (Py) ~0 k.
where p (rho) denotes a correlation coefficient. Zero correlation would then make - u
,, C= (P y) = 1 and full correlation would make
~
I C= (P y) = p
.' These agree with Eqs. 3 and 4 above.
L g -' e I m"
.---,--.-m,- , n - ,-,, ---,r,, - . ,,-, -- - - - - ,
-m -
n- --, - _-
- A e' ' u 9
-.e e
s FIGURE 8. PLOT OF VALUES OF CUMULATIVE CONDITIONAL PROBABILITIES (LOG C) VS. PROBABILITY OF FAILURE IN THE FIRST MINUTE (LOG Pj), - DERIVED FROM FIGURE 7. e W M 6 - w e G.
^ -
5 4 e - O
~
E - . - x, 1.0 ' ' ' I ' ' ' '*$ 0.1 0.5 1.0 log P, W M sm
,--u - , -, -r - - - - - - - -- --.n- - - - - - - , - . - - - - , - , , - , - - - - - - , - - - - -
_p . l.- .
.t-In summary then, the time-pattern within Table 1 can be written as l_ ~
(pl)0.87 + 0.13n M) P(n) = (Py)n-0.87(n-1) ,
"~
The two constants in the exponent sum .to 1.00, and depend upon
,1 correlation within the background noise, from minute to minu te.
- b. Moreover, Eq. 6 depends upon the siren level through Py, which varies with siren level.
t!- L.
, Next, we simplify Eq. 6 so it may be generalized to a wide i
variety of noise-level tables, not just Table 1- above.
\_
Eq. 6 is valid for all siren levels , in the presence of the particular background noise used to develop Table 1. Its general form ~ is n-1
-- P(n) =
(Py)n C
-. = (Py)n' gp ) -p(n-1) u
, (p1 )p + n (1 - p) (7)
( In logarithmic form, m ii ~ log P( n) = p + n ( 1- 0 )3 log P y
=
p log Py+n (1-p) log P (8) I ~ With logP(n) plotted against n, this is the equation of a straight i j line with vertical intercept p log P y and slope (1-p) log P .
- i '-- .
s
t A normal regression fit would solve for the two variables - and Py, separately for each of the siren levels (as shown in Figure 7, for ins tance) . However, there is a relationship above _ that implies p to be a constant, independent of the siren leve l . Therefore, we wish to collapse all curves, for all siren _. levels, to a single curve. For this purpose, we manipulate Eq. 8 as follows : F log P(n) = p + n ( 1- p) log P y ,
' ~
J log P(n) , y, log P 1 = 1 + ( n-1 ) ( 1- p ) (9) _ Hence, plotting (logP(n)/logPy) against (n-1) yields a s traight line of interceot 1 and slope (1-p) ,' independent of siren level. In other words , each curve in Figure 7 has been , normalized to its value of Py , and all curves have been , collapsed into one.
~~
We will have need below for a similar equation, but normalized to the probability at four minutes, rather than at one mi nu te . We develop this next. a In the graphs above, letter n was interpreted as progressing
~
in one-minute steps (n=1,2,3 equals t=1,2,3). However, nothing l in the ma thema tics requires this interpretation. Jy3L time , interval could be taken as the basic interval n above. In -- particular, the basic time interval could be taken as four minutes. Then four-minute minima (n=1) would combine into eight- ._ l j minute minima (n=2), and so forth. The result would be Eq. 9 above, but with n = 4t (in minutes) v s W
P h l* (n=1) (t= 4 minutes) [. Figure 9 schematically compares these one-minu te and four-minute normalizations.* For the one-minute normalization on top: n=t, and therefore n-1 = t-1, as shown on the firs t horizontal axis. Plotted horizontally is the range i v 01t-113 [_ 1<t< 4 4 j The small plotted points represent the tabulated values for these four minutes, collapsed into one line by the -Py normalization. g The line is fit by linear regression and has slope (1-p). L~ This upper portion of Figure 9 is for rotating sirens. As explained in the main text, rotating sirens are less effective in
~~'
alerting the public, since they produce their maximum siren level for only a portion of their duration. For this reason, four-minute results for rotating sirens are derived from the one-minute background-noise statistics. In the figure, the third ( ' L; horizontal scale shows the corresponding siren durations for l rotating sirens. The normalization is therefore to a four-minute
; L_
l i ~ 'i
- Note that the lines in Figure 9 rise rather than fall to the r igh t, as does Figure 7, for this reason: In Figure 7, the actual logarithms on the vertical axis are negative, since the
-- P(n) 's are less than unity. Therefore, this vertical axis i actually decreases, from zero at the top to minus-two at the bottom. For increasing n, then, the curves take on increasingly large negative values ( for example: -1, -1.5, -2). Figure 9 is normalized by -logPy , howeve r, which is also nega tive , and which turns these increasingly negative values into increasingly positive values. Therefore, the lines rise in Figure 9.
L
. /
f 2
FIGURE 9. SKETCH OF ONE-MINUTE AND FOUR-MINUTE NORMALIZATIONS TO SHOW RELATIONSHIP BETWEEN VARIABLE N AND SIREN DURATIONS. log P(n)/ log P, h ONE-MINUTE
- ! NORMALIZATlON
}., n=t
- .m
~
- Cr- Y.;
- 3 i
$. 'I.
f 1 O ! !
> n -1 (= t -1)
O. 1 2, 3, m i .. . . . - t .
~
1 2 3 4
' ' ' ; ROTATING SIREN l
4 8 12 16 DUR ATION (min) - log P(n)/ log P3 A . i - ie : FOUR MINUTE
. I I NORMAllZATION
.I Id~, n=d4 *
- >.a . L
- i ? - h I
& j. ) si l ' .I . i 3-
' y- { ". ,i : '
,, ;, ,s 1 s ; *
- *. 1 * . ~
.l. ..t
.1 .' ;O ; . ; l . . . l ; ; .l > n - 1 (= t/4 -1) -
0 1 2 3
- : : : : : : : : : : : : : : > t 1 2345 6 7 8 910111213141516
. . . . , , . , , ', , , . , , , , STATIONARY SIREN __
1 2 3 4 5 6 7 8 910111213141516 DURATION (min) W e=e
- n.- - - -
I r siren duration, and the graph extends up to a maximum of 16 minutes, fm For the four-minute normalization at the bottom of the figure: n = t/4, and therefore n-1 = t/4 - 1, as shown. Plotted horizontally is the range 1 3 t
. T I T-li3 it i S T S4 i 1<t< 16
.L The second horizontal scale shows time t and is identical to 1 the third scale, which shows duration of stationary sirens. The L_ normalization is therefore to a four-minute siren duration, and
~
the graph extends up to a maximum duration of 16 minutes.
'~
Using these equations and normalizations, the curve-fitting procedure was applied to six background-noise tables -- tables similar to Table 1 above -- developed from data measured at 74
; different indoor and outdoor locations. In this curve-fitting,
;- no linear smoothing was used, and data from all siren levels were
, used without omission. Table 2 contains the resulting slopes.
Ii - These slopes were next converted to p , assuming that they
; equal ( 1- p ) , as labelled in the table. The resulting twelve values of p were plotted against the corresponding values R of the auto correlation function, to obtain R g = -0.034 + 1.0510
=.p This regression equation has a correlation coef ficient (between
_l .! values of p and Rxx) of 0.85, which is satisfactorily high. ll i i .~s il i l r- e .- ,
f
\
TABLE 2. Listener Subclass Resulting Slopes (1-0) Location Stationary Ro tating Sirens Sirens Indoors Scenario 1 0.217 0.142 ~ Scenario 3 0.274 0.254 Outdoors Rural, day 0.164 0.177 Urban, day 0.065 0.103 _ Rural, eve / night 0.150 0.075 Urban,, eve / night 0.046 0.039 4 emsp w Gump e. W w M W
.- - _ . - - ,_ , _ . , - , _ _ . - - _ _ . , - , - - - , . , - - , - . . . - - . - , - - - , - - . - , - ,__v--, _ ,
.,e ,
i, m In the next section, we collect these results into a form of use to the reader. D.3 Sn== nary of Results Figure 10 contains the results of the analysis above. This g figure is used as follows:
. e Conve rt the four-minu te " chance of alert" to a L- " probability of failure-to-alert":
J L P= 1 - (Chance 'of alert)/100
;. e Raise this value to the exponent determined from Fig ur e 10, for the particular siren duration of interest.
-- P= (P4_,gn) Exponent (10)
L_ e Conve rt this " probability of failure-to-alert" back to a
" chance of alert":
!! .gi ll Chance of alert = 100 (1-P)
\
L W T am 0
'?
I. pt i _- a
- - ._ ~ _ ___ ___
a , FIGURE 10. GRAPH OF EXP0flENT FOR USE IN EQUAT!0ft 10. ROTATING SIREN 2 ' 3
' ' ' 8
' i '
8
' 8 8
i - EXPONENT
* @ ~ = 1 + (d/4 -1) (SI. OPE)
@l m- '
SLOPES: t* O.254
~
$ $ 3- 0.177
$j _
~
0.142 0.103
@ SCENARIO 3
@ RUR AL, D AY 0.075 -
[ : 0.039
- @ SCENARIO 1 -
i @ URBAN, DAY 7 -
@ RUR AL, EVE / NIGHT -
@ URBAN, EVE / NIGHT -
r f a f a f a t i I , t , t , e 0 2 4 6 8 10 12 14 16 - DURATION OF ROTATING SIREN, d (min) i w STATIONARY SIR EN 2 i i i i i i ' ' i
' ' ' a a a '
- EXPONENT
~
~
@; = 1 + (d/4 -1) (SLOPE)
SLOPES:
- O 0.274 0.217 0.164
~
$3 __
-~
@ SCENARIO 3 0.150 2 -7 X
@ SCEN ARIO 1 0.065
@ RURAL, DAY 0.046
_~ @ RURAL, EVE / NIGHT -
. @ URBAN, DAY _
@ URBAN, EVE / NIGHT - -
O ' ' ' ' ' ' ' ' ' ' - ' - ' - ' ~ 0 2 4 6 8 10 12 14 16 DURATION OF STATIONARY SIREN, d (min) o almn.
t' . . L. l m bar i
\
m L. APPENDIX .E
~
t a 6 e e t
%=m.
h4 h l
~
l 4 $ d w 4 w
ANALYSIS INPUT / OUTPUT DATA This appendix provides listings of computer file input and output data for the Indian Point analysis. Explanation of the terminology used for each listing is provided below. Table E.1 INDIAN-SIRENS This file contains input data for each of the Indian Point sirens as follows: ~
. Siren No. - number assigned to each siren for use by computer program. -
. Siren Name - first letter indicates whether the siren is rotating or stationary type (R or S);
the remainder consists of the siren designation, which for this plant is identical to the J Siren No.
. x,y,z - these are the physical coordinates for the .L, siren location; the x-axis is oriented east-west, the y-axis is oriented north-south, and _
the z-axis is oriented vertically. The x and y coordinates are in units of km, referenced ,
~~
to the grid shown on the Feb. 1981 NRC Emer-gency Planning Map for Indian Point (the plant center is located approximately at ~ x=587.7, y=4569.2). The z coordinates are in units of feet. _.
. SPL@l00FT - these numbers indicate the rated sound pressure level for each siren at a distance of 100 _
feet, in dB. M eme
?
s_ i t
~-
. Table E.2 INDIANEARS This file contains input data for each of the randomly L_ selected listener locations as follows:
. Site No. - number assigned to each site for use by
- computer program.
. Site Name -
designator for listener site; the first
-- letter indicates whether the site is urban or rural (U or R) .
i. L- . x,y,z - these are the physical coordinates for the listener location; the x-axis is oriented L_ east-west, the y-axis is oriented north-south, and the z-axis is oriented vertically. The x and y coordinates are in units of km, referenced to the grid shown on the Feb. 1981 NRC Emergency Planning Map for Indian Point (the plant center is located approxi-mately at x=587.7, y=4569.2). The z L- coordinates are in' units of feet. l! . Rural Road - This column indicates whether rural listener sites are within 1000 ft of major highway f ~~ '! l (NEAR) or beyond 1000 ft (FAR) . This infor-t
~mation is required for the selection of l appropriate outdoor background noise levels.
1 i l y - , - , . _ . , - ~ . ,
( Table E.3 VAL-HUDSON - This file contains input data for each of the four sample scenarios as follows: _
. Scenario No. - number assigned to each scena'rio (see ..
Appendix A) . _.
. AMOL -
molecular absorption,-in dB/1000. feet.
~
. WIND - wind direction in degrees (0 = wind from north, etc.).
~
. NRES -
residential building outdoor-to-indoor noise reduction, in dB.
~
. NCRM .-- commercial building outdoor-to-indoor noise reduction, in dB.
~"
. F1 - F8 - activity fractions. - -
F1 = fraction of people outdoors.
'~
F2 = fraction of people indoors, at home, listening to radio or TV. F3 = fraction of people indoors, at home, sleeping. F4 = fraction of people indoors, at home, neither l sleeping nor listening to radio or TV. "- F5 = fraction of people indoors, at work, in commercial establishments. _ F6 = fraction of people indoors, at work, in industrial locations. F7 = fraction of people in vehicles in urban areas - at 30 mph. .
~
F8 = fraction of people in vehicles in rural areas at 55 mph. . s d W emme
'/ ,
L
/
Table E.3 (Continued) m
. INP - in' door alert orobability curve (see Fig. 4 of text) . .
. PU30 - probability of alert for motorists in urban areas at 30 mph.
L. . PR55 - probability of alert for motorists in rural areas at'55 mph. L_ . MUL - vertical profile ~of wind speed, Sz, in ft/sec/in ft. t i v.
. ADD - vertical profile of. air temperature, a, in F/ln ft.
' Table E.4 LISTENEROUTPUT This listing provides the number, name and outdoor sound
~
pressure level (LOUT, in dB) for the " dominant" siren at each sample listener location, for each of the four sample scenarios. The results are listed in numerical order for scenarios one through four for each listener site.
l.- Table E.5 PROBS l l ! - This listing provides the final results for the analysis. Information is listed in numerical order for scenarios one through four for each listener site. This information consists of alert
-- probabilities P1 through P8 corresponding to activity fractions F1 through F8, as well as the total pro'bability of alert (PT) for each i
, m .' sample scenario at each sample listener site.
~
i A summary is provided at the end of the listing showing the rural and _ urban population followed by the total rural probability i of alert (PTRUR) , the total urban probability of alert (PTURB) , i ' I e -. l' .l ~/ l 8 9
i ' a g '
.x ..
g t
.x , - . . ,
, - op .
..., - ... -m -s s
s s- ,q - s ,
.,_ . , y . y
- s. v . .
A*, . , Q, .
} ., ~ l \
.. ~ q
'ts N ?x ' ' Table s.
E. 5 (Continiidd) s - N ~w and,t$e' total (population-weighted) probability of alert.for_the
' EPh (PTALL) . The total probability va[ues are listed in'* _
y s s. . . , ;^4 - numeri' cal order for sample scenarios one through four. , i% ( ; . , _ . s :. (, A
." ~ g ._
~
I .
% e'
%. (.' *f
=
s T * \.. ' lk l _
~ ~
. ., \ %
1
~ . ,:
% 1'%
* . %um- t, "%
~ ~_, 'h u
~' s., _
~
w s ( . - a pn n 0 9, , g, A % .
# >=
~
$ k. s . x .. - s . s ... a 1 "N % h. w
- a. **- ],, -
- q. n
- s. ,
f A 4 s' ha , g i t _ as 4 +* g.._ l s I V
.w O b me h
~
h w
\ %
ee b
-A m .x a* #
g ,Y' gr A
, *w A
- w c.\
y v = . . " E rg
% ..A' ,
+s
%~
9 1 m 4 W'
s
- I.
s.
/ TABLE E.1 f
IriDi eti- I Fila
- eIFEtm E1PEtt DH1E X Y 4 PLs100 FT I v1 5Se.150 45&3.25U 250.000 145 2 v2 578.75U 4564.400 530.000 125
~
4 R3 581.400 457d.8bu 830.0VO 125 4 P4 578.750 4577.060 770.000 125 5 P5 575.200 4578.400 83U.000 125 6 R6 573.400 4b?4.250 570.000 125 u 7 R7 576.850 4573.800 990.000 125 8 k8 580.570 4573.070 650.000 125 9 F9 576.820 4567.900 1U70.000 125 10 R10 573.050 4 5'e 9 . U O O 380.000 125
. 11 R11 58e.600 4500.000 200.UUU let 12 Pld 584.SUU 4b7e..UUU 210.UUU 145 l 13 R13 583.750 4571.95U 3?U.000 145 L 14 R14 584.500 45 t,0.12 0 e30.UUU 12b 15 P15 580.120 4554.dOU 560.000 125 le Plo 574.100 -45e.4.380 89U.U00 125 17 P17 576.180 4562.Ibu 1150.U00 125 la R16 579.830 4be4.75U 650.VOU 125 19 R19 581.600 4563.500 435.000 1E5 20 Pau 50d.000 45e.0.100 520.U00 -125
. - 21 R21 b78.880 4560.e8U 550.000 125 22 P22 578.220 4db7.eeu 61U.000 Id5
. 22 v 2,:. 501.050 4557.150 eld.UUU id5 da P24 Sub.Ubu 456b.05U 1eu.UUU 145 25 R25 584.200 4 5 e.J . 7 5 0 170.000 125 de R2e. 505.650 4 bed.450 125.UOU I d'5 27 P27 b84.700 45el.700 280.000 125 i m 28 RES 587.100 4560.3UU 90.000 125
, is 929 584.350 4559.eUU 180.000 145 l 30 P3U 587.USU 4550.450 270.000 125 31 k31 585.000 4556.4eu 180.00U 125 32 F32 586.e20 4bbe. 600 130.UUU 125 33 PJJ 504.750 4555.dbV 215.000 125
; 34 P34 634.700 4554.U50 270.000 125
.. , _ 35 P3t Sc:e.U50 4rda.UUU 2U0.000 lib l
Jo F3e 58v.4UU abbb.400 i d d . U '. U 145 l
- 37 P3T 595.e00 4bbe.150 c.s H. U U n 125
- l. 38 V38 597.b50 45de.800 4,: U.UUU Idb l
59 P39 595.b00 4557. EU d'i v . U U i; 145 40 F40 597.300 4557.900 415.Uvu 125 l 41 P41 59e. 400 4550.700 d60.UUU 145 j 42 R42 594.72U 455F.380 150.000 145
- as V42 597.420 4560.020 53H.000 1 c.5 l' 44 i 4'4 598.250 45b9.500 560.000 125 45 P4b 600.580 456d.950 530.UUU 125 4e F4e 59T.c.UU 45ec. 220 390.000 125 47 r47 595.U50 45el.Ubu deU.000 Idt l Ibumps f
8
TABLE E.1 (Continued)
~~
54 kS4 588.550 456e.100 100.000 leb 55 R55 593.950 45e3.330 430.U00 125 ' ' 56 F5d 59e.400 45et.eb6 330.000 125 57 v57 601.6c0 45e7.000 4eu.UUU 125 - 58 PSE 599.450 4de7.700 498.000- 145 by F59 e0d'.270 45ev.~350 4YU.UUU 125 60 PoO 59.'.300 45ed./00 290.000 125 ~ 61 Rol 601.870 4571.0.'O 470.000 145 ed R6d b98.000 4571.b00 530.000 145 63 he3 595.900 4571.300 490.000 1c5 64 P64 593.100 45ee.7tu 300.000 125 ~_ 65 R65- 590.900 45e7.700 270.000 125 66 966 588.200 45e0.o00 150.000 125 , 67 ke7 589.700 45e9.500 170.000 125 ~ 68 HR68 589.950 4570.700 200.000 125 6s 969 591.350 4571.450 310.000 12b 70 E70 590.100 4572.450 26 0.000 125 71 W71 538.000 4b72.1ru 100.000 125 _J 72 R72 591.e00 4b74.5bu 15U.000 125 73 F73 5y3.000 45 71. Y r' 0 450.000 125 74 P74 594.220 4573.500 370.U00 leb
~
75 P75 b9e.ee0 4574.680 750.000 125 - 7 e. 976 59Y.180 4574.550 540.000 125 77 R77 e01.800 4574.620 675.000 125 78 P78- 600.500 4576.900 650.000 lab _ 79 R79 59e.800 457e.300 450.000 125 80 F80 598.600 4578.2eu 750.000 125 81 F81 595.920 4579.870 510.000 125
~~
82 RSE 593.820 4d77.450 330.000 125 33 VB3 590.400 4b78.e20 570.UUU lib 84 R84 587.170 457e.U70 2b0.000 Id5 SS RS5 508.37U 4b81.350 200.000 145 __ 8e 90 e. SPU./UU 4584.obU 450.000 1P5 a . r . , * . O W e _.a 0 e-se S em.
m.. TABLE E.2 1 *s v. ..- a 2 U2 590.U00 4 bey,v50 150.000, - 3 E3 592.250 4570.8b0 380.000 FAF 4 U4 554.450 45eb.120 170.U00 - 6 KS 5?0.07U 4be4.800 150.UUU peg 6 96 503.720 4564.e00 130.UUU FHP
? PT 583.3du 4bt.4.350 200.000 FAR
-. E US 566.eUU 4 bel.229 30.000 -
9 U9 52e.USO a b e l , c.0 0 30.U00 - 10 V10 583.d50 4tec.100 180.UUn pep FAP 11 P11 694.450 45e4.450 490.000 12 U12 586.820 4561.100 30.U00 - 58d.550 45c.1. c:U U 300.000 FAF l' 13 14 Pls i'14 5Sc.350 abel.Ob0 46.9.100 340.00U 140.000 FAk tiEnF 15 P1b 58e.450 l" 16 R16 589.d20 4580.100 380.UVO ttEAF i FAP 17 RIT 595.850 4bie.150 500.UUU i FAF
! _. 18 P18 596.400 4d74.900 500.000 595.600 4b62. c: 00 3bO.UUU FHP l 19 R19 i 20 U20 594.100 45e2.U50 150.000 -
586.750 4558.850 220.000 FAP I 21 R21 581.070 4560.400 540.UUU tiESP v 22 P22 58e.450 4580.e50 200.000 ftEAF 23 Rd3 t 24 U24 589.ySU 4571.600 400.000 - b98.850 4572.e50 490.UOO tiEAF
' 25 R25 599.050 45e8.520 535.UUU tiEAW j 26 R2e FAP
.' - 27 P27 599.150 45e7.700 360.000 II 28 U28 59d.570 465y.400 38v.000 5 8 d.. V 0 O 4bb7.750 YO.UUU FAP i 29 W29 FAP
, 30 R3U 583.1b0 4:.57.150 282.U00 579.150 4deU.42U 515.UUU FHF 31 v31 FAR
' 32 R32 591.440 4b81.750 '700.000 596.750 4b78.47U 240.UUU FAR J4 F33 000.100 457c 200 44U.UUU tiEAP 34 P34 598.150 45el.eUO 500.000 ttEAR 35 E35 589.350 4555.e00 145.000 tiEAR 3e kJe FAF 37 P37 582.450 4556.e00 430.000 598.850 4b7y.650 840.000 FAF 38 P38
" ~ 602.27U 4570.360 440.000 tiEAF 3y 939 F e,i i- 40 F40 603,290 4b69.220 4eu.UGU 589.Ibu 456,:. 57 U 205.U00 tiEAF 41 P41
- 42 U42 591.100 4570.750 310.000 -
43 U43 5dd.150 4d53.800 25t.000 - 44 v44 o05.200 adds.450 246.UUU - 45 U4b 504.100 4b54.e50 2dU.UUH - 4e v4e 584.15U 4504.c50 cle.UUU 50J.b5U abbb.Ibe 340.000 FnF 47 i.47
" 58e.150 455t,e00 475.vv0 I4EAF 44- A48 g # hem M
/
( TABLE E.3 s
- JCEN:: H P'OL wit 1D NETS NCEC1 F1 F2 F3 F4 .F5 F6 FT FS 1 0.05 153 ~16. '41. .200 .200 .040 .200 .23i .u o-
. oac. .034
'2 U. 81 23 16. 31. .000 .000 .950 .009 .030 .010- .004 . c oe. '
3 U. 49 415 31. 31. .USO .140 .000 . S c. 9 .UUU ,U.,,,.,
,;vg ,g4g ,,,,,
4 0.4e 135 31. 41. .UOU .000 .v50 .UUU .U4v . U t e; .000 .000 INP PU30 PE55 11UL HDD 1 1.OVO 1.000 3.750 -U.350 1.000 1.000 2.25U U.170 3 1.U00 1.000 3.750 -U.170 1- 1.U00 1.000 5.e20 -0.170 , i O (
- t. -
? I
- J i
t i
%d
._~
l ' ~ t
( . TABLE E.4 LISTENEROUTPUT LYS :: LYsVENER NeetE S' IREN I: ' 'I S NENN LhiJt 1 R1 54 R54 83'.3 m. 66 R66 92'.2 66 R66 93.2 54 R54 85~.8
~
E U2 67 R67 98.5
' 68 R68 95'.2 b- 58 R68 95'.9 67 R67 99'.2 3 R3 64 R64 74'.3 73 R73 88'.5 73 R73 89'.9 64 R64 77'.2 l
' . 4 U4 25 R25 87'.9
! 14 R14
! 77'.1 14 R14
^
80'.3 25 { R25 89'.7 5 R5 52 R52 B D'. 9 52 R52 96'. D 52 R52 81'.7
, 52 R52 96~.8 , L '
6 R6
- 25 -
R25 92'.2 25 R25 77.3
! 25 R25 73'.3 25 R25 93'.4 7 R7 25 R25 91".4
- 14 R14 72'.3 i
14 R14 76'.4 25 R25 92'.8 8 UB 28 R28 91'.4 26 R26 87'.4 26 R26 88'.9 28 R2B 92'.7 I r 9 09 28 R28 85'.6 l 26 R26 l '- 93'.8 26 R26 94'.7 28 R2B 87'.7 , s 10 RID 27 R27 92'.6 25 l R25 85'.7 25 R25 87'.4 27 R27 93'.8 l t w
e
~_.
TABLE E.4 (Continued) 11 R11 55 R55 79'.7
, .. 56 RS6 80~. 8
_- 51 R51 67".3 56 RS6 83'.6 t 12 U12 28 R28 93'.8 R26 85'.5 26 26 R26 87.3
! 28 R28 94'.8 L
13 R13 2D R2D 74'.1
- 27 R27 73'.6
[_ 25 R25 65'.7 2D R2D 76'.4 f 14 R14 27 R27 70'. D 27 R27 78'.3 19 R19 76'.7 27 R27 81 *. D L 15 R15 84 R84 76'.1
. 11 R11 93'.2 L \'" 11 R11 94'.1 l 84 R84 B D'.1 16 R16 85 RB5 66~.4 l
B5 RBS 86'.6 85 RB5 88'.2 l 85 RB5 68'.4
~
17 R17 75 R75 87.1
, 82 RB2 8 0~. 6
[_ 82 RB2 83.2 75 R75 89.D 18 RIO 75 R75 103'.4 75 R75 103'.4 75 R75 93'.8
. 75 R75 103'. B
! 19 R19 46 R46 68'.8
'l' 56 55 RS6 R55 77'.7 85'.9 l 46 R46 71'.2 2D U2D 49 R49 1 D5'. 8
, 49 R49 IDS.9 49 R49 106'.1 49 R49 1 01'.2 h__ -_. _ _ __ _ ____. ._
TABLE E.4 (Continued) El R21 3D R3D 98'.9 j 3D R3D 94'. D _ i 3D R3D 84'.5 3D R3D 99'.6 22 R22 2D R2D 92'.2 20 RED 92'.3 19 R19 79^.6 2D R2D 93'.4
~
23 R23 11 R11 96'.3 85 RB5 82'.4 11 R11 77'.1 97'.2 11 ' R11 l 24 U24 68 R68 93^.1 70 R70 93'.7 70 R7D 94'.6 68 R68 94'.2 25 R25 62 R62 89'.7 76 R76 83'.9 ( 76 R76 85'.9 s- 62 R62 91'.3 26 R26 58 R58 92'.9 i L. , 58 R58 73'.1 58 R58 79'. D i 5B R58 94'.1 27 R27 58 R58 104'.2 58 R58 I D4'. 2
' ! 58 R58 99'.6 5B R58 1 D4'. 6 28 U28 41 R41 92'.2 l 42 R42 96'.3 1 42 ' RM2 97'.D 41 R41 93'.5 29 R29 32 R32 87'.9 3D R3D 89'.1 29 R29 82^. 8 3D R3D 90'.5 3D R3D 33 R33 79'.8 ; 29 R29 78'.7 ll~ 29 R29 81'.6 ll ., 33 R33 83'. D ' L_
~..
TABLE E.4 (Continued)
~~
31, R31 21 R21 83'.6 21 R21 98'.7 21 R21 99'.3 22- R22 81'.9 32 R32 83 R83 71'.1 88 R88 58'.6
- 85. R85 62'.7 l 83 RB3 75'.3 L
33 R33 79 R79 64'.6 81 RBI 62'.2 81 RBI 64~.1 79 R79 67.1 34 R34 77 R77 80'.9 78 R78 94'.4 78 R78 95'.2 77 R77 83'.9 35 R35 % RE 73'. D
' % R46 93'.1
% R46 94".1
's-44 R44 74.6 36 R36 36 R36 ID4'.2 36 R36 99'.2 36' R36 99'.5 l 36 R36 104'.5 L
37 R37 31 R31 79'.4 31 R31 79'.7 29 R29 78'.D 31 R31 82'.7 38 R38 BD RBD B7.4 BD RBD 67.6
, BD R8D 69'.1 BD RBD 89'.2 39 R39 59 R59 91'.9 El R61 86'.6 61 REI 88'.2
, 59 R59 93'.2 4D R4D 59 R59 71'. 6 59 R59 81'.7 59 R59 92'. B '
.s ,,
59 R59 72'.9
t s i TABLE E.4 (Continued) c 41 R41 36 R36 55'.3 36 R36 75'.6 36 R36 77.4 _- 36 R36 57.6 42 U42 68 R68 85~. 2
' 68 R68 90'.4 58 R68 91'.6
, 68 R68 71'.7 43 043
.._ 35 R35 92'.8 34 R34 99'.1 34 R34. 99'.6
[_ 35 R35 94'. D 44 U44. 35 R35 71'.1 35 R35 91*.2 i 34 R34 83.5 35 R35 92'.5 45 045 34 R34 85'.5 34 R34 95'.6 33 R33 87.2 34 R34 96'.4 46 U46 34 R34 92'.2 33 R33 95'.6 33 R33 86'.4 33 R33 96'.4
. 47 R47 33 R33 89'.7 33 R33 89.8 33 R33 71'.1 33 R33 91'.2 48 R48 33 R33 79'. D 33 R33 79'.3 23 R23 70.8 33 R33 82.3 49 R49 15 R15 86'.5 23 R23 75'.7 23 R23 77.5 15 R15 88'.5 l I 5D R5D 22 R22 98'.3 22 R22 1 D3'. 3 22 R22 I D3'. 7 l
22 R22 93~.7 j . N_ - I
?
PROBS TABLE E.5
~
P1 P2 P3 84 P5 P6 F7 PB PT
; LYsTkWEk 1 1 T.UDD r.DDD D'. 691 V;989 D'. 986 T.UDD r.DDD r.DDD D'.9822
'. F.UDD r.DDD D'.79D T.DDD r.DDD F.DDD r.DDD 1.DDD D.8001 1.DDD r.UDD D'.625 D.90D r.DDD r.DDD r.DDD r.DDD D'.9441 r.DDD r.DDD D.519 D'.900 r.DDD r.DDD 1;DDD r.DDD D'.5429 LYsvkken 2 J
F.DDD r.DDD D;848 T.DDD .r.DDD r.DDD r.DDD r.DDD D'.9939 T.DDD r.UDD D'.818 l'. DDD r.DDD r.DDD r.UDD 1.UDD D'.8274 r.DDD r.DDD D; 661 U.937 T.DDD r.UDD r.DDD r.DDD U;9645 _ y ,,{.0DD, T.UDD D'. 7DI D'. 998 T.DDD r.DDD r.DDD F.DDD D'.716D tzsTswen 3 I r.DDD r.DDD D;569 D'. 941 D'. 614 T.DDD r.UDD r.DDD D.8823 g i l J T.DDD r.DDD D.751 D'.999 F.DDD r.DDD r.DDD r.DDD D'.763D '
, 1.000 r.UDD D.579 D'.837 T.DDD r.UDD r.DDD r.DDD D'.9089'
! r.DDD r.DDD D'.377 D'. 716 U.7BD r.DDD r.UDD r.DDD D'.3993 i Lisvinck 4 L_ <
1.DDD r.DDD D'. 744 T.UDD r.DDD r.DDD r.DDD r.DDD D'.9898 3 T.UDD r.DDD D'. 61 D D.882 D'.7BD r.DDD r.DDD r.DDD D'.623D l r.DDD r.UDD D'.43D D'. 547 D'.902 T.UDD r.DDD r.DDD D'.7461
~
; r.DDD r.DDD U.576 0;945 r.DDD r.DDD r.DDD r.DDD D.5971
{ LYsTkWEk 5
. F.DDD r.DDD D'.659 U.982 D'.917 T.DDD r.DDD r.DDD D'.9638 f, r.DDD r.UDD D.826 T.DDD r.DDD r.DDD r.DDD 1.DDD D'.8344 L s _,
1.DDD r.DDD D.453 D'.595 D;936 T.DDD r.UDD r.DDD D.7735
. I'.UDD r.DDD D'. 671 U.985 T.UUU r.DDD r.DDD r.UDD D'. 6873 l LisYkWEk 6 L l T.DDD r.DDD U.789 T.DDD r.DDD ,r.DDD r.DDD r.UDD D'. 9916 :
; U.DDD r.UDD D.612 D'.BB4 D'.787 T.DDD r.DDD l'. D D D V;6252 ,
i L r.DDD r.DDD D'.3D9 D'. 304 D'.554 T.DDD r.DDD r.DDD D'. 6105 I
) r.DDD r.DDD D.627 D'. 971 1; DDD' r.DDD r.DDD r.UDD D.6460 ' ~
LYsTkkew 7 h 1.000 T.DDD D'.782 U.VVD r.DDD r.DDD r.DDD 1.UDD D'.9913 . l-l r.DDD r.DDD D.54D D'.772 D'. 485 T.DDD r.DDD 1.DDD D'.5480 3 L- k' r.UDD r.DDD D'.364 D'. 4DB D;743 U.DDD r.DDD r.DDD D'.6686 h 1.DDD r.DDD D;619 D'.968 r.DDD r.DDD 1.DDD r.UDD D.6377 l 1 LisYkkkk 6 1 T.UDD r.DDD D;781 T.DDD r.DDD r.DDD r.DDD r.DDD D'.9912 { 1.000 T.DDD r.DDD r.DDD D.738 D'.565 U.994 D'.814 T.UDD T.DDD r.DDD r.DDD r.DDD r.DDD 1.DDD r.DDD D'.7511 D.8960 T.DDD r.DDD D'. 618 D'.967 T.DDD r.DDD r.DDD r.UDD D'.6368 LYsVE5Ek 9 s T.UDD r.UDD D. 717 T.DDD r.DDD r.DDD r.DDD r.DDD D'.9887 T.DDD r.DDD D'. 806 T.UDD r.DDD r.DDD r.DDD r.DDD D'.8153
, r.DDD r.DDD U.645 D'.922 T.UDD r.DDD r.DDD r.DDD D'. 9561 l F.UDD r.DDD D'.547 D'. 924 U.DDD r.DDD r.DDD r.DDD D.5693 LikYanen ID
! r.DDD r.DDD D.794 T.UDD r.UDD r.DDD r.DDD 1.DDD D'.9917 .
]
r.UDD 1;DDD r.DDD l'. DDD 0;71B D'.543 D'. 974 D;776 T.DDD T.DDD r.DDD r.DDD r.DDD r.DDD r.DDD r.DDD D'. 7324 D'.8746
/ d T.DDD r.DDD D'.633 D'. 973 U.DDD r.DDD r.DDD 1.UDD D'.65D9
. - - r-, , - ---
TABLE E.5 (Continued)
~ f LziriiWen 11 ~-
1;DDD 1;DDD V;645 D.978 DiBBS l'.DDD l'. DDD l'. DDD 1;DDD l'. DDD U.659 U.935 D.917 U.9549 1;DDD l'. DDD 1;DDD r.DDD D.6739 r.DDD _' r.DDD r.DDD D'.195 D;484 D'.146 D'.B64 U.DDD U.992 r.DDD U.UDD r.DDD r.DDD r.UDD r.DDD D'.5220 ! Listuksk 12 D'.5D91 T.DDD r.DDD D.BD5 T.DDD r.DDD F.DDD 1;DDD r.DDD r.DDD r.DDD D.9922 D'. 716 D'.973 T.DDD r.DDD ' ~ F.DDD r.DDD D'.540 r.DDD r.DDD D'.73D2 U.771 1.DDD r.DDD r.DDD r.DDD r.DDD r.DDD D;646 U.978 U.872D LiSTEWEh 13 IkDD9- U.DDD r.DDD r.DDD D;6639 L_ r.DDD r.DDD D; 567- D'.939 D'.6D5 T.DDD T.DDD r.DDD r.DDD r.DDD D~.8798 D'. 5 6 1 U.008 D;576 r.DDD r
' r.DDD r.DDD r.DDD r.DDD D.5698 D.162 D;116 D'. DD D r.DDD r.DDD i
[ , ,' U. D D D, 1.DDD D'.364 U.694 U;741 r.UDD D'.5047 t.xsrmwan 14- r.DDD r.DDD r.DDD D'.3853 r.DDD r.DDD D.621 U.969 D;818 1;DDD I r.DDD r.DDD r.DDD r.DDD D.9369
' D.626 U.901 V;832 r.DDD r.DDD r.DDD r.UDD D'.369 D;418 r.DDD D'.6393 D'. 757 T.DDD r.DDD r.DDD T.DDD r.DDD D'. 442 D;812 U.921 D.6739 U.DDD r.DDD r.UDD D.4666 i LrstkWch 15 r.DDD r.DDD D;596 V;957 D'. 726 U.UDD r.DDD r.DDD D'.9123 r.DDD r.DDD D'.799 U.DDD r.DDD U.DDD r.UDD r.DDD r.DDD D.BD94 r.DDD D;637 D;914 U.UDD r.DDD i
r.DDD r.DDD D;427 r.DDD T.DDD D'.9516
~' B.792 D;897 T.UDD r.DDD r.UDD Lisrkksk 16 D;4518 r.DDD r.UDD D'. 448 D'.82D 0;DDD j f.DDD r.DDD 1;DVD r.DDD D'.712D '
r.DDD D.729 D; 981 U.UDD r.DDD T.DDD r.DDD r.DDD r.DDD D'.7425 D'.555 D;797 U.DDD r.DDD r.UUU T.DDD Y.DDD D'.215 D;521 r.DDD D'.8864 D'. DDI r.UDD r.UDD r.UUU D'.2147 l Lrsvesen 17 L r.DDD r.DDD U.735 r.DDD r.DDD r.DDD r.UUU r.DDD D'. 9894 U.UDD r.DDD D'.657 U.933 D;911 T.UDD r.UDD r.UDD D;6710 U.UDD r.DDD D'.477 D.646 U.979 U.DDD r.DDD U.DDD r.UDD l'. DDD D.8018 D'.566 D.938 T.DDD r.DDD r.DDD
~
Lrsruwkk 18 r.DDD D'.5873 F.UDD r.UDD D'.887 T.DDD r.DDD r.DDD r.DDD r.DDD D'.9955 T.UDD r.DDD D'. BBB r.DDD r.UDD r.DDD T.UDD r.DDD r.DDD r.UDD U;8933
- D.632 U.9D9 T.UDD r.UDD r.DDD r.DDD r.UDD r.DDD D'.754 T.DDD r.UUU U.9489 r.DDD r.DDD r.DDD U.7664 5 LrstEkkk 19
_ r.DDD r.DDD V;487 U;867 U.DD4 U.DDD T.DDD r.UDD r.UDD r.DDD D.7239 D'.617 U.891 D.8D4 U.UDD r.DDD r.DDD T.DDD r.DDD D.52D D'.63D4 D'. 733 r.DDD r.DDD r.DDD r.DDD D.85D4 r.DDD r.DDD D'.269 0.572 D'. 413 F.UDD r.DDD r.DDD U.282D Li.sYkkkk 20 T.DDD r.DDD U;9DS r.DDD r.DDD r.DDD r.DDD r.DDD U;9D5 r.DDD r.DDD D.9962 T.DDD l '. UDD r.DDD 1.DDD r.DDD T.DDD r.UDD D.779 r.DDD D'.9D96 r.DDD r.DDD r.DDD r.DDD r.UDDD r.DDD 1.DDD D'.724 T.DDD r.DDD 1~. 000 1.D00 r.DDD U.7379 WWum
l 7
. TABLE E.5 (Continued)
"EYkreksh~2T "
- 1;DDD 1;DVD D.852 1;DDD l'.DDD 1;DDD 1;DDD l'.DDD D.9941 1;DDD 1;DDD U.BD7 l'. DDD l'.DDD 1;DDD 1;DDD 1;DVD D;B17D l'. DDD l'.DDD U.499 D;691 U.999 1". D00 l'.UDD 1'.000 U.8272 1;DDD 1~.000 D;7D6 D'.999 1;DDD 1;DDD 1;DDD
_ 1 ~. 000 D.7205
- Li'STENEM EE ' '
l'. DDD 1;DDD U.789 1;DDD 1;DDD 1;DDD l'. DDD 1;DDD D;9916 1;DDD l'.DDD D'.790 l'. DDD 1;DDD l'. UDD 1;DDD 1;DDD D'.BUDB 110D0 l'.DDD D'.419 D;522 D'. 881 1;DDD 1;DDD l'.DDD D;7322 1;DDD 1;DDD 0; 627 D;971 1;DDD l'.DDD l'.DDD 1;DDD 016459
, Li's taban 23
- 1;DDD l'. DDD D; B29 1100D 1;DDD 1;DDD 1;DDD l'. DDD 0;9932 1;DDD 1;DDD D;679 D'. 951 U.95D l'. D DD l'. DDD l'.DDD D'.6935 j 1;DDD 1;DDD D'.376 U;433 D;779 1;DDD l'. DDD l'.UDD 0;6824 L l'.DDD 11D00 D;676 D; 987 1;DDD 1;DDD 1;DDD l'.DDD D'.6925 LisTEksk 24 1;DDD l'. D D D D.790 1;DDD 1;DDD 1;DDD 1;DVD l'. DDD D'.9919 1;DDD l'.VDD D.BD4 1;DDD 1;DDD 1;DDD 1;DDD 1;DVD D'.8138 1;DVD 1;DDD D'. 643 D'.920 1;DDD 1;DDD 1;DDD 1;DDD D*.9550
, 1.UDD 1;DDD D'. 638 D;976 1;DDD 1;DDD 1;DDD l'. DDD D.6564 ! LisYENan 25 1;DDD l'. DDD D'.764 1 *. DD D 1;DDD 1;DDD 1;DDD l'.UDD D;9906 l l'. DDD 1;DDD D.697 D;963 U.996 1;DDD 1;DDD 1;DDD D'.7118 1;DDD 1;DDD D.519 D'.732 l'. DDD l'. DDD 1;DDD l'.UDD' D;B5DD
,,,;.1DUD,,, 1". UD D D'. 590 D'. 958 1;DDD 1;DDD 1;DDD l'.DDD D'. 6181 i
LISTENEN 26 i 1;DDD l'. D D D D.797 1;DDD l'.DDD l'. D DD 1". 0D0 1;DDD D'.9919
! 1;DDD 1*.000 D.552 0;793 D'. 536 1;DDD 1;DDD l'. UDD D'.56D4 l 1;DDD l'.DDD D.4D9 U.5DD D;B6D 1;DDD l'. D D D 1;DDD D.7202 l 1;DDD l'.DDD U;637 D'. 975 l'. DDD 1;DDD l'. DDD l'.DDD U.6547 I r Li'YENEN s 27 L_ l 1;DDD l'.DDD 0;B93 1 '. DDD 1;DDD 1;DDD 1;DVD 1;DDD D;9957 1;DDD 1;DDD D*.894 1;DDD 1;DDD l'. DDD 1;DDD 1;DDD D'.0989 l 1;DDD l'.UDD D'.7DS U.968 I'.DDD l'. DD D 1;DDD l'. UUU D'. 9821 1.DDD l'.UDD D.762 l'. DDD 11000 1;DDD 1;DDD 1;DDD D'.7743 ListENEn 28
- 1;DDD l'.DDD U.79D 1;DDD 1;DDD l'. D DD 1;DDD l '. D D D D;9916 1'.000 1*.000 D'.828 1;DDD l'. DD D l'. DDD l'.DDD 1;DDD D'.837D l'. D D D IkDDD D'. 6 7 4 D'. 947 l'. UDD l'. DD D l'.DDD l'. DD D D'.9704 l'. D D D 1;DDD D'.628 D.972 1;UDD 1;DDD 1;DDD 1;DDD D'.6469 ListrNem 29 l'. DD D 1".DDD 0;744 1;DDD 1;DDD 1;DDD l'. DDD l'. U D D U.9898 1 *. D 00 l'.DDD D;757 l'.DDD 1;UDD 1;DDD l'. DDD l'. D D D D'.76BB 1;DDD l'.DDD U. 471 U.634 D'.96D l'. DDD l'. DDD 1;DDD D.7952 1;DDD 1;DVD D'.5BB D;952 1;DDD .1'.000 1;DDD l'. D D D D.6DB3 Li'stEwen 30 1;DDD 1;DDD D.646 D'.978 D;BB9 1;DDD l'. UDU 1;DDD 0;9560 i 1;DDD 1;DDD D;631 D;9D7 D'.849 l'. DDD 1;DDD 1;DDD D'.6453 1- 1;DDD 1;DDD D'. 451 D'.592 D'. 934 1'.000 1*.000 1'.000 D'. 7714
, 1'.000 .
1;DDD D*. 475 D'. 853 D'. 973 1;DDD 1.000 1'.000 D'.5000 a . v !l ~
,--.n-,,.--.. .----,..,.---,,-.-...-------------.~.,,--.-_.,,nn,-,.--.v---n--
,.r--- --- , - - .---------..,n------- , - .
j .'
- l-TABLE E.5 (Continued)
CYstr4Em 31 F.DDD r.DDD D.694 U.992 U.993 U.DDD r.DDD r.DDD D.9846 r.DDD r.DDD D;850 r.DDD l'.DDD r.DDD r.DDD r.DDD V;8577
~
r.DDD r.DDD U.7D2 D'.966 U.UDD r.DDD r.DDD r.DDD D.9810 __ U.UDD r.DDD D;456 U.831 D.94D r.DDD r.DDD r.DDD D.4BD9 ListEnan 32 r.DDD r.DDD D.523 U.9D4 U;4D9 U.DDD r.UDD r.DDD D'.8258 l'.DDD r.DDD D'.313 D;312 U.DDD r.DDD r.UDD r.DDD D.3174 U.DDD r.DDD D.10D D;D71 D.DDD r.DDD l'. U DD r.UDD D'.4796 T.UDD r.DDD D'.344 U.662 D'.679 r.DDD r.UDD r.DDD D'. 3642 e LikTENEn 33 i r.UDD r.UDD D'.419 U.779 D'.DDD r.DDD r.DDD r.DDD D.7026 r.UDD r.UDD r.DDD r.DDD r.DDD D.3788 U.378 0;435 D.DDD ii r.DDD r.DDD D.129 D.D90 V;DDD r.DDD r.DDD r.DDD D'.49D1 . L_ r.DDD r.DDD D'.19D D.49B D.DDD 1.DDD r.DDD r.DDD D.1908 , Liktknen 34 e U.DDD r.DDD D; 66D D'.982 D;918 T.DDD r.DDD r.DDD D.9640 r.DDD r.UDD D'.811 U.DDD r.UDD r.DDD r.DDD r.DDD D'.82D4 4 1;DDD r.DDD D'.651 D'.928 T.DDD r.UDD r.UDD r.DDD D.9597 r.DDD r.DDD D'.488 0;869 D'.996 T.DDD r.UDD r.DDD 0;5134-Listehka 35 T.DDD 1;DDD D*. 551 D'.928 D'.533 U.DDD l'. UDD r.DDD D'.86D2 i U.DDD r.DDD D'. 799 U.DDD r.DDD r.DDD r.DDD r.DDD D'.8DBB r.DDD r.DDD D'.636 D'.913 r.DDD r.DDD r.DDD r.DDD D'. 9511 T.DDD r.DDD D'.332 D'. 64D D.635 T.DDD r.UDD r.DDD D;35D4
~' Lrstskan 36 r.DDD r.DDD D.893 T.DDD r.DDD r.DDD r.DDD r.UDD . D'.9957
, F.DDD r.DDD D.854 T.DDD r.DDD r.DDD r.UDD r.UDD D.8615
} , -- U.DDD r.DDD D;7D5 D'.968 r.DDD r.DDD r.DDD r.DDD U.9819
' l; r.DDD r.DDD D'. 762 U.DDD r.DDD r.DDD r.DDD r.DDD U.7738 i:! LisTaman 37 __ r.DDD r.DDD D;641 D.976 D'.875 r.DDD r.DDD r.DDD D'.9520 r.UDD r.DDD D.645 D'.922 D'.886 U.UDD r.DDD r.DDD D'.6594 ('
- , F.DDD r.DDD D;391 0;463 D.818 1;DDD r.DDD r.DDD D'. 6991 U.DDD r.DDD D;469 D'. 947 D'.955 U.DDD r.DDD r.DDD D'.4940 Listawan 38
! r.DDD r.DDD D.738 U.DDD r.UDD r.DDD r.DDD r.DDD D'.9895 '. U.DDD r.DDD D'. 468 D'.627 D'. DD D r.DDD r.DDD r.DDD D'.4643 r.DDD r.DDD D'. 229 U.186 D'.D15 U.UDD r.UDD r.DDD D'.5444 r.DDD r.DDD D'.569 D'. 941 U.DDD r.DDD r.DDD 1.000 D'.5909 LikTance 39 l l r.DDD r.DDD D.786 r.UDD r.DDD r.DDD r.UDD r.DDD D'.9915 T.UDD r.DDD D.729 D; 981 U.DDD r.DDD r.UDD r.UDD D.7425 r.DDD r.DDD D'.554 D'.797 T.DDD r..DDD r.UDD r.DDD D'.8863 U.DDD r.DDD V; 624 D;97D r.DDD r.UDD r.DDD r.DDD D;6429 Listaman 4D U.DDD r.DDD D;530 D'. 91 D D;439 1.000 r.DDD r.UDD U.8341 T.DDD r.DDD D'.67D D.945 D.937 T.DDD r.DDD r.DDD D.6849 r.DDD (.DDD D'.619 U.893 U.DDD r.DDD r.DDD r.DDD D'94DD r.DDD r.DDD D'. 301 D;602 U.526 U.DDD r.UDD r.DDD 0;3167 v -
---a--------- --n.rwry,----www w-e y ,----,.-----g-.--.w,,-,,.y-,.--,. - , , ,,,,,--m-eew, --w--e-w-w,e-r-i-we.wwwe.w-- -r---+-seumww -me-ee---- -e > - ----,-w-ex--s--w--- -rimr--w-i.w.ew--e.w
f
, C' TABLE E.5- (Continu2d) g .. .-. .. .. . _ _ . . .- .
D'.751 U.DDD D.253 D'. 557 D.DDD r.DDD r.UDD r.DDD r.UDD r.DDD U.588 0;B52 0;695 D'.6DIB r.DDD r.UDU r.DDD D'.5998 U.DDD r.DDD D;382 U.443 D;793 U.DDD r.DDD ! D.836 r.DDD U.6BB2 j T.DDD D'.DDD D.236 U;DDD r.DDD r.DDD r.DDD
, LYstkkkh 42 ' '
D.DIDD [ U.DDD r.DDD D; 713 U.DDD r.DDD .1;DDD 1;DDD., r.DDD .F,9BB5..,
~
1
-W.1DD -' U. DDD ' T. 77T '"13DU'T. DDD " U.'0DD',,r. UDD r.DDD U.7824 T.DDD r.DDD D'.6D3 U.872 r.DDD ILDDD l'.DDD r.DDD r.DDD D.92B3 r.DDD D~.279 D;581 U. 4' 48 T.DDD r.DDD r.DDD LYsikkch 43 '
~
D'.2930
- 6 s 1;DDD 1;DDD U;796 r.DDD r.DDD 1;DDD r.DDD T.DDD r.DDD l'. D D D D.9918 D.853 1;DDD r.DDD r.DDD 1;DDD r.UDD D.8606
} r.DDD r.DDD D.7D6 V; 968 T.UDD 1;DDD r.DDD r.DDD L [ Y.DDD r.DDD D'.9823 D.635 D.974 r.UDD r.UDD l'. D D D r.DDD D'.6534 Lisikwon 44
; ; 1.DDD r.DDD D;523 D'.9D4 D'. 4DB r.UDU Y.DDD r.UDD D'.8256 L r.DDD r.DDD D.7BD r.DDD r.DDD r.DDD 1;DDD r.UUU D;79DB
' - 1;DDD . r.UDD D.483 0;659 U.991 U.DDD r.DDD r.UDU D'. B DBB r.DDD r.DDD D.615 D.966 T.UDD 1;DVD 1;DVD r.DDD D'.6339
, .__Ysiti2W~45 L /
,- T.DDD r.DDD D.717 U.DDD r.DDD r.DDD r.DDD r.DDD D;9BB6 T.DDD r.DDD D.822 U.DDD r.UDD r.DDD 1;DDD r.UDD D.8314
-l T.DDD 1.DDD D'.539 U.769 T.DDD r.DDD r.DDD r.UDD D.8705 L F.DDD r.DDD D.667 U.984 T.DDD r.DDD r.DDD r.DDD D.6835 LisYrkkn 46 i
T.DDD r.DDD D'.79D r.DDD 1;DDD r.DDD r.UDD r.UDD D;9916
.' ~
1.000 r.DDD D.822 U.DDD r.DDD r.DDD r.DDD r.DDD D.B314 i
..- F.DDD r.DDD D;527 D'. 746 T.UDD r.DDD r.UDD r.DDD l j.,
' D'.858U r.DDD r.DDD U;667 D'. 984 U.UDD 1;DDD r.DDD 1;DDD D.6835 LYstewkh 47 '
'L w.. T.UDD r.DDD D'. 763 1 *. DD D l'. DDD r.DDD r.DDD r.DDD U.9905 r.DDD r.DDD D'.765 r.DDD r.UDD r.DDD r.UUU r.DDD U.DDD l'. DDD D.7765 D.267 D'. 23B U.407 T.DDD r.DDD r.UDD D.5731
_ l T.DDD r.DDD U.597 D'.957 T.DDD r.DDD r.DDD 1*.000 016172 i i LYstruen 48 . I r.DDD r.DDD D'.635
~
U;974 0.059 f.' BUD l;DVD r.DDD i D'.9478 I 1".000- T.DDD' D.639 D;916 D*. B 71 U.DDD r.DDD 1;DDD
" t
~
T D~.6537 D'. 262 D'.23D D'.39D 1;DDD 1;DDD r.DDD D.569D l r DDD DDD.r.DDD r.DDD D.464 D'. 84D D'.949 f.UDU r.DDD r.DDD D'.4885 LisYkkak 4'9 '
, _ r.DDD r.UDD' U.729 T.DDD r.DDD r.DDD 1; DDD r.DDD D.9891 F.DDD r.DDD D'.591 U.855 D'. 7D4 l'.DDD r.DDD r.DDD D.6D22 F.DDD r.DDD U.383 U.-446 D'.796 U.DDD r.UUU r.DDD D'. 6896 t
l'.DDD r.DDD D'. 559 U.934 T.DDD r.DDD r.DDD 1;DDD D;5 BID Listaken 50 F.DDD r.DDD U.846 T.UDD r.DDD r. BUD f.DDD r.DDD D'.9939 T.DDD r.DDD D.BB7 U.DDD r.UDD 1.DDD r.DDD r.DDD U.8924 U.DDD r.UDD U.752 r.DDD r.DDD r.DDD r.UDU 1.000 D'.9997 T.DDD r.DDD D'.631 0.973 U.DDD 'U.DDD r.UDD r.DDD 0;6496 hUnnLi bWahn korbLhYYoipis ? 14645'4411D929 a .
>Ynun >Ybbs
. ' 4 . ; c. .
>YhLL
- D'. 931 D'. 979 D.951 -
D'. 7D1 D'.BDD D". 744 1
~~
! D.776 0;908 U.833
! D.527 5;629 9.571 *
.. " E i.
. ..s ., . ,
i e m n . a .
.e . ,;a , . .u,,. . . . .u .
. ..,...m..-- - -- - - -
_ . . . n . nn_ --- A- - - - - -
- E NUREG/CR-2655
~
?NL-4226 AN
.}
Evaluation of the Prompt Alerting Systems at Four Nuclear Power Stations 'l Manuscript Completed: February 1982 Date Published: March 1982 j Prepared by D. A. Towers, BB&N G. S. Anderson, BB&N D. N. Keast, BB&N J. L. Kenoyer, PNL Technical Coordinator A. E. Desrosiers, Ph. Project Manager i Bolt Beranek and Newnan, Inc. i
' 10 Moulton Street Cambridge, Massachusetts 02238 Pacific Northwest Laboratory Richland, Washington 99352 Prepared for Division of Energency Preparedness Office of Inspection and Enforcement U.S. Nuclear Regulatory Commission Washington, D.C. 20555 NRC FIN No. B2311 1
i n! F
SUMMARY
The purpose of this study was to provide examples of the analytical procedure developed in PNL-4227 for. the evaluation of the effectiveness of siren systems for alerting the public in the vicinity of a nuclear power plant. Evaluations of the prompt alerting siren systems at-four U.S. nuclear power facilities are presented in this report. These facilities are Trojan, Three Mile Island, Indian Point, and Zion. Site-specific information was used for each system evaluation. The analytical procedure is sunmarized and details of computations for each evaluation are given. o tii
ABSTRACT This report presents evaluations of the prompt notification siren systems at the following four U.S. nuclear power facilities: Trojan, Three Mile Island, Indian Point, and Zion. The objective of these evaluations was to provide examples of an analytical procedure for predicting siren-system effectiveness under specific conditions in the 10-mile emergency planning zone (EPZ) surrounding nuclear power plants. This analytical procedure is discussed in report number PNL-4227. l l V
I FOREWORD i The work presented in this report was prepared by Bolt, Beranek and Newman Incorporated under subcontract No. B-A2740-A-V which, in turn, was funded under a Related Services Agreement with the U.S. Department of Energy Contract DE-AC06-76RLO 1830.
*e l
i I l l l l i i i l N vil l l - - - - - - .- . -.
TABLE OF CONTENTS page Section 1. INTRODUCTION.................................... 1-1
- 2. EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE TROJAN NUCLEAR POWER STATION. . . . . 2-1 2.1 Estimating Siren Sound Levels Out of Doors at Listener Sites............. 2-1 2.2 Estimating Indoor Sound Levels of Sirens. . . 2-6 2.3 Assumptions about Chance of Alert.......... 2-8 2.4 Alerting People Ou t of Doors. . . . . . . . . . . . . . 2-10 2.5 Alerting People Indoors................... 2-15 2.6 Alerting People in Motor Vehicles. . . . . . . . . 2-19
- 3. EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE THREE MILE ISLAND NUCLEAR POWER STATION......................................... 3-1 3.1 Estimating Siren Sound Levels Out of Doors at Listener Sites............. 3-1 3.2 Es tima ting Indoor Sound Levels of Sirens. . . 3-7 3.3 Assumptions about Chance of Aler t. . . . . . . . . . 3-9 3.4 Alerting People Out of Doors. . . . . . . . . . . . . . 3-11 3.5 Alerting People Indoors................... 3-16 3.6 Alerting People in Motor Vehicles......... 3-18
- 4. EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE INDIAN POINT NUCLEAR PCWER STATION. . . . . . 4-1 4.1 Estimati'ng Siren Sound Levels Out of Doors at Listene r Sites. . . . . . . . . . . . . 4-1 4.2 Estima ting Indoor Sound Levels of Sirens. . . 4-5 4.3 Assumptions abou t Chance of Alert. . . . . . . . . . 4-8 4.4 Alerting People Out of Doors. . . . . . . . . . . . . . 4-10 4.5 Alerting People Indoors................... 4-15 4.6 Alerting People in Motor Vehicles......... 4-19 Ix
TABLE OF CONTENTS (Cont.) Page
- 5. EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE ZION NUCLEAR POWER STATION. . . . . . . . . . . . . . 5-1
' .5.1 Estimating Siren Sound Levels
~
Out of Doors at Listener Sites............. 5-1 5.2 Estimating Indoor Sound Levels of. Sirens. . . 5-5 5.3 Assumptions about Chance of Alert.......... 5-8 5.4 Alerting People Out of Doors. . . . . . . . . . . . . . . 5-9 5.5 Alerting People Indoors................... 5-17 5 '. 6 Alerting People in Motor Vehicles......... 5-23 . APPENDIX A. Population-Weighted Random Selection of Listening Points at' Trojan Site..................A-1 B. Test Cases (Sample Scenarios) for Trojan Site.............................................B-1 C. Siren Locations for Trojan EPZ...................C-1 . D. estimeeien of A,,,................................D-1 . E. Dependence of Alert Upon Siren Duration. . . . . . . . . . E-1 F. Random Selection of Population-Weighted Listening Points at Three Mile Island Site.......F-1
- G. Test Cases (Sample Scenarios) for Three Mile Island Site.......................~..........G-1 H. Siren Locations for Three Mile Island E PZ . . . . . . . . H- 1 I. Analysis Input / Output Data for Three Mile Island......................................I-l J. Random Selection of Population-Weighted Listening Points at Indian Point Site............J-l K. Sample Scenarios for Evaluation of Siren Alerting at Indian Point. Site....................K-1 L. ' Siren Locations for Indian Point EPZ.............L-1 M. Analysis Input /Cutput Data for Indian Point. . . . . .M-1 .
N. Random Selection of Population-Weighted Listening Points at Zion Site....................N-1 O. Sample-Scenarios for Evaluation of Siren Alerting at Z i o n . . . . . . . . . . . . . . . . . .' . . . . . . . . . . . . . . . O- 1 P. Siren Loca t ions for Z ion EPZ . . . . . . . . . . . . . . . . . . . . . P-1 Q. Analysis Input / Output Data for Zion............. 0-1 X
LIST OF FIGURES page FIGURE 2-1. Flow of compu ta tion s . . . . . .'. . . . . . . . . . . . . . . . . . . . . . 2- 9 2-2. Chance of alert for awakening people asleep. . . . 2-10 2-3. Distributions for detection b ou t of doors . . . . . . . . . . . . . . . . .y people
.................. 2-14 2-4. Chance of alert for people indoors at home
( 4-minu te s ta tiona ry siren ) . . . . . . . . . . . . . . . . . . . . 2-17 2-5. Chance of alert for people indoors at home (4-minute rotating stren)...................... 2-18 3-1. Flow of computations........................... 3-10 3-2. Chance of alert for awakening people' asleep.... 3-12 3-3. out of doors......................ple
. Distributions for detection by peo
............. 3-15 3-4. Chance of alert for people indoors at.home
( 4-minu te sta t iona ry siren ) . . . . . . . . . . . . . . . . . . . . 3-19 3-5. Chance of alert for people indoors at home (4-minute rotating siren)...................... 3-20 3-6. Chance of alert for people indoors at work in commercial / ins titutional establishments. . . . . 3-21 4-1. Flow of computations............................ 4-9 4-2. Chance of alert for awakening people asleep.... 4-11 4-3. Chance of alert for people indoors ( 4 -m i n u te ro t a t in g s i re n ) . . . . . . . . . . . . . . . . . . . . . . 4 -14 l 4-4. Chance of alert for people indoors at home i ( 4 -mi n u t e ro ta t i ng s i re n ) . . . . . . . . . . . . . . . . . . . . . . 4 -17 l 4-5. Chance of alert for people indoors at work in commercial / institutional establishments..... 4-18 5-1. Flow of computations........................... 5-10 5-2. Chance of alert for awakening people asleep. . . . 5-12 5-3. Chance of alert for people outdoors (4-minute stationary siren).................... 5-15 5-4. Chance of alert for people outdoors (4-minute rotating siren)...................... 5-16 5-5. Chance of alert for people indoors at home ( 4-minute stationary siren) . . . . . . . . . . . . . . . . . . . . 5-20 xi
s ( LIST OF FIGURES (Cont.) Page 5-6. Chance of alert for. people indoors at home (4-minute rotating-siren')....................... 5-21 e, 5-7. Chance of alert for people indoors at wor'k i.n commercial / institutional establishments..... 5-22 9 3 e s. t xii
LIST OF TABLES page TABLE 2.1 Summary of Trojan siren system eva l u a t ion re s ul t s . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 2-2 2.2 Meterological conditions for the four sample scenarios used to evaluate the Trojan siren system............................. 2-5 2.3 Atmospheric attenuation (A a caused by wind and temperature grah$e)nts............... 2-7 2.4 Siren alerting ability for generalized categories of outdoor environments. . . . . . . . . . . . . 2-13 2.5 Assumed activities and background noise
. environments for people . indoors................ 2-16 2.6 Siren alerting for motorists. . . . . . . . . . . . . . . . . . . 2-20 3.1 - Summary of TMI siren system. evaluation results.......................................... 3-2 3.2 Meteorological conditions for the four scenarios used to evaluate the TMI siren system.......................................... 3-6 3.3 Calculation of atmospheric attenuation, A atm' caused by wind and temperature gradients (see appendix D for details).................... 3-8 3.4 Siren alerting ability for generalized l
ca tegories of outdoor environme nts. . . . . . . . . . . . . 3-14 - l 3.5 Assumed activities and background noise l environments fet people indoo rs . . . . . . . . . . . . . . . . 3-17 3 .' 6 Siren alerting for motorists................... 3-22 4.1 Summary of Indian Point siren system l evaluation results.............................. 4-2 1 l 4.2 Meteorological conditions for the four sample scenarios used to evaluate the Ind ian Point siren sys tem. . . . . . . . . . . . . . . . . . . . . . . 4-6 4.3 Calculation of atmospheric attenuation, Aatm, I caused by wind and temperature gradients ( see appendix D for details ) . . . . . . . . . . . . . . . . . . . . 4-7 i 4.4 Minimum background noise levels for generalized categories of outdoor environments (see Fig. 4-3 for distributions)............... 4-13 s xiii
I w LIST OF TABLES (Cont.) page 4 4.5 Assumed activities and bacground noise environments for people indoors................ 4-16 4.6 Siren alerting for motorists................... 4-20 5.1 Summary of Zion siren system evaluation results......................................... 5-2 5.2 Meteorological conditions for'the four sample scenarios used to evaluate.the Zion siren system............................... 5-6 5.3 Calculation of atmospheric attenuation, A a tm ' caused by. wind and temperature gradients ( see Append ix . D for de tails ) . . . . . . . . . . . . . . . . . . . . 5-7 5.4 Minimum background nois'e levels for genera-lized categories of outdoor environments (see Figs. 5-3 and 5-4 for distr'ibutions)...... 5-13 5.5. Assumed activities and background noise
~
environments for people indoors................ 5-18 5.6. Minimum background noise levels for gener-alized' categories of indoor activities / environments................................... 5-19 5.7. Siren alerting for motorists................... 5-24 I Xiv-
- 1. INTRODUCTION This report presents evaluations of the prompt alerting siren systems at the following four U.S. nuclear power stations:
Trojan (Oregon) Three Mile Island (Pennsylvania)
+
Indian Point (New York) Zion (Illinois) The purpose of these evaluations was to aid in the develop-ment of an analytical procedure for predicting siren-system ef-fectiveness under defined conditions in the vicinity of nuclear power plants. The resulting analytical procedure is outlined in a separate report [1]. Because the analysis procedure evolved over the course of the four above evaluations, some of the calculations differ somewhat from one site evalua' tion to another. For example, all of the calculations for the Trojan evaluation were done manually. As a result, some broad generalizations regarding estimations of parameters such as shielding, air absorption, and a tmospheric attenuation were adopted in order to render the analysis trac-l table. Subsequent evaluations employed a computer program, resulting in more precise calculations. In addition, the ana-lyses for Trojan and Three Mile Island employed many different background noise level categories for the estimation of siren detectability outdoors. However, the results of these evalua-tions indicated that such a specific description of outdoor noise environments was not warranted, and therefore subsequent analyses used a more generalized estimation procedure for outdoor detecta-bility. i s l l-1 t
"'~.
e u_ - ] As a result of the evolutionary process described above, the results of the four evaluations, although quite similar, a're not directly comparable. With this qualification, the remaining sections of ,this report summarize the methods and results of the four evaluations. - o
..,8
+
ar a t e d 1-2 j
o i-2. EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE TROJAN NUCLEAR POWER STATION This section summarizes the evaluation of the siren alerting system for the Trojan Nuclear Power Station. The procedure that was used consists of a detailed analysis of siren alerting capa-bility at each of 50 randomly chosen listener locations, under four different " sample scenario" conditions. The random selec-tion process for listener sites is described in Appendix A and the four test cases (sample scenarios) are included in Appendix B. The analysis is based on siren location as sh~own on maps provided in Appendix C. The results of the evaluation fo'r Trojan are summarized in Table 2.1 and indicate that the chance of alert is estimated to vary between 65% and 100%, depending on the sample scenario under consideration. The remainder of this section' describes the procedure used to arrive at this conclusion. 2.1 Estimating Siren Sound Levels Out of Doors at Listener Sites The first step in the procedure is to determine the siren in the vicinity of each selected listener site that is expected to produce the highest sound level at that site for each sample scenario. This choice is not always obvious, because the sound level caused by a particular siren at a given listener site de-l pends not only on the sound output of the siren and its distance from the listener, but also on shielding and atmospheric effects (particularly wind direction). Therefore, it'is generally neces-sary to evaluate several sirens in the vicinity of each listener site in order to determine the dominant one. As a general rule, the closest, highest-rated, nonshielded sirens are selected for evaluation at each site. Furthe rmore , sirens should be chosen such that they are distributed north, south, east, and west of the site (or in any other four mutually perpendicular directions) 2-1 l
r z- s- %_ s _
* . g
., v tj ~-s .L * *
,% ~
- m ;s
, u
.- 4 ., s, s' r 3 _ s .-
, ' TABLE 2.1. s
SUMMARY
OF TROJAN SIREN SYSTEM EVALUATION RESULTS. j
; ~ ,s .s -
s w < g .s
\
__s u - <T 4 x -
,- m Chance of Alert ..
.-Population-Scenari urban Rural We'ighted Average
- No. Description y w__
(%) (%) . (%) 1 Warm Summer Weekend Day i-100 100 100
- (clear to partly. cloudy) 2 Summer Weekday Night 78 72 , 76 (clear to partly cloudy)
. 3 Winter Weekday Evening 97 '89 95 (cool, damp, and overcast) ~\ 4 Winter Night - 7 67 ~ 60 s. 65 (during rainstorm)
- Based on a total urbas populationof 46) o00 and a total rural population of 18,600,- s
- N' ,q . x _
3. 4
*f* 1
\
x k > g s A. N k we 2-2
,y w _ _ - **
where possible to account for different wind directions. For the Trojan analysis, one to four sirens were evaluated at each of the 50 listener sites. Sites at which only one siren was considered were located so close to the chosen' siren that the selection of additional sirens was obviously not warranted. The next step in the procedure is to establish the outdoor sound level produced by the selected sirens at each listener location. This is accomplished by applying adjustments to the e rat'd sound level of the siren as follows: L(listener) = L(siren) -Ad-As -Aair ~ Aatm' where L(listener) is the outdoor siren sound pressure level at the listener site (d B ) , L(siren) is rated sound pressure level of the siren at 100 ft (dB), A d is the distance attenuation (dB), A s is shielding attenuation (d B ) , Aair is the air absorption (dB ) , and A atm is the atmospheric attenuation caused by wind and temperature gradients (dB). The rated sound pressure levels for the Trojan sirens were obtained from the manufacturer's literature as follows: Federal Signal Corporation Rotating " Thunderbolt" sirens = 125 dBC @ 100 ft Stationary Model STA10 or STL10 sirens = 115 dBC @ 100 ft Stationary Model 5 sirens = 107 dBC @ 100 ft Stationary Model 2 sirens = 102 dBC @ 100 ft Stationary Model LCS-1 sirens = 86 dBC @ 100 ft The first two adjustments (for distance and shielding) are the same for all four test cases and are based on in forma tion obtained from USGS maps. Distance attenuation beyond 100 ft is 2-3
O calculated by assuming sound propagation from an acoustic point source with a reduction of 6 dB per distance doubled. It is calculated as follows: d Ad = 20 log 10I l00 I' where d 'is the siren-to-listener distance (ft). Shielding attenuation (As) is estimated based on the degree of break in the line-of-sight from siren-to-listener. Sirens are assumed to be at a height of 50 ft above terrain level and the listeners at a height of 5 ft. The break in line-of-sight is obtained by using ground contour information from USGS maps. For the Trojan analysis, a shielding attenuation of 15 dB was in-cluded if the break in line-of-sight was 50 ft or more. Other-wise, no attenuation was' assumed. The corrections for air absorption and a tmospheric effects depend on the meteorological conditions for the particular scena- ~ rio. The assumed conditions for the Trojan site are provided in Table 2. 2 .for the . four test cases, based on local weather infor-mation.* In terms of air absorption, these conditions indicate attenuations ranging between 0.6 and 0.9 dB per 1000 f t, depend-ing on the scenario [2,3]. For simplicity in the analysis, an average value of 0.8 dB per 1000 ft was assumed for all cases.
- Thus,
^ air"lb0 '
where d is the siren-to-listener distance (ft).
- Trojan Plant Environmental Impact Report - Amendment L, Fig.
2.3-4, Table 2.3-13 (March 1973). 2-4
TABLE 2.2. METEOROLOGICAL CONDITIONS POR THE FOUR SAMPLE SCENARIOS USED 'ID EVALUATE 'ITIE TROJAN SIREN SYSTEM. Relative Temperatur Scenario Tempera ture Humidity No. Wind Conditions
- Gradient (%) (*F) 1 10 mph frcm the north through- -2 C /100 m ; 50 75 out the region, except upslope Class A in the canyons 2 5 mph from the south in the +1 5 C/100 m; 90 55 river valley, dovnslope in the Class E canyons 3 3 mph from the south, calm in 11 C/100 m; 80 55
'the canyons Class E ~
14 15 mph from the south, 5 mph +1 C/100 m; 90 35 downslope in the canyons Class E
- Note: Weather data from the Trojan Plant indicate occasional conditions when the wind speed at lower elevations exceeds that at higher elevations. This unusual occurrence is assumed to be a measurement artifact, and has been ignored in this analysis.
2-5 {
The adjustment for a tmospheric gradient ef fects (Aa tm) is based on siren-to-listener azimuth with respect to wind direction cnd on wind and temperature gradient characteristics. Table 2.3 summarizes the calculation procedure for determining Aatm for occh scenario at the Trojan site. A more detailed description of the . estimation procedure for A atm can be found in Appendix D. Application of the above calculations yields the estimated outdoor sound pressure level for various sirens at each sample listener site, for each of the four scenarios. For the balance of the analysis, only the highest siren level at each listener site is generally used. An exception to this rule is made at listener sites where the sound level of a stationary siren is estimated to be between 0 and 6 dB lower than the sound level of a rotating-type siren, which had been determined to be the loudest siren. In such cases, the stationary siren was selected for further analysis. The reason for this exception is that the maximum sound level produced by a rotating siren is not continu-ous, and thus the total acoustic energy at the listener (as measured by the" single event noise exposure level, or SEL) is approximately 6 dB less than for a stationary (i.e., continuous) siren with tre same maximum sound level. 2.2 Estimating Indoor Sound Levels of Sirens a The result of the above calculations is a single outdoor siren sound pressure level at each of the 50 sample listener locations for each of the four test cases. Corresponding indoor levels are then obtained by subtracting typical values for resi-dontial building sound attenuation. For test cases 1 and 2 (cummer), residential windows were assumed to be partly open; for test cases 3 and 4 (winter) residential windows were assumed to l be closed. For the frequency region within the 500 Hz octave l - l 2-6 l
TABLE 2.3. ATMOSPHERIC ATTENUATION (Aata) CAUSED BY WIND AND TEMPERATURE GRADIENTS Sample Listener Site Siren-to-Scenario Postion with Respect Listener Aatm No. to Siren Position
- Distance (f t) (dB)
Upvind 0 - 1000 0 1000 - 2000 10
>2000 20 Crossvind 0 - 4000 0
>h000 10 Downwind (all) 0 Upvind 0 - 1000 0 2
1000 - 2000 10
>2000 20 Crossvind (all) 0 Downwind (all) 0 3 Same as Scenario No. 2, except A, =0 in canyons k Same as Scenario No. 2
# Defined with respect to the smaller angle ($) between the source-to-receiver directional vector and the vind directional vector as follows: Downwind for $ = 0 - h5 Crossvind for $ = 45 - 135 Upvind for $ = 135 - 180 .
2-7
l band, the sound attenuation into buildings is estimated to be 16 l dB for test cases 1 and 2 and 27 dB for test cases 3 and 4 [4] . l j 2.3 . Assumptions about Chance of Alert The outdoor and indoor siren levels calculated by the. above procedure provide some of the information r3 quired for the ana-lysis of the chance of alert. In addition, it is necessary to I know the level of interfering background noise at the listener locations. Figure 2-1 is a flow chart of the analysis computations. The analysis is divided into components-(rows) that correspond to L the possible activities of people for the various scenarios. The major components relate to peopl'e (1) at home (outside or in-l side), (2) at work,'or (3) in motor vehicles. The chance of alert is estimated for each activity component .and is.then multi- , p1'ied by the . fraction of people likely to be engaged in that activity (activity fraction). The results are summed to obtain l the overall chance of alert for each listener location and for each test case. Overall chances of alert for the variou's scenario (test _ case) conditions are then obtained by averaging the chances for all rural and/or urban' sample listener sites. l l Note that all estimates assume siren signal duration of 4 minutes; an average of the "3 to 5 minutes" called for in Appendix 3 of NUREG-0654. The effects of different siren signal durations are~ discussed in Appendix E. Siren detectability is a function of the siren signal level i and of the background noise level in a " critical frequency band" centered at the signal frequency. For this analysis, outdoor and indoor detectability'is estimated based on the signal-to-noise (S/N) difference in the 630-Hz 1/3-octave f~r equency band. The 2-8 l
1
- 5 st, 9
.e 1.-i 1:. 1:. h:J
- 1 b n
2
. . s M
, , , s ,
- 2. , , ,- -- --
ff fh :.., ! a 8 i n c$_ w
.k i
1 di 1 j .,!_ -_ -. __ l, I s , m) <
'll m
Il
!}
{
}}
t II ; .!!.{ .!](. g; 55 i o ss is e k ka i~
'O ,
, , u
'a W t
i :1a ei i h O T ,, g t --
- gs 11.1 _li .
Y.r i t II , e s m Z s e o H
.T P
O k h I: ;]!3 jaj ' a:l g
'as y .
g o o U 5 5 j 11.1I); $ I Ij a g rx
't y 7:.
r t
.. 4 j!'at jsa o E5 ' , ,
3 5S o jlz!
=2 eeeee l im 1 :n 3f D' VV Jt N
s . .
- : :- -1" e i & BE II H 3 r3 r rr A 2-9
I chosen criterion for alerting is that the given signal level must b2 9 dB or more above the minimum background noise level at any time during a 4-minute period for people who are not sleeping (i.e., a S/N difference of 9 dB). The chance of alert during sleep is based on the indoor siren Single Event Level ( S EL) -a measure of' total acoustic energy - and the sleep-awakening model developed by the U.S. Environmental Protection Agency [5] . The graph used for estimating the chance of alert during sleep is shown in Fig. 2-2; for the Trojan analysis, the curve for the chance of awakening one out of two sleepers was used. 2.4 Alerting People Out of Doors For the analysis of the ability of siren's to alert people out of doors, background noise levels are based on noise measure-mants conducted by BBN in the vicinity of the Trojan Plant in , March 1981. These _ measurements con'sisted of collecting 1-minute statistical summaries of background noise for a period of 1 hour
~
ct various . types of locations. The summaries provide the Lgo (sound level exceeded 90% of the time) for 1-minute samples of data in the 1/3-octave frequency band centered at 630 Hz.* The data were used to calculate the chance of detection for various siren levels and signal durations based on the background noise levels and their variability. Generalized types'of background noise environments were then established so that all sample listener sites would be included with one of these general cate-gories. In each category, the siren sound level necessary to clert is 9 dB greater than the minimum background noise level that could exist in any 4-minute period (1 minute for rotating sirens), adjusted for the probability distribution of such minima. This is handled by assigning a " median alerting level"
*The L 90 was used as a conservative estimate of the minimum sound level i hsua
100 l y 99.9 -l
-7 99.8 99.5 '
/ '
/
, / / i
/ / I 95
/
f f f
- 1 Out of 3 Sleepers, /
' /
g g / 1 Out of /
/
g / / [ t / / /^ 2' 3 / / / ' S* m , f' / / l,o
// /
f/ , l lr,/ f . 10 t SEL = L + 24 For Stationary Sirens
; SEL = L + 18 For Rotating Sirens.
5 For 4 Minuts Siren Operation The Middle Curve is Used For 2 This Analysis 1.0 0.5 0.2 0.1 0.06 0 40 50 80 70 80 90 100 110 120 130 140 INDOOR SIREN SE L ( db ) -
, < FIG. 2-2. CHANCE OF ALERT POR MAKENING PEOPLE ASLEEP.
2-11
4 for each background noise category and adjusting these levels in cccordance' with probability distributions' generalized from the data. The medi'an alerting levels for each background noise cate-gory are listed in Table 2.4. These are keyed to corresponding distributions shown in Fig. 2-3. For example, assume that a rotating siren produces 54 dB at a given location adjacent to a major traffic artery. Table 2. 4 indicates that the median alerting level-at such locations is 55 dB and that the applicable distribution on Fig. 2-3 is No. 5. The siren level minus the median alerting level is 54 - 55 = -1 dB. From distribution No. 5 on Fig. 2-3, we read 24% probability of alerting at -l'.dB. Note that probabilities of greater than 99% on Fig. 2-3 are treated as 100%, and those less than 1% are treated as 0%. Outdoor background noise in urban areas and along rural roadways is caused predominantly by motor vehicle traffic. It is generally insensitive to seasons of the year, but varies markedly with time of day. Minor traffic variations (i.e., less than a ' factor of 2 in traffic volume) have little effect on the background noise. In rural areas remote from roadways, outdoor background noise can be seasonal (birds, insects, etc.) and can. vary with the weather (wind, raln, waterflow, surf). Few people live or work in such " natural" acoustic' environments. During the analysis of the Trojan alerting system there were no instances where outdoor noise limited the effectiveness of the sirens. Note that results are given separately for stationary sirens and rotating sirens. This~is because rotating sirens would actually produce their estimated sound level during about one St-RR
TABLE 2.4. SIREN. ALERTING ABILITY FOR GENERALIZED CATEGORIES OF OUTDOOR ENVIRONMENTS. Median Alerting Level' (dB) Applicable Distribution 1 Rotating Sta tiona ry Rota ting Sta tiona ry Generalized Background Siren ' Siren Siren Siren Noise Environment (4 min) (4 min) (4 min) (4 min) I. URBAN A. Adjacent to Major Traffic Artery 55 53 No. 5 No. 3
- 3. Be: note frcan hjor Traffic Artery kB k6 No. 5 No. k II. RURAL A. Within View of hjor Noise Sources
- 1. Highway I-5: g3 gg 3,, g y,, g
- 2. Highway US-30' 48 k8 No. 6 No. b
- 3. Port of Icogview 53 52 No. 3 No. 2
- 3. Re: note frcza hjor Ncise Sources
- 1. No Wind or Vater Flow Noise' kl kl No. 3 No. 1
- 2. Subject to Wind Noise h5 kk No. 5 No. 3
- 3. Subject to Water Flow Noise' 57 57 No. 1 No. 1 III. INDUSTRIAL' 55 5k No. k No. 2 NOTES:
- 1. See Fig. 2-3.
- 2. Alerting levels apply for sites within 500 ft, with view angle (0) of 180* to highway; beyond 500 ft. levels should be reduced by 10 legge (D/500), where D = dist. fr a highway in fts for view angles less than 180*. levels should be further reduced by 10 logge (180/0).
- 3. Alerting levels apply for sites within 1600 ft, with view angle (8) of 180* to highway; beyond 1600 ft levels should be reduced by 10 legte (D/1600). where D = dist. frecs highway in fts for view angles less than 180 ,8 levels should be further reduced by 10 logge (180/0).
- k. Alerting levels apply at 300 ft fra strea:st for other distances adjust levels by 10 logie .
j ( 300/ distance) .
- 5. Alertinc levels apply at 1000 ft from sources for other distances adjust levels by 20 logie l (1000/ distance).
p l 2-13
f. 100 No.1 No. 2 No. 3 No. 4 No. 5 No. 6 99.9 99.8 99.5 99 : +1 - >+2 -- > +3 +4 +5. +6 98 95 !
. I 80 y 70 E
a l 50 7, E 40 ~ y'" 30 l E u 20 f 10 5 l
- 2. l I
1.0 0' li- - 2: -3:: -4 : 5 0.5 0.2 0.1 0.05 N O'
-1 0+1 -2 0+2 3 0 +3 4 0 +4 -5 0 +5 -6 0 4 l SIREN LEVEL MINUS MEDIAN ALERTING LEVEL FIG. 2-3. DISTRIBUTIONS FOR DETECTION BY PEOPLE OUT OF DOORS. .
91 - 1 6
~
quarter of the presumed 4-minute operating time at any particular listener location. Thus, the results for rotating sirens are based on 1-minute statistics rather than on 4-minute statistics. In summary, information regarding siren . type, estimated siren' sound level, background noise category at the listener site, and test-case conditions are used in conjunction with Table 2.4 and Fig. 2-3 to estimate the chance of siren detection outdoors. 2.5 Alerting People Indoors For the analysis of alerting people indoors at home, three types of activities are considered. These are (1) listening to radio'or TV, (2) sleeping, or (3) other activities that range from quiet to noisy situations. Table 2.5 provides the percen-tages assumed for various activities for each scenario. For people listening to radio or TV, the chance of alert'is 100%. For people sleeping, the chance of alert is calculated from the indoor siren SEL using the relationship shown in Fig. ! 2-2 for the chance of awakening one out of two sleepers. For all other indoor activities, the chance of alert is based on classi-fications of actual indoor background noise measurements under a wide variety of conditions. Results for test cases 1 and 3 are provided in Fig. 2-4 for 4-minute stationary sirens and in Fig. 2-5 for 4-minute rotating sirens. Thus, given the siren type, indoor siren level, and test case condition, these figures are used to estimate the chance of alert for indoor activities other than sleeping or listening to radio or TV. w 2-15 1
TABLE 2.5. ASSUMED ACTIVITIES AND BACKGROUND NOISE ENVIRONMENTS FOR PEOPLE INDOORS. Percentages of People Engaged in Various Activities Indoors (1) Indoor Noise Environment At flace Listening to Obviously Busy and Obviously Scenario of Business TV/ Radio Sleeping Noisy 8 Active' isolated' Quiet *
- 1. Waru Bisweer Veekend --
50 -- -- 15 10 25 Day (clear to partly eloudy)
- 2. Sisweer Weekov Night 5 -
95 - - - -- (clear to partly c1oudy)
- 3. Winter Weekosy During ,-- 20 -- 5 50 20 5 Evening Casauting Ilours (cool, damp, ove rc as t)
- 4. Winter Mirjat During .5 --
95 -- -- -- -- fluin s tnrw _ NOTC1;
- 1. Vacuum cleaning, diehuneher, shower, vent fan en , etc.
- 2. Dinner conversation, kitchen work, playing music, children at play, ate.
3 Molse-producing activity in adjacent ronse, oof t background muele, etc. l
%. Heading, study, eating nione. ]
me' 2-16 r- - --
- O e
100 N 99.9 99.8 99.5 I 99 9, /1
,5 //
. .1f
;- Sc.narw @
y- [ E 70 f Scenario @ E w / Y s J7 / 40 u 30 ) ' 1 n 1 / 10 , 5 2 ( 1.0 0.5 l l 0.2 l 0.1 0.06 0 h 10 20 30 40 50 80 70 80 90 100 INDOOR SIREN LEVEL ( dB ) FIG. 2-4. CHANCE OF ALERT FOR PEOPLE INDOORS AT BOME (4-MINUTE STATIONARY SIREN). I 2-17 l
100 i 99.9 99.8 99.5 I
,, J /
98 95 f.7 g sc ) [ Scenario @ E 70 y, # f s, / / 40 y .
/ g Scenario Q) f' u 30 20 / -
10 [ : 5 I .A, g ,, 2 1.0 1: 0.5 f 0.2 l 0.1 0.05
.of0 10 20 30 40 50 60 70 80 90 100 INDOOR SIREN LEVEL ( de l FIG. 2-5. CHANCE OF ALERT FOR PEOPLE INDOORS AT HOME (4-MINUTE
, ROTATING SIREN). l 2-18
l l For the analysis of alerting at work, two activity cate-gories are considered: (1) commercial / institutional, and (2) industrial environments. In the case of Trojan, only the night-time scenarios (test cases 2 and 4) include people at work. For these cases, essentially all nighttime work was assumed to occur in industrial environments and none of these people is likely to be alerted because of building attenuation and high background noise levels. Thus, the ~ chance of alert was assumed to be 0% at work at night for the Trojan analysis. In actuality, it is possible that these people would be alerted by some means (phone, radio) other than sirens, and hence our estimates may be low as a result. 2.6 Alerting People in Motor Vehicles The analysis for the alerting of motorists is based on the assumption of an average siren signal strength and spacing throughout the EPZ. The probability that a motorist will pass within the alert range of a siren during its 4-minute operation is estimated as follows: C = 2R+d x 100 (not to exceed 100%) where C is the chance of alert (%),.R is the maximum. alert distance (ft), d is distance traveled in 4 minutes (ft), and L is the average siren spacing (ft). Separate analyses were carried out for urban and rural areas of the Trojan EPZ. The average urban siren produces a sound level of 125 dB at 100 ft, and the average rural siren produces a sound level of 119 dB at 100 ft. Alerting ability was evaluated by using the re-sults of a study for the Society of Automotive Engineers (SAE) [6]. Siren alerting levels for speeds of 55 mph and 30 mph with windows shut or open were first determined from the SAE study results. The average siren source levels for rural and urban 2-19
~
areas'were then reduced to alerting levels in accordance with the propagation models from current NRC guidelines (i.e., 10 dB/ double distance) [7). In this manner, the maximum alert distance (R) was calculated for each driving condition. The distance traveled in 4 minutes (d) was calculated based on speed for each case, and the average siren spacing (L) was estimated to be 4,785 ft for urban areas and 6,895 ft for rural areas. The calculations of alerting ability for motorists are sum-marized in Table 2.6. The results indicate that the chance of alert is expected to be 100% for all conditions applicable to the Trojan analysis. TABLE 2.6. SIREN ALERTING FOR MOTORISTS. 4-min whicle whicle Rxy!. Signal Max. Alert Ravel Avg. Siren Chan
!beed Window for Alert Dist. , R dist. , d Spacing, L of Alert
~
Ama (nph) (bnditim (dB) (ft) (f t) (ft) (t) 55 closed 96 750 19,360 4785 100 open 90 1130 19,360 4785 100 URBAN l 30 closed 89 121'O 10,560 4785 100 open 86 1500 10,560 4785 100 55 closed 96 490 19,360 6895 100 l open 90 750 19,360 6895 100 RURAL 30 closed 89 800 10,560 6895 100 open 86 980 10,560 6895 100 2-20
- 3. EVALUATION OF THE PROMPT ALERTING SYSTEM FOR THE THREE MILE ISLAND NUCLEAR POWER STATION This section summarizes the evaluation of the siren alerting system for the Three Mile Island Nuclear Power Station (TMI).
The procedure that was used consists of a detailed analysis of siren alerting capability at each of 50 randomly chosen listener locations, under four dif ferent " sample scenario" conditions.
~
The random selection process for listener sites is described in Appendix F and the four test cases (sample scenarios) are included in Appendix G. The analysis is based on existing and proposed siren locations as of 30 June 1981. Maps that shown the siren locations are provided in Appendix H.
, The results of the evaluation for TMI are summarized in Table 3.1 and indicate that the chance of alert is estimated to vary between 49% and 90% depending on the sample scenario under consideration. The remainder of this report describes the procedure used to arrive at this conclusion. Input and output data for the analysis are included in Appendix I.
3.1 Estimating Siren Sound Levels Out of Doors at Listener Sites The first step in the procedure is to determine the siren in the vicinity of each selected listener site that is expected to produce the highest sound level at that site for each sample scenario. This choice is not always obvious, because the sound l 1evel caused by a particular siren at a given listener site depends not only on the sound output of the siren and its dis-tance from the listener, but also on shielding and a tmospheric effects (particularly wind direction). Therefore, it is generally necessary to evaluate several sirens in the vicinity of each listener site in order to determine the dominant one. As a general rule, the closest, highest-rated, nonshielded sirens are selected for evaluation at each site. Furthermore, sirens are 3-1
\
TABLE 3.1.
SUMMARY
OF T?ll SIREN SYSTEM EVALUATION RESULTS. Chance of Alert Scenario Population-Urban Rural Weighted Average
- No. Description -(%) (%) (%)
1 Warm Sumner Weekday Af ter-noon (clear to partly cloudy) 96 88 -90 2 Surmer Weekday Night (clear to partly cloudy) 82 66 70 - 3 Winter Weekday Evening (cold and overcast) 89 76 80 4 Winter Night (during snowfall) .66 42 49
- Based on a total urban population of 46,573 and a total rural population of 119,722.
9 3-2
1 a chosen such that they are distributed north, south, east, and west of the site (or in any other four mutually perpendicular directions) where possible to account for different wind direc-tions. For the TMI analysis, four. or six sirens were evaluated at 46 of the 50 listener sites. Only two or three sirens were considered at the remaining four sites. These sites were either located at the fringes of the EPZ such that sirens could not be chosen in all directions, or they were located so close to one or two sirens that the selection of additional sirens was obviously not warranted. The next, step . in the procedure is to establish the outdoor sound level produced by the selected sirens at each listener location. This is accomplished by applying adjustments to the rated sound level of the siren as follows: - L(listener) = L(siren) -Ad-As -Aair - Aa tm ' where L(listener) is the outdoor siren sound pressure level at the listener site (dB ) , L(siren) is rated sound pressure level of the siren at 100 ft (dB), Ad is the distance attenuation (dB), A s is shielding attenuation (d B ) , Aair is the ai~r absorption (dB), and A atm is the atmospheric attenuation caused by wind and temperature gradients (dB ) . The rated sound pressure levels for the TMI sirens were estimated based on anechoic chamber performance data, obtained with the cooperation of the Metropolitan Edison Company. These data indicate sound pressure levels of 142.9 dBC and 145 dBC for stationary and rotating ' sirens respectively, measured at a dis-tance of 2 meters. These levels were reduced by 23.7 dB to extrapolate to the level at a distance of 100 fee t (see distance adjustment discussion below) and then increased by 3 dB to l account for the presence of a ground plane for sirens in the 3-3 l l
i field. The resulting rated sound pressure levels at 100 ft are therefore 122 dB for TMI stationary sirens and 124 dB for TMI rotating sirens. The first two adjustments (for distance and shielding) are the same for all four test cases and are based on information obtained from USGS maps. Distance attenuation beyond 100 ft is calculated by assuming sound propagation from an acoustic point source with a reduction of 6 dB pe.r. distance doubled. It is calculated as follows: d Ad = 20 log 10 I l00) ' where d is the siren-to-listener distance (ft). Shielding attenuation , As) is estimated using the following formula for the attenuation of a rigid straight barrier for sound incident from a point source [21: ~
'A ""
s
= 20 log + 5 dB for N > -0.2 tanh /2nN
= 0 dB for N < -0.2 l
N is the Fresnel number (dimensionless): N=if(A+B-d) Where A '= wavelength of sound, f t (1.79 ft for 630-Hz siren tone) i d = straight-line distance between source and rece ive r, ft 1 A+B = shortest path length of wave travel over the
- barrier between source and receiver,- ft
+ sign = receiver in the shadow zone (i.e., barrier obstructs line-of-sight) l l 3-4 l
- sign = receiver in the bright zone (i.e., barrie,r doesn't obstruct line-of-sight)
When N is negative, the above equation for As is evaluated by replacing N with lNl, r.nd by replacing tanh with tan. Shielding attenuation is limited to a maximum of . 24 dB . based upon a large body of experimental data. For the TMI analysis, sirens are assumed to be at a height of 52 f t above terrain level, listener sites are assumed to be at a height of 5 f t above terrain level, and barrier heights are' obtained from ground contour information on USGS maps. The adjustments for air absorption and atmospheric effects depend on the meteorological conditions for the particular scenario. The assumed conditions for the TMI site are provided in Table 3.2 for the four test cases, based on local weather information.* In terms of air absorption, these conditions indicate the following attenuation rates based upon temperature and relative humidity [31: Scenario Aair (dB per 1000 ft) 1 0.88 2 0.79 3 0.55 4 0.64 1 The adjustment for atmospheric gradient ef fects (Aa tm) is
' based on siren-to-listener azimuth with respect to wind direction
*Three Mile Island Nuclear Station Unit 2-Environmental Impact Report, Chapter 2.
3-5 1
i TABLE 3.2. METEOROLOGICAL CONDITIONS FOR 'IEE FOUR SAMPLE SCENARIOS USED 'IU EVALUATE 'IEE 'INI SIREN SYSTEM. Relative Scenario Temperature Humidity Temperature No. Wind Conditions * ' Gradient (%) (OF) 1 5 mph from the east -1.00F/100 ft 65 85 Class A 2 5 mph from the northwest +0.5 F/100 ft 80 65 Class E 3 3 mph from the southeast -0.5 F/100 ft .70 40 Class D ' 4 15 mph from the west -0.5 F/100 ft 90 25 - Class D
*At 100 ft above ground level.
I l l I 3-6
and on wind and temperature gradient characteristics. Table 3.3 summarizes the calculation procedure for determining A atm IOC each scenario at the TMI site. A more detailed description of the estimation procedure for Aa tm can be found in Appendix D. Application of the above calculations yields the estimated outdoor sound pressure level for various sirens at each sample listener site, for each of the four scenarios. For the balance of the analysis, only the highest siren level at each listener site is generally used. An exception to this rule is made at listener sites where the sound level of a stationary siren is estimated to be between 0 and 6 dB lower than the sound level of a rotating-type siren, which had been determined to be the loudest siren. In such cases, the stationary siren was selected for further analysis. The reason for this exception is that the maximum sound level produced by a rotating siren is not continu-ous, and thus the total acoustic energy at the listener (as measured by the single event noise exposure level, or SEL) is approximately 6 dB less than for a stationary (i.e. , continuous) siren with the same maximum sound level. 3.2 Estimating Indoor Sound Levels of Sirens The result of the above calculations is a single outdoor siren sound pressure level at each of the 50 sample listener locations for each of the four test cases. Corresponding indoor levels are then obtained. by subtracting typical values for resi-dential building sound attenuation. For test cases 1 and 2 (summer), residential windows were assumed to be partly open; for test cases 3 and 4 (winter) residential windows and storm windows were assumed to be closed. For the frequency region within the 500 Hz octave band, the sound attenuation into buildings is esti-mated to be 16 dB for test cases 1 and 2 and 31 dB for test cases 3 and 4 [4]. For commercial buildings, the outdoor-to-indoor 3-7
w TABLE 3.3. CALCULATION OF ATMOSPHERIC ATTENUATION, A atm' CAUSED BY WIND AND TEMPERATURE GRADIENTS (SEE APPENDIX D FOR DETAILS). Siren-to-Listener Distance, D (Ft) Relative to Xn ( Ft)_ Aatm (dB) D 1 1.2 X o 0 1.2'Xo < D 1 1. 7 Xo 5
.l.7 X o < D 1 2.4 X o 10 2.4 Xo < D 1 3'.4 Xo 15 3.4 X o <D 20 Computation of Xn X=o .f 1057/[eScos$-a~ -
Scenario 1 2 3 4 Wind Direction, O w 90 315 135 270 AT F (150'-50') -1 +0.5 -0.5 -0.5 a = ce =.AT/(In 150'-in 50') -0.91 +0.46 -0.46 -0.46 Wind Speed, V2 ft/sec @ 100ft 7.3 7.3 4.4 22 e6 = v2 /(In 100' - in 2') 1.87 1.87 1.12 5.6
- W 3 R/S = 5'/50' O.1 f(R/S) 0.45 r ,
X o (min) @ & = 0 634' 890' 840' 429' &c = Cos-1 ( ) 119 76 114 95 3-8
s noise reduction is estimated to be 31 dB, assuming closed and sealed windows for all four scenarios. 3.3 Assumptions about Chance of Alert The outdoor and indoor siren levels ' calculated by the above procedure provide some of the information ~ requ' ired for the ana-lysis of the chance of alert. In = addition, it is necessary to know the level of interfering background noise at the listener locations. Figure 3-1 is a flow chart of the analysis computations. The analysis is divided into components (rows) that correspond to i the possible activities of people for the various scenarios. The major components relate to people (1) at home (outside or inside), (2) at work, or (3) in motor vehicles. The chance of alert is estimated for each activity component and is then multiplied by
~
the fraction of people likely to be engaged in that activity (activity fraction). The results are summed to obtain the overall chance of alert for each listener location and for each test case. Overall chances of alert for the-various scenario (test case) con-ditions are then obtained by averaging the chances for all rural and/or urban sample listener sites. Note that all estimates ! assume siren signal duration of 4 minutes: an average of the "3 to 5 minutes" called for in Appendix 3 of NUREG-0654. The effects of different siren signal durations are discussed in Appendix E. Siren detectability is a function of the siren signal level and of the background noise level in a " critical frequency band" centered at the signal frequency. For this analysis, outdoor and indoor detectability is estimated based on the signal-to-noise l
~
3-9
y s. x c , , s w g. I Z 1 u gy c;ig 9 rpm?:'o+, J p sty $ 4$ IJ
'Gb,_"e.
G j II , ~ ; uIf s {' ~' x,. [{ I j o.i cp { l] 3 , 4 t>
, \
ji}\ jij g! Ta 5 T~ ' II
~
.s g w a,f Tal. d, ge, 4 e, .
E a; 4
!)
k$
~'
\
fl f6^ l h, f o s
>i k\
v o, e g y 1h ja 4 cla]- 5 g, 1a[ss s 4 e ij 1
!E gg 1
)< '! fu
[lj g, r ,, .j 8 ,fd
, m 1
1 IA r
$I 8
r N o
=
r b 2
-a
- - ? $u 1 i
6s60s 111 i 11 b
~
~ v1 1 u
V . a e y
~
y 3 O ti r3' et
- 0
. N 3-10
~ .--
(S/N) difference in the 630-Hz 1/3-octave frequency band. The chosen criterion for alerting is that the given signal level must be 9 dB or more above the minimum background noise level at any time during a 4-minute period for people who are not sleeping (i.e., a S/N difference of 9 dB). The chance of alert while sleeping is based on the indoor siren Single Event Level (SEL) -a measure of total acoustic energy - and the sleep awakening model developed by the U.S. Environmental Protection Agency [5) . The graph used for' estimating the chance of alert during sleep is shown in Fig. 3-2; for the Three Mile Island analysis, the curve for the chance of awakening one out of two sleepers was used. 3.4 Alerting People Out of Doors For the analysis of the ability of sirens to alert people out of doors, background noise levels are based on noise measurements conducted by BBN in the vicinity of the Trojan Nuclear Plant in
. Oregon, near the Susquehanna Steam Electric Station in Pennsyl-vania, and upon the body of data in BBN files. The data typically consisted of statistical summaries of background noise at various
. types of locations. The summaries provide the L90 (sound level exceeded 90% of the time) for 1-minute samples of data in the 1/3-octave frequency band centered at 630 Hz.*. The data were used to calculate the chance of detection for various siren levels and signal durations based on the background noise levels and their variability. Generalized types of background noise environments were then established so that all sample listener sites would be included with one of these general categories. In each category, the siren sound level necessary to alert is 9 dB greater than the minimum background noise level that could exist in any 4-minute period (1 minute for rotating sirens), adjusted for the probabil-
- *The L90 was used as a conservative estimate of the minimum sound level 3-11
100 [ I 99.9 j f l "5 99 / i
/ ;
" /
/
1 Out of 3 Sleepers, / a y /
/ 1 Out of /
/
l }
/
[ j' [, ~ N 1 Sleeper w [/ l l *0 I 3
// f r
/ '
c y
,/
l ~ 10 b / SEL = L + 24 For Stationary Sirens
; SEL = L + 18 For Rotating Sirens.
6 For 4 - Minute Siren Operation The Middle Curn is Used For 2 This Analysis 1.0 0.5 0.2 0.1 0.06 O 4 s0 .0 ,0 .0 .0 ,. ,,0 ,, ,, ,.0 INDOOR SIREN SEL ( db ) - FIG. 3-2. CHANCE OF ALERT POR / WAKENING PEOPLE ASLEEP.
~
3-12
ity distribution of such minima. This is handled by assigning a
" median alerting level" for each background noise category and adjusting these levels in accordance with probability distribu-tions generalized from the data.
The median alerting levels for each background noise category are listed in Table 3.4. These are keyed to corresponding distri-butions shown in Fig. 3-3. For example, assume tha t a rotating siren produces 53 dB at a given urban location during the daytime. adjacent to a major traffic artery. Table 3. 4 indicates that the median alerting level at such locations is 54 dB and that the ap-plicable distribution on Fig. 3-3 is No. 5. 'The siren level minus the median alerting level is 53 - 54 = -: dB. From distribution No. 5 on Fig. 3-3, we read 24% probabil'ity of alerting at -1 dB.
, Note that probabilities of greater than 99% on Fig. 3-3 are treated as 100%, and those less than 1% are treated as 0%.
Outdoor background noise in urban areas and along rural roadways is caused predominantly by motor vehicle traffic. It is generally insensitive to seasons of the year, but varies markedly with time of day. Minor traffic variations (i.e., less than a factor of 2 in traffic volume) have little effect on the background noise. In rural areas remote from roadways, outdoor background noise i can be seasonal (birds, insects, etc.) and can vary with the weather (wind, rain, waterflow, surf). Few people live or work in such " natural" acoustic environments. As shown in Table 3.4, rural, non-coadway background noise is selected to be dependent on windspeed. Note that results are given separately for stationary sirens and rotating sirens. This is because rotating sirens would actu-ally produce their estimated sound level during about one quarter of the presumed 4-minute operating time at any particular listener 3-13 I
Ia TABLE 3.4. SIREN ALERTING ABILITY FOR GENERALIZED CATEGORIES OP - OUTDOOR ENVIRONMENTS. Median Alerting Level (dB) Applicable Distribution Rotating Stationary Rotating Stationary Generalized Background Siren 51ren Siren Siren Noise Environment ' (4 min) (4 min) (4 min) (4 min)
- 1. URBAN A. Roadway
- 1. Daytime 54 52 No. 5 No. 3
- 2. Evening 49 48 No. 4 No. 3 e 3. Nighttime 43 43 No. 3 No. 2
' 5. Non-Roadway
- 1. Daytime 50 48 No. 5 No. 4
- 2. Evening 48 47 No. 4 No. 3
- 3. Ni;,ht time 42 41 No. 3 No. 2
- 11. RURAL A. Roadway
. 1. Limited Access Highway 63 61 No. 6 No. 4
- 2. Other Highway 3 51 50 No. 6 No. 4 B. Non-Roadway
- 1. No-Vind Noise 5 28 27 No. 3 No. 1 -
- 2. Subject to Wind Noise (See Note) (See Note) No. 5 No. 3 111. l'NDUSTRIAL 55 54 No. 4 No. 2 NOTES:
- 1. See Fig. 3-3.
- 2. Alerting levels apply f or sites within 500 f t, with view angle (0) of 180 to highway; beyond 500 ft, levels should be reduced by 10 log 10 (D/500), where D=dist. from highway in ft; for view angles less than 1808, levels should be further reduced by 10 logio (180/0).
- 3. Alerting levels apply for sites within 1600 f t, with view angle (8) of 180* to highway; beyond 1600
( ft, levels should be reduced by 10 log 10 (D/1600), where D-dist. from highway in it; for view angles 8 less than 180 , levels should be further reduced by 10 los10 ( 80M). 6 Wind speed e 1 emh.
- 5. Median' Alcrting Lcyc1 (with wind)
- Median Alcrting Level (so wind) + 15 log 10
- where S = average wind speed in sph.
6. Alerting levels apply at 1000 ft f rom source; for other distances adjust levels by 20 log 10 (1000/ distance). I l l l l { l l 3-14
100 No.1 No.2 No. 3 No.4 No.5 No. 6 99.9 99.8 99.5 99 +1 - ' + 2 --- '+3 +4 .+5 +6 98 95 9e 80 S f 70 E
- a 50 j
f 8 40
~
y 30 I, ,\ b 20' f 10 5 = l i 2 l . 1.0 0' 1'>- -2: -3:: -4: 5 l l 0.5 0.2 0.1 0.05 0' 1 0+1 -2 0+2 -3 o +3 4 0 +4 5 0 +5 6 0 +6 SIREN LEVEL MINUS MEDIAN ALERTING LEVEL i l FIG. 3-3. DISTRIBUTIONS FOR DETECTION BY PEOPLE OUT OF DOORS. I 3-15 l P
location. Thus, the results for rotating sirens are based on 1-minute statistics rather than on 4-minute statistics. In summary., information regarding siren type, estimated siren sound level, background noise category at the listener site, and test-case conditions are used in conjunction with Table 3.4 and
~
Fig. 3-3 to estimate the chance of siren detection outdoors. 3.5 Alerting People Indoors For the analysis of alerting' people indoors at home, three types of activities are considered. These are ( 1) listening to radio or TV, (2) sleeping, or (3) other activities that range from quiet to noisy situations. Table 3.5 provides the percentages assumed for various activities for each scenario. For people listening to radio or TV, the chance of alert is 100%. For' people s'leeping , the chance of alert is calculated from the indoor siren SEL using the relationship shown in Fig. 3-2 for - the chance of awakening one out of two sleepers. For all other indoor activities, the chance of alert is based on classifications of actual indoor background noise measurements under a wide variety of conditions. Results for test cases 1 and 3 are provided in Fig. 3-4 for 4-minute stationary sirens and in Fig. 3-5 for 4-minute rotating sirens. Thus, given the siren type, indoor siren level, and test case condition, these figures are used to estimate the chance of alerting for indoor activities other than sleeping or listening to radio or TV. For the analysis of alerting at work, two activity categories are considered: (1) commercial / institutional, and (2) industrial environments. For the TMI analysis, it was assumed that 75% of the working population are in commercial establishments while the 3-16
TABLE 3.5. ASSUMED ACTIVITIES AND BACKGROUND NOISE ENVIRONMENTS FOR PEOPLE INDOORS. Percentages of People Engaged in Various Activities Indoors (1)
, Indoor Noise Environment At Place Listening to Obviously Busy and 3bviously Scenario of Business .TV/ Radio Sleeping Noisy' Active' Isolated' Quiet *
- 1. Warm Summer Weekday 41 27 5 -
4 5 to Afternoon (clear to # partly cloudy)
- 2. Sumer Weekday Night 4 - 96 - - - --
(clear to partly cloudy) *
- 3. Winter Weekday During --
20 - 5 50 - 20 5 Evening Commuting Hours (cold and ' overcast) ,
- 4. Winter Right During 5 - 95 -- - -- --
Snowfall , 1 NOTES:
- 1. Vacuum cleaning, diahwasher, shower, vent f an on, etc.
- 2. Dinner conversation, kitchen work, playing muste, children et play, etc.
- 3. Noise-producing activity in adjacent room, sof t background muelc ' etc.
- 4. Reading, study, eating alone.
3-17
/
\
remaining 25% are in industrial locations. For commercial loca- - tions, the chance of alert is based on the statistics of background noise measured in a typical office environment, using Fig. 3-6. For industrial locations, it has been assumed that 100% of the people are likely to be alerted by some means of communi-cation other than sirens. 3.6 Alerting People in Motor Vehicles The analysis for the alerting of motorists is based on the assumption of an average siren signal strength and spacing throughout the EPZ. The probability that a motorist will . pass within the alert range of a siren during its 4-minute operation is estimated as follows: C = 2R d x 100 (not to exceed 100%) where C is the chance of alert (%), R is the maximum alert distance (ft), d is distance traveled in 4 minutes (ft), and L is the average siren spacing (f t) . Separate analyses were carried out for urban and rural areas of the TMI EPZ. The average urban siren produces.a sound level of 125 dB at 100 ft, and the average rural siren produces a sound level of 123 dB at 100 ft. Alerting ability was evaluated by using the results of a study for the Society of Automotive Engineers (SAE) [6]. Siren' alerting levels for speeds of 55 mph and 30 mph with windows shut or open were first determined from the SAE study results. The average siren source levels for rural and urban areas were 3-18
, . ~ _ _ . _ _ _ _ _ _-
l 100
" O.
99.9 99.8 99.5 j 99
= /1
* //
.1(
2 sc _ in -o l / v .0
.y / s__,. e
$w I Y y W /
$. 20 -/ /
- 1 /
10 / /. -* 5 i l 2 1.0 D.5 0.2 0.1 0.06 l 0 T 30 g 50 80 70 80 90 goo INOOOR SCREN LEVEL ( dB l
. -4. C C ALERT R PEOPLE INDOORS AT HOME ( 4-MINUTE-g 3-19
M 100 l 99.9 99.8 99.5
" 1 /
c
.- 98 LF
* / /
/ .i )/ '
Scenario @ a: 70
= 1 1 5 ," / /
o / '/ ga u 30
'y j S_ w 10
/ /
- / : .p 1.0 1
0.5 i 0.2 0.1 0.05 l 0 0 10 20 30 40 50 60 70 80 90 100 INDOOR SIREN LEVEL ( db ) PIG. 3-5. CHANCE OF ALERT POR PEOPLE INDOORS AT HOME (4-MINUTE ROTATING SIREN). 3-20
100 99.9 99.8 4 MINUT E STATIONARY SIRENg p-4 MINUTE ROTATING SIREN 99.5 99 - 98 o y . i d f
=
- 80 e
E 70 w
< 60 o
g
- A '
f/ 5 30 J 20 10 5 2 1.0 0.5 l 7 0.2 ! 0.1 l 0.06 t ,- 0 0 10' 20 30 40 50 60 70 80 90 100 INDOOR SIREN LEVEL (dB) FIG. 3-6.- CHANCE OF ALERT FOR PEOPLE INDOORS AT WORK IN COMMERCIAL / INSTITUTIONAL ESTABLISHMENTS. 3-21 i 1 .. .- . - .
then reduced to alerting levels in accordance with the propagation ~ models from current NRC guidelines (i.e., 10 dB/ double di'tance) s [7]. In this manner, the maximum alert distance (R) was calcu-lated for each driving condition. The distance traveled in 4 minutes (d) was calculated based on speed for each case, and the average s.iren spacing (L) was estimated to be 5,560 ft for urban areas and 11,850 ft for rural areas. The calculations of alerting ability for motorists are sum-marized in Table 3.6. The results indicate that the chance of alert is expected to be 100% for all conditions applicable to the TMI analysis. TABLE 3.6. SIREN ALEITDC POR PDIORISTS. 4-min Vehicle Vehicle Ibg3. Signal Max. Alert 'Itavel Avg. Siren Chance Speed Window for Alert Dist., R dist., d Spacing, L of Alert Area (nph) (bndition (dB) (ft) (ft) (ft) (%) _. 55 closed 96 610 19,360 5560 100 open 90 920 19,360 5560 100 URBAN 30 closed 89 980 10,560 5560 100 open 86 1210 10,560 5560 100 55 closed 96 650 19,360 11,850 100 open 90 980 19,360 11,850 100 RURAL 30 closed 89 1060 10,560 11,850 100 open 86 1300 10,560 11,850. 100 l l t l l 3-22 l -
- 4. EVALUATION OF THE PROMPT ALERTING S'YSTEM FOR THE INDIAN POINT NUCLEAR POWER STATION This section summarizes the evaluation of the siren alerting system for the Indian Point Nuclear Pow'er Station. The procedure that was used consists of a detailed analysis of siren alerting capability at each of 50 randomly chosen . listener locations, under four different " sample scenario" conditions. The random selection process for listener sites is described in Appendix J and the four test cases (sample scenarios) are included in Appendix K. The analysis is based on existing and proposed siren locations as of 25 August 1981. Maps which show the siren loca-tions are provided in Appendix L.
The results of the evaluation for Indian Point are summa-rized in Table 4.1 and indicate that the chance of alert is estimated to vary between 57% and 95% depending on the sample scenario under consideration. The remainder of this report describes the procedure used to arrive at this conclusion. Input and output data for the analysis are included in Appendix M. 4.1 Estimating Siren Sound Levels Out of Doors at Listener Sites The first step in the procedure is to determine the siren in the vicinity of each selected listener site that is expected to produce the highest sound level at that site for each sample scenario. This choice is not always obvious, because the sound level caused by a particular siren at a given listener site de-pends not only on the sound output of the siren and its distance from the listener, but also on shielding and atmospheric ef fects (particularly wind direction). Therefore, it is generally neces-sary to evaluate several sirens in the~ vicinity of each listener site in order to determine the dominant one. As a general rule, the closest, highest-rated, nonshielded sirens are selected for evaluation at each site. Furthe rmo re , sirens should be chosen 4-1
i TABLE 4.1.
SUMMARY
OF INDIAN POINT SIREN SYSTEM EVALUATION RESULTS. Chance of Alert I Scenario Population-Urban Rural Weighted Average
- No. Description (%) (%) (%) ,
1 Wam Sumer Weekday Af ter-noon (clear to partly cloudy) 98 93 95 2 Sumer Weekday Night (clear to partly cloudy) 80 70 74 - I 3 Winter Weekday Evening
'(cold and overcast) 91 78 83 ,
4 Winter Night . (during snowf all) 63 -53 57 Abased on a total urban population of110,928 and a total rural population of 146,454. a. i l 4-2
i such that they.are distributed north, south, east, and west of the site (or in any other four mutually perpendicular directions) where possible to account for dif ferent wind directions. For the Indian Point analysis, four or five sirens were evaluated at each of the 50 listener sites. The next step in the procedure is to establish the outdoor sound level produced by the selected sirens at each listener location. This is accomplished by applying adjustments to the rated sound level of'the siren as follows: L(listener) = L(siren) -Ad-As ~Aair ~ Aatm' where L(listener) is the outdoor siren sound pressure level at the listener site (dB ) , L(siren) is the rated sound pressure level of the siren at 100 f t (dB), A d is the distance attenuation (d B ) , A s is the shielding attenuation (dB), Aair is the air absorption (dB), and A atm is the atmospheric attenuation caused by wind and temperature gradients (dB) . The rated sound pressure level for all of the Indian Point sirens was taken to be 125 dB at a distance of 100 ft, based upon the siren manufacturer's rating; all sirens are rotating type units. The first two adjus tmen ts (for distance and shielding) are the same for all four test cases and are based on information I obtained from USGS maps. Distance attenuation beyond 100 ft is calculated by assuming sound propagation from an acoustic point source with a reduction of 6 dB per distance doubled. It is calculated as follows: Ad = 20 log 10 (100)
- where d is the siren-to-listener distance (ft).
4-3 1
5 Shielding attenuation (As) is estimated using the following formula for the attenuation of a rigid straight barrier for sound incident from a point source [ 2] : A f2iB s= 20 log + 5 dB for N > -0.2
< tanh /2iN
= l0 dB for N < -0.2 1
4 k N is the Fresnel number (dimensionless): N=f(A+B-d). Where A = wavelength of sound, ft (1.79 ft for 630-Hz siren tone) d = straight-line distance between source and receiver, ft A+B = shortest path length of wave travel over the barrier between source and receiver, ft
+ sign = receiver in the shadow zone (i.e., barrier obstructs line-of-sight)
- sign = receiver in the bright zone (i.e., barrier doesn't
~ '
obstruct line-of-sight) When N is negative, the above. equation for A s is evaluated by replacing N with lNl, and by replacing tanh with tan. Shielding attenuation is limited to a maximum of 24 dB based upon a large body of experimental data. For the Indian Point analysis, sirens are assumed to be at a height of 52 ft above terrain level, listener sites are assumed to be at a height of 5 ft above terrain level, and barrier heights are obtained from ground contour information on USGS maps. The adjustments for air absorption and atmospheric ef fects depend on the meteorological conditions for the particular scena-rio. The assumed conditions for the Indian Point site are 4-4
2 provided in Table 4.2 for the four test cases, based on local weather information.
- In ' terms.of air ab, sorption, these con-ditions indicate the following attenuation rates based upon temperature and relative ~ humidity [3] .
Scenario Aair (dB per 1000 ft) 1 0.85 2 0.81 3 0.49 4 0.46 The adjustment for atmospheric gradient effects ( Aatm) is based on siren-to-listener azimuth with respect to wind direction and on wind and temperature gradient characteristics. Table 4.3 summarizes the ' calculation procedure for determining A a tm f0# each scenario at the Indian Point. site. A more detailed descrip-tion of the estimation procedure for A a tm can be found in Appen-dix D. Application of the above calculations yields- the estimated outdoor sound pressure level for various sirens at each sample ' listener site, for each of the four scenarios. For the balance of the analysis, only the highest siren level at each listener site is used. 4.2 Estimating Indoor Sound Levels of Sirens The result of the above calculations is a single outdoor siren sound pressure level at each of the 50 sample listener locations for each of the four tes t cases. Corresponding indoor i
- Final Facility Description and Safety Analysis Report for Indian Point No. 3 Nuclear Power Plant, Section 2.6 (1973).
4-5
t TABLE 4.2. METEOROLOGICAL CONDITIONS FOR 'IEE FOUR SAMPLE
~
SCENARIOS USED 'IU EVALUATE 'IEE INDIAN POINT SIREN SYSTEM. Relative Scenario Temperature Humidity Temperature No. Wind Conditions
- Gradient (%) (OF) 1 10 mph from the SSE from the south in the riv;er- -1.00F/100 ft 65 80 Class A valley 2 6 mph from the NNE; from +0.50F/100 ft 80 70 the north in the river Class E valley 3 10 mph from the northwest -0.5 F/100 ft 70 30 Class D ,
4 15 mph from the southeast -0.5 0 F/100 ft 90 30 _ Class D
*At 100 ft above ground level.
w 4-6
TABLE 4.3. CALCULATION OF A'IMOSPHERIC ATTENUATION, Ag, CAUSED BY WIND AND TEMPERATURE GRADIENTS (SEE APPENDIX D FOR DETAILS) .. Siren-to-Listener Distance, D (Ft) Relative to Xn (Ft) Aatm (dB) D 1 1.2 X o o 1.2'X, < D 1 1.7 Xo 5 1.7 X, < D 1 2.4 X, 10 . 2.4 X, < D 1.3.4 X, 15 3.4 X, < D 20
' Ccmputation of Xo 47S -----
I o= y . f g/R 1057/[eScos$-a Scenario 1 2 3 4 Wind Direction, O o General 157.5 22.5 315 135 i',' valley , 0 180 - -
- AT F (95'-7') -0.9 +0.44 -0.44 -0.44 a - as = AT/in 95'-in 7') -0.35 +0.17 -0.17 -0.17
- Wind Speed, V ft/sec@l00ft 14.7 2
8.8 14.7 22 s8 = V2 /(in 100' - In 2') y 3.75 2.25 3.75 5.62 J V R/S 0.1 f(R/S)
- Xo (min) @ $ = 0 #522' .633' 534' 439
& = Cos" (g) ,
950 850 93 92 0 4-7
( levels are then obtained by subtracting typical values for com-marcial or residential building sound attenuation. For test cases 1 and 2 (summer), residential windows were assumed to be partly open; for test cases 3 and 4 (winter) residential windows were assumed to be closed (with storm windows). For the fre-quency region within the 500 Hz octave band, the sound attenua-tion into buildings is estimated to be 16 dB for test cases 1 and 2 and 31 dB for test cases 3 and 4 [4]. For' commercial build-ings, the outdoor-to-indoor noise reduction is estimated to be 31 dB, assuming closed and sealed windows for all four scenarios. 4.3 Assumptions about Chance of Alert The outdoor and indoor siren levels calculated by the above procedure provide some of the information required for the ana-lysis of the chance of alert. In addition, it is necessary to know the level of interfering background noise at the~ listener locations. ~ Figure 4-1 is a flow chart of the ' analysis computations. The analysis is divided into components (rows) that correspond to the possible activities of people for the various scenarios. The major components relate to people (1) at home (outside or in-side), (2) at work, or (3) in motor vehicles. The chance of alert is estimated for each activity component 'and -is then multi'- plied by the fraction of people likely to be engaged in that activity (activity fraction). The results are summed to obtain the overall chance of alert for each listener location and for ecch test case. Overall chances of alert for the various scena-rio (test case) conditions are then obtained by averaging the chances for all rural and/or urban sample listener sites. Note that all estimates assume siren signal duration of 4 minutes: an 4-8
. : .n ..
,,nco a =. .3. c:.a , '-
. c.. ; .c .
,,* _1 4,4 s ,
v c . . y_. f 1 n.v. t T 11 & [
.. 1 1 1 4J m
i,
,)
,E Ij g S
m, y.: l
,1 s
I. 1 __ j -. _, -- l il I ji. . 11 11 !!
. I, ' ' '
t t TT ;
.!jj .tij
-_ __ _ _ _ _ __ __ J, J, Il b5 ta i *
) ,
*ti -
tt I si
*A
'):
.k t
1.i :1. .r 1 T 7- T t -- fllr ' l1. }} .II t TT . I. .. l l
. r-l A%- \; t$
l 1! jij: 1 _l 'jal- @ la!i g s- g l$ 1 A .. A [* D
] ]
1 1 1 1 1 o Ill !!! % O g .4 , ,, o g u A.. geese ()! (!)111 : 8 EM D . V V S
~
it
.iis .
g i 52 r r3 M rf f O H 4-9
cverage of the "3 to 5 minutes" called for in Appendix 3 of NUREG-0654. The effects of different siren signal durations are discussed in Appendix E. Siren detectability is a function of the siren signal level and of the background noise level in a " critical frequency band" centered at the signal frequency. For this analysis, outdoor and indoor detectability is estimated based on the signal-to-noise (S/N) difference in the 630-Hz 1/3-octave frequency band. The chosen criterion for alerting is that the given signal level must be 9 dB or more above the minimum background noise level at any time during a 4-minute period for people who are not sleeping (i.e., a S/N difference of 9 dB). The chance of alert while sleeping is based on the indoor siren Single Event Level (SEL) - a measure of total acoustic energy - and the sleep-awakening model developed by the U.S. Environmental Protection Agency.[51 The graph used for estimating the chance of alert during sleep is shown in Fig. 4-2; for the Indian Point analysis, the curve for the chance of awakening one out of two sleepers was used.
~
4.4 Alerting People Out of Doors For the analysis of the ability of sirens to alert people out of doors, background noise levels are based on noise measure-monts conducted by BBN in the vicinity of the Trojan Nuclear plant in Oregon, near the Indian Point Nuclear Power Station in New York, and upon'the body of data in BBN files. The data typi-cally consisted of statistical summaries of background noise at various types of locations. The summaries provide the L 90 (sound level exceeded 90% of the time) for 1-minute samples of data in the 1/3-octave frequency band centered at 630 Hz.*
- The L 9 level.0 was used as a conservative estimate of the minimum sound -
4-10
- , 'i u. 3 - . . .
, .y ,
l 1 l 108 . l [
~
- l
*^ I j j
/ /
*^
j j j gg l
.. l l
/
1 Out of 3 Sleepers, j # g o_ .
/
/.
10ut of ) /
. j- / /' 2 5-"'"/
l ". / / /mi
- f .,- y S-
[/ l* f 30
//
v r
/~
/
=
y - 1e SEL = L + 24 For Stationary Sirens
; SEL = L + 18 For Rotating Sirene.
5 [ For 4. Minute Siren Operation P The Middle Curve is Used For
. 2 This Analyus 1.o 0.5 M
8.1 (Le5 e-45 BB S0 70 00 90 100 110 120 130 140 LNDOOR SIREN SEL ( m ) - FIG. 4-2. CHANCE OF ALERT POR ANAKENING PEOPLE ASLEEP.
~
l 4-11
f These data were used to estimate the range of background noise levels that are likely to exist during any 4-minute period (1 minute for rotating sirens) for a variety of outdoor environ-ments. The results are summarized in Table 4. 4, which specifies the background noise environment for urban and rural areas. Only daytime noise levels are presented since the . nighttime scenarios assume that essentially no people are outdoors at night. .The siren sound level necessary to alert is 9 dB greater than the minimum background noise level that could exist in any 4-minute period (1 minute for rotating sirens) , adjusted for the probabi-lity distribution of such minima. The chance of alert for people outdoors was determined for each scenario at each listener site using Figure 4-3. Outdoor background noise in urban areas and along rural roadways is caused predominantly by motor vehicle traffic. It is generally insensitive to seasons of the year, but varies markedly _ with time of day. Minor. traffic variations (i.e., less than a factor of 2 in traffic volume) have little effect on the back-ground noise. In rural areas remote from roadways, outdoor background noise can be seasonal (birds, insects, etc.) and can vary with the weather (wind, rain, waterflow, surf). Few people live or work in such " natural" acoustic environments. Note that rotating sirens would actually produce their estimated sound level during about one quarter of the presumed 4-minute operating time at any particular listener location. Thus, the results for rotating sirens are based on 1-minute statistics rather than on 4-minute statistics. 4-12
i
- s. i. (, . .
1 . . w. l l TABLE 4.4. MINIMUM BACKG3OUND NOISE LEVELS FOR GENERALIZED CATEGORIES OF OUTDOOR ENVIRONMENTS (SEE FIG. 4-3 FOR DISTRIBUTIONS). Generalized Background Range of Minimum Background Noise Environment Noise Levels for a 1-Minute Periodl,2 (dB) I. URBAN-DAY (Includes Rural locations within 21-57 1000 ft. of major roadways) II. RURAL-DAY (Except Rural locations within 17-48 1000 ft. of major roadways) NOTES:
- 1. Refers to the range of the minimum (L90) sound pressure levels in the 630 Hz one-third octave band during the specified time period.
- 2. Applicable for analysis of rotating sirens operated for 4 minutes.
- 3. Urban locations are defined as the pink " building exclusion" areas of topographic maps, or as those communities with a population density exceeding 2000 people per square mile.
Major roadways are defined as roadways with more than one lane in each direction.
- 4. Rural locations are taken to be all sites not classified as urban (above).
4-13
t t
,_ p 6
e _d, 00 99.9 99.8 99.5 99 98 RURAL-DAY 3 [ g s ; / A
.0 / /
e ,0 / / g ,0 / / t ) A
! .,0 / / \ _
e' -/ / \ s 20 // tuR..u-oxv
- 10 lI 5 / /
2 / / 1.0 l 0.5 l a, . 0.1 0.06 OT / 0 10 20 30 40 50 80 70 30 OUTDOOR SIREN LEVEL (dB) FIG. 4-3. CHANCE OF ALERT FOR PEOPLE INDOORS (4-MINUTE ROTATING SIREN). 4-14
4.5 Alerting People Indoors For the analysis of alerting people indoors at home, three types of activities are considered. These are (1) listening to radio or TV, (2) sleeping, or (3) other activities that range . from quiet to noisy situations. Table 4.5 provides the percentages assumed for various activities for each scenario. For people listening to radio or TV, the chance of alert is 100%. For people sleeping, the chance of alert is calculated from the indoor siren SEL using the relationship shown in Fig. 4 for' the chance of awakening one out of two sleepers. For all other indoor activities, the chance of alert is based on classi-fications of actual indoor background noise measurements under a wide variety of conditions. Results for test cases 1 and 3 are provided in Fig. 4-4 for 4-minute rotating sirens. Thus, given the indoor siren level and test case condition, this figure was used to estimate the chance of alerting for indoor activities other than sleeping or listening to radio or TV. For the analysis of alerting at work, two activity cate-gories are considered: (1) commercial / institutional, and (2) industrial environments. For the Indian Point analysis, it was assumed that 75% of the working population are in commercial es-tablishments while the remaining 25% are in' industrial locations. For commercial locations, the chance of alert is based on the statistics of background noise measured in a typical office envi-ronment, using Pig. 4-5. For industrial locations, it has been assumed that 100% of the people are likely to be alerted by some means of communication other than sirens. 4-15
3 s . t .' v . s~ l 1 ,
\ '
s l-
, t s 1, g w TABLE 4.5. ASSUMED ACTIVITIES AND BACKGROUND 11OISE ENVIRONMENTS '
FOR PEOPLE INDOORS.
~
as 5
- s Percentages of People Engaged in Various Activities Indoors (5) 'N
._, , Indsor snise Environeent At Place Listening to Obviously Busy and ]bviously Scenario l3 of Business TV/ Radio Sleeping hoisy 8 Active 8 Isolated' Quiet'
, , 1. Warm $Mer We-\ day 41 27 5 - - - 8 5 14
- s. Attec noon (clear to s,
+ pactly cloudy)
- 8. Suemer Wee' a day Night 4 --
96 - - - ,s~ (clear to partly cloudy) -
- 3. Winter Weekday Durir. --
20 -- 5 50 20 5 Evening
- Commuting Hours (cold and overcast) t 4 , Winger Night Durits 5 -
93 - - - -- Snowfall ' s g .- -
. \
NOTES: -
- +. s
- 1. Vacuum cleaning, diahwasher. ' shower. vent fan on, etc. %[N.
- 2. Dinner coneeregcios, kitchen work, playing musik..Nhildren at play. etc. "
- 3. Noise-product g activity in adjacent room, su?t' t!'14 . round music, eccJ
- 4. Reading, study."asting alone.
' ) ' %,
_ J s- ( ,, *
- 3 s
% v
, s 4
r *
% x t
-.- y N
\
N i n s
* * \ ',
~
4 A g
* ~ ,
.s '
% 3; {
' N 4-16
~
100 99.9 99.8 J 99.5
- 9. 1 /
9, If
,, /f
. //
g , /A/ 70
- n. -
s r o s l ) Io J
-y j l
s_e w u 30
,, f f 10
, I ,#"
1.0 0.5 f 0.2 0.1 1 0.os o o to 20 30 40 50 00 70 so so too INDOOR SIR EN LEVEL ( db ) l l FIG. 4-4. OkmICE OF ALERT FOR PEOPLE INDOORS AT HOME (4-MINUTE RMATING SIREN). 4-17
9 0 l 100 G 99.9 99.8 4 MINUTE STATIONARY SIRENg f 4 MINUTE ROTATING SIREN 99.5 99 _ t 98 95 '
= $~! ~
80 e 5 70 w
;i so N
a 50 J7 l 40 ? 8 30 1 - . . 20 10 5 2 1.0 0.5 0.2 0.1 0.05 2 0 10 20 30 40 50 80 70 80 90 100 INDOOR SIREN LEVEL (dB) FIG. 4-5. CHANCE OF ALERT FOR PEOPLE INDOORS AT WORK IN COMMERCIAL / INSTITUTIONAL ESTABLISHMENTS. 4-18
9 4.6 Alerting People in Motor vehicles The analysis for the alerting of motorists is based on the '- assumption of an average siren signal strength and spacing throughout the EPZ. The chance that a motorist will pass within . the alert range of a siren during its 4-minute operation is . estimated as follows: C = 2R+d L x 100 (not to exceed 100%) where C is the chance of alert (%), R is the maximum alert distance (ft), d is distance traveled in 4 minutes (ft), and L is the average siren spacing-(ft). Separate analyses were carried out for urban and rural areas of the Indian Point EPZ. The average urban or rural siren produces a sound level of 125 dB at 100 ft. Alerting ability was evaluating by using the results of a study for the Society of Automotive Engineers (SAE). [61 Siren alerting levels for' speeds of 55 mph and 30 mph with windows shut or open were first determined from the SAE study results. The average siren source levels for rural and urban areas were then reduced to alerting levels in accordance with the , propagation models from current NRC guidelines (i.e., 10 dB/ double distance) [7] . In this manner, the maximum alert distance * (R) was calculated for each driving condition. The distance traveled in 4 minutes (d) was calculated based on speed for each case, and the average siren spacing ( L) was estimated to be 4,890 ft for urban areas and 12,530 ft for rural areas. The calculations of alerting ability for motorists are summarized in Table 4.6. The results indicate that the chance of alert is expected to be 100% for all conditions applicable to the Indian Point analysis. 4-19
; TABLE 4.6. SIREN ALERTING FOR MOTORISTS.
Vehicle vehicle Reqd. Signal Max. Alert 4-min Avg. Stren Chance Speed Window for Alert Dist. R Travs1 Dist., d SpacingL of Alert Area (sph) Condition (ft) (d8) (ft) (ftg (5) l UnaAM 30 closed 89 980 10,560 open 4.890 100
; 86 1.210 10,560 4,890 100 i
ButAL $5 closed 96 650 19,360 12,530 100 Open 90 980 19,360 12,530 100 t a I i l 4-20
i
- 5. EVALUATION OF THE PROMPT ALERTING SYSTEM POR THE ZION NUCLEAR POWER STATION This section summarizes _ the evaluation of the siren alerting system for the Zion Nuclear Power Station. The procedure that was used consists of a detailed analysis of siren alerting capa-bility at each of 50 randomly chosen listener locations, under four different " sample scenario" conditions. The random selec-tion process for listener sites is described in Appendix N and the four test cases (sample scenarios) are included in Appendix
- 0. The analysis is based on exis' ting and proposed siren loca-tions as of 15 October 1981. Maps which show the siren locations are provided in Appendix P.
The results of the' evaluation for Zion are summarized in Table 5.1 and indicate that the chance of alert is estimated to vary between 58% and 97% depending on the sample scenario under consideration. The remainder of this report describes the
~
procedure used to arrive at this conclusion. Input and output data for the analyses are included in Appendix Q. 5.1 Estimating Siren Sound Levels Out of Doors at Listener Sites The first step in the procedure is to determine the siren in
-the vicinity of each selected listener site that is expected to produce the highest sound level at that site for each sample scenario. This choi'ce is not always obvious, because the sound level caused by a particular siren at a given listener site de-pends not only on the sound output of the siren and its distance from the listener, but also on shielding and atmospheric ef fects (particularly wind direction)~. Therefore, it is generally neces-sary to evaluate several sirens .in the vicinity of each listener site in order to determine the dominant one. As a general rule, the closest, highest-rated, nonshielded sirens are selected for evaluation at each site. Furthermore, sirens should be chosen 5-1
i TABLE 5.1.
SUMMARY
OP ZION SIREN SYSTEM EVALUATION RESULTS. Chance of Alert Scenario Popula tion-Urban Rural Weighted Average
- No. Description (%) (%) (%)
1 Warm Summer Weekday After-noon (clear to partly cloudy) 97 96 97 2 Summer Weekday Night (clear to parely. cloudy) 81 74 80 3 Winter Weekday Evening (cold, overcast, light ' 90 85 89 precipitation) 4 Winter Night (windy) 59 51 58
- Based on a total urban population of 268,629 and a total rural population of 33,201.
5-2
such that they are distributed north, south, east, and west of the site (or in any other four mutually perpendicular directions) where possible to account for different wind directions. For the Zion analysis, four or five sirens were evaluated at 46 of the 50 listener sites. Only two or three sirens were considered at the remaining four sites; these sites were located at the fringe of siren coverage such that sirens were not present~in all directions. The next step in the procedure is to establish the outdoor sound level produced by the selected sirens at each listener location. This is accomplished by applying adjustments to the rated sound level of the siren as follows: L(listener) = L(siren) -Ad-As -Aair ~ Aa tm ' where L(listener) is the outdoor siren sound pressure level at the listener site (d B) , L(siren) is the rated sound pressure level of the siren at 100 ft (dB), Ad is the distance attenua, tion (dB), A s is the shielding attenuation (dB) , A air is the air absorption (dB), and Aatm is the atmospheric attenuation caused by wind and temperature gradients (dB ) . l The rated sound pressure levels for all the proposed Zion sirens were obtained based on information prov' ided by Commonwealth Edison, and are as follows: ACA Dual-Tone Rotating Sirens = 123 dBC 0 100 ft ACA Single-Tone Rotating Sirens = 126 dBC @ 100 ft
.Whelen Electronic Rotating Sirens = 124 dBC @ 100 ft ACA Stationary Sirens =
115 dBC @ 100 ft The rated sound pressure levels for existing sirens to be employed in the Zion system were taken to be 125 dBC for rotating 5-3 1
( units and 115 dBC or 100 dBC for stationary units, all at a distance of 100 ft. The first two adjustments (for distance and shielding) are the same for all four test cases and are based on information obtained from USGS maps. Distance attenuation beyond 100 ft is calculated by assuming sound propagation from an acoustic point source with a reduction of 6 dB per distance doubled. It is calculated as follows: Ad = 20 log 10 (IU)' where d is the siren-to-listener distance (ft). Shielding attenuation (As) is estimated using the following formula for the attenuation of a rigid straight barrier for sound incident from a point source [21: 1 - As" < l20 log tanh /77R
+5 dB for N > -0.2
= 0 dB for N < -0.2 N is the Fresnel number (dimensionless):
N=* (A + B - d) where A = wavelength of sound, ft (1.79 ft for 630-Hz siren tone) d = straight-line distance between source and receiver, ft A+B = shortest path length of wave travel over the barrier between source and receiver, ft
+ sign = receiver in the shadow zone (i.e., barrier obstructs line-of-sight) - sign = receiver in the bright zone (i.e., barrier doesn't obstruct line-of-sight) 5-4
1 When N is negative, the above equation for As is evaluated by replacing N with lNl, and by replacing tanh'with tan. Shielding attenuation is limited to a maximum of 24 dB based upon a large body of experimental data. For the Zion analysis, sirens are assumed to be at a height of 25-60 ft above terrain level, listener sites are assumed to be at a. height of 5 f t above terrain level, and barrier heights are obtained from ground contour information on USGS maps.. The adjustments for air absorption and atmospheric ef fects depend on the meteorological conditions for the particular scenario. The assumed conditions for the Zion site are provided in Table 5.2 for the four test cases, based on local weather information.* In terms of air absorption, these conditions indicate the following attenuation rates based upon temperature and relative humidity [2,31: Scenario Aair (dB per 1000 f t) 1 0.85 2 0.85 3 1.0 4 2.0 The adjustment for atmospheric gradient ef fects (Aatm) is based 'on siren-to-listener azimuth with respect to wind direction and on wind and temperature gradient characteristics. Table 5.3 summarizes the calculation procedure for determining Aa tm fr each scenario at the Zion site. A more detailed description of the estimation procedure for A atm can be found in Appendix D.
- Commonwealth Edison, Zion Nuclear Power Station Weather Data Re co rd s.
5-5
TABLE 5.2. METEROLOGICAL CONDITIONS FOR TEE FOUR SAMPLE SCENARIOS USED 'IO EVALUATE THE ZION SIREN SYSTEM. Relative Scenario Temperature . Humidity Temperature No. Wind Conditions
- Gradient ( ". ) (OF) 1 11 mph from the southeast -1.30F/90 ft 60 71 2 12 mph from' the WhV +1.10 F/90 ft 60 70 3 11 mph from the Nhv -0.70F/90 ft 95 17 4 33 mph from the WSW -0.80 F/90 ft 76 13
*At 125 f t above ground level.
5-6
TABLE 5.3. CALCULATION OF ATMOSPHERIC ATTENUATION, A ata' CAUSED BY WIND AND TEMPERATURE GRADIENTS (SEE APPENDIX D FOR DETAILS) . Siren-to-Listener Distance,0(Ft) Relative to Xn (Ft) Aatm (dB) D 1 1. 2 I o 0 1.2'X, < D 1 1.7 Xo 5 1.7 X, < D 1 2.4 X, 10 2.4 X, < D 1 3.4 X, 15 3.4 X, < D 20 Computation of Xn X, = .f 1057/[eBCos$-a Scenario 1 2 3 4 Wind Direstion, e . 1300 2900 3280 2510 AT F-(125'-35') -1.3 +1.1 -0.7 -0.8 a = az = AT/ (In 125 '-In 35 ' ) -1.02 +0.86 -0.55 -0.63 Wind Speed, V ft/see @ 125ft 16.3 17.2 15.8 48.4 2 V1 ft/see @ 35ft 10.1 8.9 11.6 32.3 l ZS - (V2 -Y1 )/(in 125' - In 35') '4.87 t 6.52 - 1.27 12.65 y s R/S 0.1 f(R/S) 0.45 e n X (min) @ $ = 0 436' 444' 784' 290'
&c = Cos (f) 1020 82 1160 93o 5-7
Application of the above calculations yields the estimated outdoor sound pressure level for various sirens at each sample listener site, for each of the four scenarios. For the balance
~
of the analysis, only the highest siren level at each listener site is generally used. An. exception to this rule is made at listener sites where the sound level of a stationary siren is estimated to be between 0 and 6 dB lower than the sound level of a rotating-type siren which had been determined to be the loudest siren. In such cases, the stationary siren was selected for fur-ther analysis. The reason for this exception is that the maximum sound level produced by a rotating siren is not continuous, and thus the total acoustic energy at the listener (as measured by the single event noise exposure level, or SEL) is approximately 6 dB less than for a stationary (i.e. , continuous) siren with the same maximum sound level. ( S.2 Estimating Indoor Sound Levels of Sirens - The result of the above calculations is a single outdoor siren sound pressure level at each of the 50 sample listener locations for each of the four test cases. Corresponding indoor levels are then obtained by subtracting typical values for com-mercial or residential building sound attenuation. For tes t cases 1 and 2 (summer), residential windows were assumed to be partly open; for test cases 3 and 4 (winter) residential windows were assumed to be closed (with storm windows) . For the fre-quency region within the 500 Hz octave band, the sound attenua-tion into buildings is estimated to be 16 dB for test cases 1 and 2, and 31 dB for test cases 3 and 4 [4}. For commercial build-ings, the outdoor-to-indoor noise reduction is estimated to be 31 dB, assuming closed and sealed windows for all four scenarios. t 5-8
5.3 Assumptions about Chance of Alert The outdoor and indoor siren levels calculated by the above procedure provide some of the information required for the analy-sis of the chance of alert. In addition, it is necessary to know the level of interfering background noise at the listener locations. Figure 5-1 is a flow chart of the analysis computations. The analysis is divided into components (rows) that correspond to
~
the possible activities of people for the various scenarios. The major components relate to people (1) at'home (outside or inside), (2) at work, or (3) in motor vehicles. The chance of alert is estimated for each activity component and is then multi-plied by the fraction of people likely to be engaged in that activity (activity fraction). The results are summed to obtain the, overall chance of alert for each listener location and for each test case. Overall chances of alert for the various scena-rio (test case) conditions are then obtained by averaging the chances for all rural and/or urban sample listener sites. Note that all estimates assume siren signal duration of 4 minutes: an average of the "3 to 5 minutes" called for in Appendix 3 of NUREG-0654. The effects of different siren signal durations are discussed in Appendix E. Siren detectability is a function of the siren signal level and of the background noise level in a " critical frequency band" centered at the signal' frequency. For this analysis, outdoor and indoor detectability is estimated based on the signal-to-noise (S/N) dif ference in the 630-Hz 1/3-octave frequency band. The chosen criterion for alerting is that the given signal level must be 9 dB or more above the minimum background noise level at any 5-9
'/ DE Pt mes LaPose . ICE tvt G N DE Pt anos uPOW BCE ER. Apg
^ m o,,,_ '~l ** - o - F-T,47 w- .T.O. O - , ,;,,,,,,-
C.T. ., -
- T ,*' ~ ,'/'.,.*'. -~ a-
- m. .+
* * '$ ~
i U y ' ' ~ e a_ g#, ...,- ,r h =>m g _.g &'-&I:::.,l M M e & -L-J--llEj = IE_I ED--L'+
'. a - f- 3 LiEj = l '= 1 CM-EG
- ',nt. t,.,;_.tv .
e r =[. j : ; . -@ Se '
~
I See VlaesCLES
**es S FIG. 5-1. FLOW OF COMPUTATIONS.
l time during a 4-minute period for people who are not sleeping (i.e., a S/N difference of 9 d'B). The chance of alert while sleeping is based on the indoor siren Single Event Level (SEL) - a measure of total acoustic energy - and the sleep-awakening model developed by the U.S. Environmental Protection Agency [ 51.. The graph used for estimating the chance of alert during sleep is shown in Fig. 5-2; for the Zion analysis, the curve for the chance of awakening one out of two sleepers was'used. 5.4 Alerting People Out of Doors For the analysis of the ability of sirens to alert people out of doors, background noise levels are based on noise mea-surements conducted by BBN in the vicinity- of the Trojan Nuclear Plant in' Oregon, near the Zion Nuclear Power Station in New York, and upon the body of data in BBN files. The data typically l consisted of statistical summaries of background noise at various types of locations. The summaries provide the L 90 (sound level exceeded 90% of the time) for'l-minute samples of data in the 1/3-octave f requency band centered at 630 Hz.* These data were used to estimate the range of background noise levels that are likely to exist during any 4-minute period (1 minute for rotating sirens) for a variety of outdoor environments. The results are ( summarized in Table 5.4, which specifies the background noise environment for urban and rural areas. Only daytime noise levels are presented since the nighttime scenarios assume that essen-l tially no people are outdoors at . night. l l The siren sound level necessary to alert is 9 dB greater l than the minimum background noise level that could exist in any l l l l *The L 90 was used as a conservative estimate of the minimum sound level 5-11
100 [ g
- I 99.9 l 99.8 99.5
, / ! l
. / / I
/ l l , ,
1 dut of 3 S!'espers,
, / )
g eg F. / /^
/
/.1 Out of 2 Sempers
/
/
m p yn 1 2 g /
- , / / / ^ ' s~
f/ f l~3,
// / -
2a
/
l v < 10 SEL = L + 24 For Stationary Sirems 5 SEL = L + 18 For Rotating Siruna. p,, 4 , y;,,,, 3wan Operation The Middle Curve is Used For 2 This Analysas 1.0 0.5 (12 0.1 0E O l 40 50 80 70 80 90 100 110 120 130 140 INDOOR SIREN SEL ( db ) l l FIG. 5-2. CHANCE OF ALERT FOR MAKENING PEOPLE ASLEEP. i 5-12
TABLE 5.4. MINIMUM BACKGROUND NOISE LEVELS FOR GENERALIZED CATEGORIES OF OUTDOOR ENVIRONMENTS (SEE FIGS. 5-3 AND 5-4 POR DISTRIBUTIONS) . Generalized Background Noise Environment RangeofMinimumgackground Noise Levels (dB) 1-Minute Period 4-Minute Period 3 I. URBAN-DAY (Includes Rural locations within 21-57 21-57 1000 ft. of major roadways) II. RURAL-DAY (Except Rural locations within 17-48 17-47 1000 ft. of major roadways) NOTES:
- 1. Refers to the range of the minimum (L90) sound pressure levels in the 630 Hz one-third octave band during the specified time period.
- 2. Applicable for analysis of rotating sirens operated for 4 minutes.
- 3. Applicable for analysis of stationary sirens operated for 4 minutes.
- 4. Urban locations are defined as the pink " building exclusion" areas of topographic maps, or as those communities with a population density exceeding 2000 people per square mile.
Major-roadways are defined as roadways with more than one lane in each direction.
- 5. Rural locations are taken to be all sites not classified as urban (above). -
5-13
( 4-minute period (1 minute for rotating sirens), adjusted for the probability distribution'of such minima. The chance of alert for people outdoors was determined for each scenario at each listener cite using Figs. 5-3 and'5-4. Outdoor background noise in urban areas and along rural roadways is caused predominantly by motor vehicle traffic. It-is generally insensitive to seasons of the year, but varies markedly with time of day. Minor traffic variations (i.e., less than a factor of 2 in traffic volume) have little ef fect on the back-ground noise. In rural areas remote from . roadways, outdoor background noise can be seasonal (birds, insects, e tc. ) and can vary with the weather (wind, rain, waterflow, surf). However, few people live or work in such " natural" acoustic environments. ; Note that results are given separately for stationary sirens - and rotating sirens. This is because rotating sirens would cetually produce their estimated sound level during about one quarter of the presumed 4-minute operating time at any particular listener location. Thus, the results for rotating sirens are based on 1-minute statistics rather than on 4-minute statistics. In summa ry , information regarding siren type, estimated ciren sound level, background noise category at the listener site, and test-case conditions were used in conjunction with Figs. 5-3 and 5-4 to estimate the chance of siren detection outdoors at the Zion Site. 4 5-14
8
- 1M f
99.9 99.8 99.5 99 i ;
,, I I /
,, - /
. /T /
RU R A L - DAY-
.0 b 70 60 Y
[ [ \ RBAN-DAY U s" e, I ,>
/ /
j 30 i
' i / #
/
= l 20 / /
10 ' // 5 I / r
/
i 2 /
' I / I I 0.5 ^I I i 1 0.2 l
0.1 0.06 0 0 10 20 30 40 50 60 70 80 OUTDOOR SIREN LEVEL (d8)' PIG. 5-3. CHANCE OF ALERT POR PEOPLE OUTDOORS (4-MINUTE STATIONARY SIREN). 5-15
100 l CO 99.9 99.8 99.s 99 98 / es RURAL-DAY s i / f ee y /
" / /
g70 / / g / sa
/ /
sa / / e" / / N-g= \ f[ tuR As.OAv 20 10
' ; /
' I I 1.0 l 0.s 0.2 j
0.1 l 0.05 l OT 0 10 20 30 # M N OUTDOOR SIREN LEVEL (dB) FIG. 5-4. CHANCE OF ALERT POR PEOPLE OUTDOORS (4-MINUTE ROTATING SIREN). 5-16
- 5. 5. Alerting People Indoors For the analysis of alerting people indoors at home, three types of activities are considered. These are (1) listening to radio or TV, (2) sleeping, or (3) other activities that range from quiet to noisy situations. Table 5.5 provides the percen-tages assumed for various activities for- each scenario.
For people listening to radio or TV, the chance of alert is 100%. For people sleeping, the chance of alert is calculated from the indoor siren SEL using the relationship shown in Fig. 5-2 for the chance of awakening one out of two sleepers. For all other indoor activities, the chance of alert is based on genera-lized categories of actual indoor background noise measurements under a wide variety of conditions. The ranges of minimum back-ground noise levels for these categories are listed in Table 5.6. Results for test cases 1 and 3 are provided in Fig. 5-5 for 4-minute stationary sirens and in Fig. 5-6 for 4-minute rotating sirens. Thus, given the siren type, indoor siren level and test case condition, these figures were used to estimate the chance of alerting for indoor activities other than sleeping or listening to radio or TV. For the analysis of alerting at work, two activity cate-gories are considered: (1) commercial / institutional, and (2) industrial environments. For the Zion analysis, it was assumed that 75% of the working population are in commercial establish-ments while the remaining 25% are in industrial locations. For commercial locations, the chance of alert is based on the statis-tics of background noise measured in a typical of fice environ-ment, using Figure 5-7. For industrial' locations, it has been assumed that 100% of the people are likely to be alerted by some means of communication other than sirens. 5-17
(. TABLE 5.5. ASSUMED ACTIVITIES AND BACKGROUND NOISE ENVIRONMENTS FOR PEOPLE INDOORS. I d. Percentages of People Engaged in various Activities Indoors (I) Indoor Noise Environment At Place Listening to Obviously Busy and %viously Scenario of Business TV/ Radio Sleeping Noisy' Ac tive 8 Isolated' Quiet'
- 1. Warm sua.or Weekday 41 27 5 -
4 S 16 Afternoon (clear to , partly cloudy)
- 2. Summer Weekday Night 4 **-- 96 -- -- -- --
(clear to partly cloudy)
- 3. Winter Weekday Durina -
20 - 5 50 20 5 Evening Commuting Ilours (cold and overcest)
- 4. Winter Night During 5 -
95 -- - -- -- i Snoufall NOTES:
- 1. Vacuum cleaning, diehuesher, shouer, went fan on, etc.
- 2. Otaner converset ton, kitchen work, pteying muele, children at play, etc.
- 3. Nosee produc e
.g e.-ttwity in adjacent room, soft background music, etc.
4 Reading, study, eating stone. J w 5-18
i t TABLE 5.6. MINIMUM BACKGROUND NOISE LEVELS POR GENERALIZED CATEGORIES OF INDOOR ACTIVITIES / ENVIRONMENTS. Generalized Range of Minimum Backgroynd Activity / Environment Noise Levels in dB 1-Min. Period 2 4-Min. Period 3 4 At home, obviously noisy (i.e., vacuum cleaning, 41- 76 41- 73 d is hwa'she r, shower, vent fan on At home, busy and active 4 (i.e., dinner conver-sation, kitchen work, 21-64 21-54 playing music, children at play) At home, isolated 4 - (i.e., noise-producing activity in adjacent 23-49 23-38 room, soft background music) A't home, obviously quiet 4 (i.e., reading, study, 11-39 11-28 eating alone) At work, office and commercial 28-49 28-45 NOTES:
- 1. Refers to the range of the minimum (L sound pressure levels in the 630 Hz one-third octave 90) band.
- 2. Applicable for analysis of rotating sirens operated fo.r 4-minutes.
- 3. Applicable for analysis of stationary sirens operated for 4-minutes.
- 4. To simplify the procedure, these are combined into a single indoor range on the basis of the activity fractions in Table 5.5.
5-19
# f 100 M
99.9 99.8 99.5 99 I
,, /1
,, //
.0
./ /
# 8C :
[ C 70 Scenario @ ' [ y Scenario @ 8 1 ~ f s ,,,, JY / - u 30 / ' 7 ..
- 1 .
/
10 5 2 1.0 i l 0.5 0.2 0.1 l 0.05 l C l 0 0 10 20 30 40 50 60 70 80 90 100 INDOOR SIREN LEVEL ( dB ) FIG. 5-5. CHANCE OF ALERT POR PEOPLE INDOORS AT BOME (4-MINUTE STATIONARY SIREN). 5-20
9 4 99.9 99.8 99.5 1
,9 1 /
99
,, /f
. //
g B. / / [ Scenario h a: 70
;, sr Ei [ /
s .so 0 / / u 30
.-y j s-,,. w
,, / /
10
/ .
" lJ
, f#
1.0 ] 0.5 0.2 0.1 0.05 0 , _ 0 10 20 30 40 50 60 70 80 90 100 INDOOR SIREN LEVEL ( db ) FIG 5-6. CHANCE OF ALERT POR PEOPLE INDOORS AT HOME (4-MINUTE ROTATING SIREN). 5-21
a . 100 99.9 99.8
. 4 MINUTE STATIONARY SIRENg f4 MINUTE ROTATING SIREN 99.5 99 a 98 95 '
' d/
- 80 e
$ 70 w
Y 4m a, JV a g 40 U, 5 30 , J 20 1 10 5 2 10 0.5 0.2 0.1 0.05 0 l 0 10 20 30 40 SO 60 70 80 90 100 INDOOR SIREN LEVEL (dB) FIG. 5.7 CHANCE OF ALERT POR PEOPLE INDOORS AT WORK IN COMMERCIAL / INSTITUTIONAL ESTABLISHMENTS. 5-22
5.6 Alerting People in Motor Vehicles The analysis for the alerting of motorists is based on the assumption of an average siren signal strength and spacing throughout the EPZ. The chance that a motorist will pass within the alert range of a siren during its 4-minute operation is estimated as follows: 2R+d C= g x 100 (not to-exceed 100%) where C is the chance of alert (%), R is the maximum alert distance (ft), d is distance traveled in'4 minutes (ft), and L is the average siren spacing (ft). Separate analyses were carried out for urban and rural areas of the Zion EPZ. The average urban siren produces a sound level of 123 dB at 100 ft and the average rural siren produces a sound level of 124 dB at 100 ft. . Alerting ability was evaluating by using the re-sults of a study for the Society of Automotive Engineers (SAE) 16] . Siren alerting levels for speeds of 55 mph and 30 mph with windows shut or open were first determined from the SAE study results. The average siren source levels for rural and urban areas were then reduced to alerting levels in accordance with the / propagation models from current NRC guidelines (i.e., 10 dB/ double distance) [7]. In this manner, the maximum alert distance (R) was calculated for each driving condition. The distance traveled in 4 minutes (d) was calculated based on speed for each case, and the average siren spacing (L) was estimated to be 5,045 ft for urban areas and 19,240 ft for rural areas. The calculations of alerting ability for motorists are sum-marized in Table 5.7. The results indicate that the' chance of alert is expected to be 100% for all conditions applicable to the Zion analysis.- 5-23
't , .
3
.N .; *
. ~ .
JN r a k N wy " TABLE.5.7. SIREN ALERTING FOR MOTORISTS. m. N
? Vehicle vehicle Reqd. Signal Max. Alert 4-min Avg. Stren Chance Speed Window for Alert Dist. R Travel Dist., d Spacing L of Alert Ared- (mph) Condition (68) (ft) (ft) (ft) (1)
URBAN 30 Closed 89 1.000 10.560 5.045 100 Opena 86 1.200 10.560 5.045 100 RURAL 35 Closed % 700 19,360 19.240 too
\
- Open 9u ' , - 1.000 19,360 19.240 100
% s - ,
s 6 e
~
6 { u << - ..
., t .. +
% N
,\:
4
. 's
. , s 2- ,
t i A.* w
- b N
,, 4 "g r g-
+
~,
,, 4 s ,4'-
. g, .
i i ,
\
4 /. i . s s' ' 5-24
i REFERENCES
- 1. Keast, D.N., Towers, D.A., and Anderson, G.S., " Procedure for Analyzing the Ef fectiveness of Siren Systems for Alerting the Public," BBN Report No. 4774 ( Februa ry, 1982). -
- 2. L. L. Beranek, ed., " Noise and Vibration Control,'" McGraw-Hill Book Company (1971).
- 3. C.M. Harris, ed., "Hardbook of Noise Control," Chapter 3, McGraw-Hill Book Company (1979).
- 4. D.A. Driscoll, J.P. Dulin, Jr., and D.N. Keast, " Attenuation of Northern Dwellings to a Linear Source of Noise."
Presented at the 95th Meeting of the Acoustical Society of America, Providence, R.I. (May 1978).
- 5. U.S. Environmental Protection Agency, " National Roadway Noise Exposure Model," Draf t Report (1980).
- 6. R.C. Potter et al., "Ef fectiveness of Audible Warning Devices on Emergency Vehicles," U.S. Department of Transportation, Report DOT-TSC-OST-7 7-38, W a sh ing' ton , DC (August 1977).
- 7. " Criteria for Preparation and Evaluation of Radiological Emergency Response Plans and Preparedness in Support of Nuclear Power Plants," Nuclear Regulatory Commission, NUREG-0654, Revision 1, Appendix 3, Part C. 3.e, Washington, DC (November 1980). (The Federal Emergency Management Agency also publishes this document as FEMA-REP-l.)
i I l i l 1 5-25
APPENDIX A: POPULATED-WEIGHTED RANDOM SELECTION OF LISTENING POINTS AT THE TROJAN SITE The objective of the listener-site-selection process is to identify 50 randomly selected residential locations within the 10-mile EPZ surrounding the Trojan Nuclear Plant. It was arbi-trarily decided that 40 sites were to be in rural areas (population density below 2000 persons /sq mi) and 10 sites were to be in urban areas (population density above 2000 persons /sq mi). Of the rural ~ sites, 20 were to lie within 5 miles of the plant and 20 were to lie between 5 to 10 miles from the plant. These ground rules were established based on site-specific information.
~
The various steps used in the site selection procedure are described below:
- 1. The boundaries of urban and rural areas were defined on a set of USGS topographical maps covering the EPZ.
Those regions denoted by USGS as " building omission areas" on the maps (pink shading) were assumed to be e urban (with uniform population density), and all other areas were assumed rural. The urban areas included only the communities of Longview and Kelso.
- 2. A population distribution drawing (see Fig. A-1) consisting of a 10-mile-radius circle divided into annular sectors defined by interior circles and radii was superimposed on the U.S.G.S. maps.
Population distribution information consisted of the number of people within each annular sector. These data were used to population-weight the random selection process for rural sites as de-scribed in Step 3 below. A-1
a TROJAN NUCLEAR PLANT i DOPt8L AleON 84 POPUL Ai80% wit HIN &NNUL AR RING $ R ADI AL Os5f ANC[ of $11[
- 0. 8 g,6 I l 176 t - 2 487 *N 2 593 3.J 3322 3 2185 34 2315 4 4983
- 4. $ 3406 $ 3389 S. 30 $6307 30 64626 s e .* .
# 4
, 4 15062 3 4 3 4 ,,,,, . ,,/ afLSO 1308
' ~ . *
- 6., 4. y . .. .,,
rog 'J-e0
$8 132 * "**
226* ,, 256 gn
/ s>>..y 3 3 ,
260 ,,3 0
,,6 s 2>>
. V .
- 34. 64 0
sn [ 32 122 6, 23 gg 0 8 44 - 23 t,,
- 63
...s ,
(/)04C0 . - t .
- 20. ..
l6 io
,,0 ,, ,, ,,,
2n so o 6 o4 o 2a ,, '.. - 6 60 3.
,0,.
, j..
326 8*2 sia 4
- 46 ,2si 0
ISO g3 259 33 221 28 67 200 569 g 9 20e set 896 0 FIG. A-1. 1980 PROJECTED POPULATION DISTRIBUTION WITHIN 10 MILES AND RANDOMLY SELECTED LISTENER SITES (APPROXIMATE). A-2
3 Each annular sector was first assigned a number. A range of numbers was then assigned to each sector according to the population in that sector. For example, Sector no. 1, due north of the alte, had a population of 90 and thus was assigned numbers 1 through 90. Sectors 2 through 4 (moving clockwise) had zero population and were therefore not assigned any numbers. Sector 5 had a population of 17 and was assigned numbers 91 through 107. T'his process was continued until each number between 1 and 18,600
~
(the total estimated rural population) was assigned to a particular sector. A random number generator (available, for example, on a Texas Instruments Model TI-59 hand calculator)'was then used to select 20 numbers at random between 1 and 8,293 (repre-senting sectors within 5 mile ~s of the plant) and 20 numbers at random between 8,294 and 18,600 (representing sectors between 5 and 10 miles from the plant). Each number selected represented one site (to be chosen later) within the sector assigned to that number. Thus, sectors with larger populations would have a greater probability of including chosen listener sites.
- 4. Having determined the sector location for each rural listener site, the next step in the procedure involved selecting the actual location of each site within its respective sector. This was accomplished by first overlaying a rectangular coordinate grid on each sector of interest on the USGS map. The grid was composed of boxes with dimensions of A-3
/
4 approximately 1000 ft sq, and each box was assigned an X and a Y coordinate according to its location on the grid. The grid-was positioned so that the X axis was oriented in the east-west direction and the Y axis was oriented in the north-south direction, and so that all parts of the sector of interest were covered by a positive (X,Y) coordinate pair box. A random number generator was thin used to select random pairs of numbers within the X and Y ranges including the sector of interest. Each X,Y pair was used to locate a particular 1000 ft sq box on the USGS map. If no residences were inside the square area or if the area fell outside of the sector of interest, the coordinate pair was disregarded and another pair was chosen at random. This process was continued until a square area including one or more residential structures was found in the sector of interest. The listener site was then chosen to be any residence within the randomly selected square area. In this manner, the sample of 40 rural listener sites was selected. ! 5. The selection procedure for urban sites was similar to that for rural sites, except that each distinct urban area was treated as a sector and population was assumed to be uniformly distributed throughout each urban'se~ctor. Thus, random number pairs were used to select square' areas on the grid, and a listener site was chosen anywhere in that area provided that the site-fell within the urban sector of interest. In this manner, the sample of 10 urban listener sites was selected. A-4
The above procedure resulted in a random sample of 50 listener locations, distributed throughout the EPZ as shown roughly on Fig. A-1. 4 4 A-5
APPENDIX B: TEST CASES (SAMPLE SCENARIOS) FOR THE TROJAN SITE
- 1. Warm summer weekend day, weather clear to partly cloudy
, People 70% out of doors 20% indoors 10% in motor vehicles (windows open)
Buildings: Windows open Vind: 10 mph - from the north throughout the region upslope in the canyon Temperature Gradient: -2*C/100 m; Class A Relative Humidity: 50%
- 2. Summer weekday night, weather clear to partly cloudy People: 95% indoors, sleeping 4% indoors, at work '
1% in motor vehicles (windows closed) Buildings: Windows open Find: from the north on rid'ges and plateaus east and west of the site 5 mph - from the south in the river valley downslope in the canyons Temperature Gradient: +1.5*C/100 m; Class E Relative Humidity: 90% 1
- 3. Winter weekday during evening commuting hours Cool, damp, and overcast People: 70% indoors 25% in motor vehicles (windows closed) 5% out of doors Buildings: Windows closed B-1
. 4 i
( Vind: 3 mph - from the south
. calm in the canyons Tempera ture Gradien t: +1*C/100 m; Class E Relative Humidity: 80%
- 4. Winter night during rainstorm People: 95% indoors, sleeping 4% indoors, at work 1% in motor vehicles (windows closed)
Buildings: Windows closed Wind: 15 mph - from the south 5 mph - downslope in the canyons f Tempera ture Gradien t: +1*C/100 m; Class E Relative Humidity: 90% ~' Sources: EIR, Amendment'1, March 1973, Fig. 2.3.4, Tables 2.3.3 and 2.3.13. Site-specific wind velocity profile inversion at Trojan has not been considered. B-2 s
- - , , . - - - -- - .-w ,
APPENDIX C: SIREN IDCATIONS FOR THE TROJAN EPZ This appendix provides siren locations for the Trojan EPZ on a set of maps (1-6). A siren location map index is provided which shows the relationship of individual maps to the Trojan EPZ. Table C.1 provides information on the type and rating for each siren, as well as a guide for locating the ' sirens on the maps. G l I t 1 ! C-1 l
- - - - --- --ry mm- ,, , , - - - - - _ _ - . - - - - - - - - - -+ ,,,.7 , .c - --.,y-
. l p
{ V . ..~~ ) . ~ .>.- -
/.1 % m A _ ~ yn . . . y --
{$p r g.M winggXfg.) 3... k -3 4 ' ;{, g j ~.m p gg *
~
! I
?- 19 i ,-
~
~
ag ., A + -
~
7 :v n g
+ .N.
,(( . ew :n. h ,
h,p .G.C[,
~
% ky-p:Y
~
?g$N'$gl/.l
~
b.; f'J ;j?.(3 . h; ? C W'; 1
.Q
,1 f.Q:Rl ? ~~
N . XLs'~%E3IE&f ,MM_f,t r e M4 ' ., k ' 3 / ~~
-' ' a ' / -
S $ h h kk F a , b c p-l./4 '-T g? h f -
,g e
I f.9 S 4 ..-)* ' $'a h h R y^ h
? ,,,n.Y& r, -
Q-a WB&'l l) h
'7.
, Nb % 1.@5
%.o, ,
r,
' . ' W '-s;;
~
, / ^ < r~ *hv P -(
^
f , n N !, , w.'w% I l5' 3,, :, c j D. W r. N:??yi, l
'$h b h' $
hQ). ,)/ p) .~ _ I,. '
?'ly ,- - ., <-
2 i 3, %W,.gi v: V s w:>;w,:., - W =- ,p'sp'( ~ ~
~,,.*~-
-,9.
..t<i.;i.. b..g j.~
- m =. ~ . a '%. w .e
. . . , Q.
.- c.
7 I .. ~ . . q' w:rT %u,9- .
... t
,1
= 4'
~v
' s k ; : _..V f W' 0 g> < s.
-'/-M c L / v. - ' 3.
'$ /y> w$N' W, 0-t ] .? .A3- -A h:f: f;:t$nk k..-gll*;.c,..::%:
li 1.:
, Yh?AY .*%.Q9%., }- <h
/y ~ a..
.l hN(T .h, ".T..i;,., '
, I , x' .f y,hk- ..hp.
s
- , [. ,,,7 [.h. ,- l
;. ' r . ' ~ =
,. g j 3-g--c .-y .:; i9:. [ ? +- :y. ' ' <' . c.e j x:q99-
' 5%. , c-t. ~ ; .J r "n .4 . , ,
- n. v.,
_' t j.7 3, M . , H TRodAN EMEF7-4.EMcY WARMikib.6Y5TE1'i .. . _6lRE}J L.OGATick fMAE WDEX _ C-2
., <, N (c, '
1g
- \ ' _.
(
' \ '. , q" ' . 'Al LOW > ROVE ~ '] //-_: ,h i
'N I ._.
\x\
M. o. w
% .r o
1,0 Coda. m/ s. ~ P~ " D" q
' 's #r* %
%. g N %~ Y "Q*%gy ,
- yy-.+ f
? MD.
e ----
\ 1 4r 4834 W~
%1
+p N a -
g +
.. ,t N ~.1 o.
r o.,; -
;s e -
,. : ::; 4,, . %'Ngs x ..
N i cy *+soy I _y _3 l Downing- * ' ' g 4 * [-- \ s A*-3 s'9m a* 2 -
- r. -
/
s. m
.y sO
's N
( - s * . yQ s y x \ El0 ' O ,ts /\
- I w I' -
I \" ~. cox 404l AD. , Pe f 5
%l :* <r. ,
,,,_ , e
.c <c ' o%* ktp
, '!*p RLLD RO AM o-
!ss ,
r q, y* o . E 4037 Q.& '- a"
- @g g
Mg e w e !r:
/ a vsee s aC
/-, root f
~ F aratu J I-1 X
/. 4 4:47 j 8'88
- j %f
-/
a n. ((% -
/- P ==
elena
* *' * =g N 'q,
# . g? .- :
5 9.3.'ALSTON , _
-m.s wonocam 43 g,<<. 4# 4' l*
'Toy l ww. '
28Z
, way //
i
'. , y q , ,o %,,- ~ _ p .
' s [ e , * '- ,
~ ~r. &
"[:
i s E
\ p',# i Zo23 *
% . - *>4 g '%
N / cauuan r no.
'1 5 %
w: 1. i 9 y
#,e s f
~
/ 'I i h1 CEDAR
~
*o n '% $0 A h
! N --g J*
q- g
\ . ,
l
~
4 [h coco *Cao.- X 'm ER mo. S t. i 2r (' 9 Q. oo,=
'* 3 t . .
~f. A
V i Q '
/,o*7 m /
m '
?,Y (NSwede W town c
j M.. (
\ 2 g
% .. o.2.5, y ,
1s . fri S
- i.
- s
- MAP 1 l . . .
. I? C-3 % - - -
i g ,, yr- -me e.ed < ' l .
}
i i l'
.l,'
'BEXCgNIin
(<. *) [ ' wkgl ' Y
#, ' s. n.. - ~%, {
mshsl. ._
. 1 1 .-
%::,~hhf{ho3 a 1
\_f ["# v' g
f f V-17 v.E
'A ~
k
, s 4 w-ii7 4
~-
' i
.l L +. v-95 s.-
i
)l ~ .: . e
- x;h . ,
- a. c w-ior n .
l , ,, g -
.M 4 7 , ,
/A 4 [/,
.w.i ;
A#W st
-i w,uz ,lh;- ! W_. 90-I J
Mj ,,4 ,,% a sr[fCV4og.% q-4 ~
, go w ""s pgv-N g p.
E)d j I ( w-as .
.whii[% #% c "^ " . ', ..
../
, k ' N/. \My,' hj
~
\, "' ~ "
g-5 N -
- h[(eg . -
N, '
'O' J T
P 8 ,
, ./
N g ro,' ( t J y _. s e.g. w /%stu .,,,
~~
\ ~.
K g' i .'
- A ~, *
. V ~
.s, buffo. ? l t
g t zim , ug. ,
. \ ,,'N g -
'[db5Ad} dis
{'* ~
, f ,;;;
i f ,,y;;; 2 ,- N, p ,g, - 7 _=
,,,, . =: r _ _ ,
gI w ._ t J ::-;;. .p :.: .:,:. . ) ' 3.s ..., pg,, R A I N IERg= (.9
\ q '
s i .v/ ,.72 ,M.w-7t, s,,, -.- h
.l[ 2 s.
r ,g 4(0-as ' t-g % ' ,, - i LI DB - G g\ F ~t A t'; g$ & , 1 i
*~
; , '%.u%-
'L -Q \,e g\ 1 \ 'i ^??*b a I' g.
&? ,-(~./
'. 2j /' .sl n50-27 7:1
~ i l
.- - S -
g J . (
; , ~ -
;;M' 90, *y -iz)d_ Q\Q >. . = --
Se ern e , r ,'%, K' o 0-3 s'
^fl, .
[' ./ Hill [ 50-2/, ~ "~
\ 0 Eo
,J-1 o-
'Q ( '-5
- s. '
y. i ., /l '
.( F. r 'E-
- s 130- ',
i- , ~ L._.
'.' ww ) idi ' 3 4.%l ;;P"'
- i
""*s f J' t, v,. s o-w... , . i- - MAP 2.: .~ ._.
'f O-26 \ = ^ v' ~, e ..s _ . i . c-4
ser
\
.rmpg e. ,.w:e
')
i a w:. v:.: ys
- :;^.
~w A: m l
~ x. ~y NEAc L*-@. -
s Q *'C q' f. . _~ _.%s.4EEWAw V' ' , r wy
', s o
;: .e d,p. \ +\
e ~ 9. 4 ,.; ,
- y' , '
f ,Y...,.+.>
\
%4; % i
[ M:C".. 7kkI'Q4
' %. '4 f.
i~ a./ . .
% ) J'g j~dj.Q:iW .
.hr h,g (( ' ' . C 41MBIA, my9- ,,;.
., gx. 7. x7.c s .'
e~. " .,
- " n.
~~ : u' ~g,W s . 4 5, 5 y Q 1. H :
1 3.-J. e _gJ'DP U A c
.* a j
~g
't i 1A
' f.
- .9 2..
nv
, .; cFi .Mn,.' . Q$
r s a a. , g N g [ 3 /# 4 y s, t 5 5 i. . x . u. e t'>a.d- f' [ ." N, i ..
+ .
1 N ) .;* * ,
-t' a -
j w.
/,1.?i:.h % .
/ '&.. ;W. ,. m i- N _. um w,.,m if_ 9.m.6
\^ . g' w&. ...Y%
m %"# . '
;l <
- M b*
' N.4 ui n,e i,h"$ tl$'Ee a>;,.w ' C+ M' 4,E , . 1 K. , 9/
3,<;, - A i, .. ik ;
- . 3sl,- W 8'
\, gNv(b
.1
,g ;g. t;y , 7;
~+-
,, f ;
.+ ., , t 1
N.. M i
} VIg V f. ..
u m M ,w,. 2..(m,//,,
.,o I, ,
.,s q
t r,e -p .a *-
\
g"y m
~
% s j, > % Q ;; .,. n,
\. * $ ll E
.g g 't I
'N f , q.yg.
f, Y % < , {_j ' F
\.,
.y '4
,. s 4 <i . ,.
N,
, ,4 , a% i
.g ' .
i
/,
'% s : wm-j #s
%, 3'
%'# iJ .
i A' o ~.c,
+
'[/S*!f.
-}.
- r o ,
. .. . . ,, o
, / vla _ w t, s M
'i s : ;-
,a*# **
C-5 e MAF' Z a : c- - ~ - -
l
.__ , 9 48 8 W
, . .% -, s t . i ..
N11!,;9
,,,,,f, c.
a' f izy y - i w,. p FJ ' , .N .J
?
e...W' ! - 1 .
- 3. x I _
@ Hil.l.
E
.,.." ./
O sj'. ,, - l . w ,y / . .
' l Nf < l l
.;. l \
- , r N u I - =.-
'Ot c o e,
~
. ee e e.
i 9- M.3) /,/,Cp 1 1 5 II :
...., L, . ... I I q , 'g*- .
i
's c-
. M. ;., ; I
/: ...t.L m.
s . s , t o J
~, I I
i
., I -[ Tm -
03 - { {,"\ \ :)
.s .c q.97, i J' (ar ,
l 1[jg- , 7 , ** l n
.- g. b. . 'q
~ '
.$ k,!;?.,' _
fN'
- r . ,1 BUE'~e'ICR)S \. l
,l / -
r-- h y d. .kh5 [. [gi, ,j' \ l ___ . i ; i '\. _\ \ l hf@g-gu_0Ep?k%pot L ~ >uw.h k [ k: .y/ .
\
_> \.\._. q!%Il4hfh ;T *'d 3g 3%. 's'{'\ - I _. m . . s , i r + a/P;j
,.f t-J.-.' i 5 g,v'.:f e.y . ., l, l
,a s .
L..y; .4
~"
,g %
n-
*13 "'
O[ .,of,. 7; 1""\, % ' Lfy,ViyfY f y}' i& 94 / IN) R W
?~l$?'
r.}l a$$f f s p y. W_?' x x,
\\ ('
_ ic. ..
, e. =
a
- g.
f j. * //38 Y M [ M, '* 8' \
- g':
' J.:. / a_
I N__ 'w.j n o g - j', . o
% ..,\
t
".n.q --- 6 ,,,y-1 q} b_._\.. ..
b 7. s . 3 s
.m - ,/
e '
, a.. y 4
'b. y '.
. '.11-
. u.
s . 2....; L w
}, .- l E yp
* ,'; . Q p { u y 'f f ' $A " ". _. -3;.? M % @g., J< '
g--/ -
~ -
< r_ - &g \ ' ' s\. N
'- a 1 ?"". L'- - ; !
4 .y', mar u
, _ , , . .J h * , ^*'"
s b ,, ,,
'k\: . . .
$ gog E .-,./ I
.... r. -
p g_g . - WiO ~ "
, .% N ...T...._._,
b# ']h .qi ~ 7_ .%f ' .._. c-6
~
~
_ C. _ \
--M~ AE Zb
)
i 1 ,--- ,-
' t f- L. 1 l
t' - ie 4 . .. (
\-c-z oT ~. ... . t \ ,. -
}/.A <- - >
m i ..# _- g,3 p" , il 4 ' n j
' ~
..7 g. / t., ..
l .. -
% g k
s i
'g W-s5 ~
n- '
\ a\ '
s u ,, l \s no
, ~,s \, ,,,,,,
, c
\ ~
~
,._. 5% 6k. .' ~ ~'
%/&g ,\
c, is " j n l y s u t
< g,,
! , .8,, ,.
x
, _ - -_.y& -
- m. , . _. ;, g s w:- N
.na x bV M *
%,, l j .s v %?W .
' + , '. e E*@ -.
E%.EY h- '
~
u M a;,- Af ' x-
),,,
vC , y +je h*"y/ww..,..
- j. .
.e +N h
~.
St,?- s' e3 , V37
)5 e#g., M s
y
,=\ f l is 4 . ua l :.
7c E w-r 7 -' s $
- 3 ,
, 1- -
W 7f # "'" ' '
- s. \ [e V'k "f&. ,,, , ,; ,/v. a.
i u-s+ " w-55l s.
/e , w 53 e
- s %
\
\
w- su
%v f;
m ( 1,. ; ,' -
}(,. ,:i j ,.
/
st 'J, 3 w.5g , , , < - I hk% . r 3- .-
/.52 .
3
*i 0
. wse ; 1 y y ,
w.43 , _ _ g g 3 ,,
'" ~ " "
.__..C-7
~ ,
"*I S.
g , l . V$
. n on co. +.,4- '
!
- 2073' I
. ? *
. 0 3t * .sf \
i
\
l
\
s [ O l Y
~
} \ j t
[j i smuss mo.\. , i ~ zios ' .j 5
) i
/ l Apiary 0-S naw J .I} (
, l. -
I a-y
y. ._- N/ :-
s a y , ;4 .g h o.1 s - v . I
, s 3.. . .
i i >.
; - \ g:.w I .n -- m m _ _
s a ( s \ , i +,a i
\;
\) - - '
i g%e
+c
, ( 0-+ j , ; o.3o f-g.
... k '
./ g)
N: ;
}}
/ s
/ ./ ,.- m,- n .
\: 2 ;'t\,r 2 o.m.
.g,. g's.
/ -
- r. .
y 1 . 6 h ,, esh 5 / l} lf / =
. ~ .. j e21 c l' .
e , s
's r s ,oor
- - r- :
o.a 8 %.. I e+j != '.
)
g . y ~
~ e,
/ r x ...
\
\ -% ~,
_r,,f '
~
g- , . s ,
~
+e rh '
E pp , , .o M,%eg .i
\ - -
A '- --- *44;, '\
? ~- wc l-
/ 3 1 s; %y i '.9 %
- X- .
.f-
, i
/ <
/ ,. -:
+o w ' u ,
< z
/- --
1 ? 7 Q _e .4 . n. i .
/
- /% / a*" %r.1 U 7s Mt. 8 /-
- q.
D, - ~ y l^ v., Cedars un -
' x f e M~AP 4-b ,o 3 / i /
...... . . - - . c_, ..- --
'- II o.23 9'
~
7,o.g g g - ( N- Ngg%.;( O-2.9
~
~: zoso **
N l 1
&g*. i ( '!. 9 ,,38I p' ' q2 I gs .t-. 's
.u.
, Nr;6 Y g- ~ca m'.,. on p c , , , *s i'3 no. i i '. v .3 . /
) i i i.
i l *dp7
; ab~st2 3
\ .
J
's i! rid 1' V $27 _
$ 0.n A %;. . ,
l
- N. f 'N
\ +e Si ; . ,
\\.-
% 94 4, # e oY2 ao. A 9, h'
x g 2. . O b ;al g,f ,.N 2 4 - 1 , g g .
' )?.,, Q_ [*
9 : ch \
\
; 'Nr [ l, l[ jy"".E'I I o-si
}
ko
~
"' 5 " "
i , % .
- a 9 N ,f
% .. / 'q
\
}eg h5gt% S Q q, 'gi '
r,,o 2. Eo
~ .' 807, 0-+7 . *, 0% EE. M, l '
no % Tier , c a. 1.g .
^ ^ ^ fL.g(, . , b
, i' '
E\ Shilohp*
,B asin .
zu ." - "
,. fh . .
a-
.*? ,. g ,,
f
! O nuntenS no k \ o O-57 / Sozs jg s s.. 'ss **, , '
s s,
=. .-
t e Tv . . .; \
\
\
e,P'3 '? /\,' gssun m. Canaan; y .33[t _44mug,'g o
!! \ 'g
- s. [
b p""U j 'N, - -
,,jLN us sk"" , '.~ g
\ lo-sa /
m - ISLAND c=E l2 l ,
) - _ - - #{ ,
l
\lPA .
\ -
J i 'w \. p i% . OL\
/ 3 ITrenholrn'o s 7 %'* '
,\
q 5 x y A a o
\ ,
, 3g \
j g
$ "re M(E9-ssgg,, ' * *
- l h g k'. k 7 . m l b -
./ E,'*** i %% 'g n
/ .m MAP 5
-. _ ._. c-9 ._
I y.m. .
\ .
tr' '\* Y
,',', 4
- g. *o y3 g e.
w * *t-3 QP W.42
~ * **
4~m C
' \n._
,p
/
)3 ^
w y, y.
' } e w44
'.. I
. ,.37 k
e
\ w31 -
g'
#m M '<
/
et
,l k i%_.M 4
. f ~%
e e i
'Ed3sf e=iig *l
- e. @ j f
( , 3,
$ { d' A. ,W-3..
1 _ ._.
'i- ,,[
% W3 / V' 'l . f 4et s 0 cc
=g,
(_ (11 o #, ll ,, ,, .. e <
) R3 4 mc N T S, .. ,j,i i
' N, *y fg
- g. O,. e
[ .' l \ **k)
>. a . n y i
/
3 * ,- q ..p..
,.i .
,- so n
'. W-l N.,' sI ,
E.rt W.,,. m e s, no...W-i3:( 3 w.s n . A ,'f
* *mi
- i _w.zs mN s e. ' #. . o,S - '
,j w-io . /
\ } \f'r 30 . .
c
- f. T l y [
,, y .s
\,,'y& v.~.vr =4 a
,, s.
E s. N_.
~ zr a .
'di&,f ,w nn(' ]& n w-c.l .
*\ , .
W-5 .% i n e- ...- - 1 3 .
/ .
M'"i *~'j*ibl 1
~ '
r i 5
\.
y wg *
*. - M S .om.g
/
- ' 3
\r,< ;"o-<
i.. , ., '- ;. .g" , 1. I : # i 's . E' . . N* '
. 8 J' '
\ t .
> ,s (p .e. g g) 4-
\* 5 Kr' e, / u ;s y>r 4
.k. x_
S. } ( [ ,. . a ***, s j
'J i .e .o n
, -( -
t
,,,,,,,- i}(
h, g 0:st no s, --a - L 4 :. At155 : e
.,.,,1
! 1 f I / W-17 a'/#, -. =* _ ., .
. i 33f. 3 I Q] 'j t
f , ., e t
. 1 -
. tven. e* po'8
, .. g k's l k , .,e .-q 4',. i, Y\ W00Dt.AND fIA .c.. , n l ., \n;;., -. .
.n u <
" !!; - \n x'. #
. M I i,k \.
*sD 1 *
/~ QllQ ~jj lz , y,
, T, . 'A : .
I .~'Il C-10 .. t. MAF' ~ 6_
TABLE Col. TROJAN SIREN INFORMATION. Location Rated SPL Siren No. (Map No. ) (dB @ 100 ft) Type
- 0-1 1 125 R 0-2 2 125 R 0-3 4 102 S 0-4 4 86 S 0-5 4 86 S 0-6 1 125 R 0-7 1 125 R 0-8 4 125 R 0-9 2 125 R 0-10 2 107 S 0-11 2 125 R 0-12 2 125 R 0-13 1 125 R 0-14 1 125 R 0-15 5 125 R 0-16 5- 102 S 0-17 5 115 S 0-18 5 125 R 0-l~9 2 125 R 0-20 2 125 R 0-21 5 102 S 0-22 5 125 R 0-23 5 102 S 0-24 5 107 S 0-25 1 102 S 0-26 2 107 S 0-27 5 102 S 0-28 5 125 R 0-29 5 102 S 0-30 4 102 S 0-31 4 125 R 0-32 4 102 S 0-33 5 86 S 0-34 5 115 S 0-35 2 107 s 0-36 2 102 S 0-37 2 102 S 0-38 5 2 86 S 0-39 2 102 S 0-40 2 102 S 0-41 5 102 S 0-42 5 102 S
- Rotating (R) or Stationary (S)
C-ll
[ TABLE C.l. TROJAN SIREN INFORMATION (Cont.). Location Rated SPL Siren No. (Map No. ) (dB 9 100 f t) Type
- 0-43 5 125 R 0-44 5 102 S o 0-45 5 102 S 0-46 5 125 R 0-47 5 125 R 0-48 5 107 S 0-49 5 102 S 0-50 5 115 S 0-51 5 115 S 0-52 5 125 R
*0-53 5 125 R 0-54 5 125 R 0-55 5 102 S 0-56 5 107 S 0-57 5 125 R 0-58 5 102 S 0-59 6 125 R 0-60 6 125 R 0-61 3 102 S ,
0-62 5 125 R W1 e 125 R -- W2 6 125 R W3 6 102 S W4 6 86 S WS 6 86 S W6 6 102 S W7 6 102 S W8 6 86 S W9 6 102 S t W10 6 86 S Wil 6 102 S W12 2 102 S W13 6 86 S W14 6 107 S W15 6 102 S W16 5 125 R W17 6 125 R W18 6 102 S W19 6 102 S W20 6 86 S W21 6 86 S l l
- Rotating (R) or Stationary (S) c-12 l
l
TABLE C l. TROJAN SIREN INFORMATION (Cont). Location Rated SPL Siren No. (Map No.) (dB e 100 ft) Type
- W22 6 102 S W23 6 102 S W24 6 86 S W25 6 102 S W26 6 102 S W27 6 102 S W28 6 102 S W29 6 125 R W30 6 102 S W31 6 102. S W32 6 102 S W33 6 102 S W34 6 115 S W35 6 86 S W36 6 86 S W37 6 102 S W38 5 115 S W39 2 125 R W40 6 102 S W41 6 86 S W42 6 86 S W43 3 102 S W44 6 86 S W45 6 102 S W46 6 102 S
, W47 6 102 S , W48 , 6 102 S W49 6 102 S W50 3 102 S
- W51 3 107 S
, W52 3 86 S i W53 3 86 S W54 3 102 S i W55 3 86 S W56 3 107 S W57 3 86 S
- W58 3 102 S W59 3 107 S W60 3 86 S W61 3 86 S
, W62 6 86 S l
- Rotating (R) or Stationary (S)
C-13 n,- -
[ TABLE C.l. TROJAN SIREN INFORMATION (Cont.) ~ Location Rated SPL Siren No. (Map No.) (dB 9 100 f t) Type
- W63 2 102 S W64 2 107 S W65 2 102 S 8
W66 3 102 'S W67 3 102 S W68 3 102 S W69 3 102 S W70 3 102 S W71 3 102 S W72 2 115 S W73 2 102 S W74 3 102 S W75 3 107 S
.W76 2 107 S
. W77 2 125 R W78 3 102 S W79 3 102 S W80 -
3 102 S W81 3 107 'S W82 3 107 S W83 3 -102 S -- W84 3 102 S W85 3 107 S W86 3 86 S W87 3 102 S W88 3 102 S W89 2 102 S W90 2 125 R W91 - 2 102 S W92 2 102 S W93 3 107 S l- W94 3 86 S W95 2 102 S l W96 2 102 S l W97 2b 125 R W98 26 125 R W99 2 115 S W100 2b 102 S W101 2b 102 S W102 2b 125 R W103 2b 125 R
- Rotating (R) or Stationary (S)
C-14
r- , TABLE C.1. TROJAN SIREN INFORMATION (Cont.) Incation Rated SPL Siren No. (Map No.) .(dB 9 100 f t) Type
- W104 2b 125 W105 R 2b 125 R W106 2b 107 S W107 2b 125 R W108 2b 125 W109 R 2 125 R W110 2a 125 Will R 2a 125 R W112 2a 125 W113 R 2a 125 R Wll4 2a 125 Wil5 R 2a 125 R Wil6 2a 125 Wil7 R 2a 125 R 5 R i
l C-15
1
~
TABLE C l. TROJAN SIREN INFORMATION (Cont.) Location Rated SPL Siren No. (Map No.) (dB 9 100 f t) Type
- W104 2b 125 R W105 2b 125 R
. W106 2b 107 S W107 2b 125 R W108 2b 125' R W109 2 125 R W110 2a 125 R Will 2a 125 R Wil2 2a 125 R W113 2a 125 R Wil4 2a 125 R Wil5 2a 125 R Wil6 2a 125 R W117 2a 125 R Wil8 2a -
125 R Wil9 2a 125 R W120 2b 107 S W121 2a 102 S W122 2a 102 S W123 2a 102 S W124 6 107 S - ( l 1 l
- Rotating (R) or Stationary (S)
C-16 i
APPENDIX D: ESTIMATION OF A ATM The speed of sound in air increases with the square root of the absolute temperature. When the atmosphere is in motion, the speed of sound is the vector sum of its speed in still air and , the wind speed. The temperature and wind in the atmosphere near the ground are almost never uniform. Hence, atmospheric nonuni-formity produces gradients of the speed of sound, and thus refraction (bending) of sound wave paths. Near the ground, this refraction can have a major effect on the apparent attenuation of sound propagated through the atmosphere. For the purpose of this procedure we have assumed a horizontally stratified atmosphere in which temperature and wind speed vary only with the logarithm of height above the ground. During the daytime, temperature normally decreases with height (lapse), so that sound waves from a source near the ground are refracted upwards. In the absence of wind, an " acoustic shadow" forms around the source (Fig. D-la) into which no direct sound waves can penetrate. Marked attenuations are observed at receiving points well into the shadow zone - it is just as if a solid barrier had been built around the source. At night a temperature increase with height is common near the ground (inversion) and our " barrier" disappears as in Fig. D-lb. Near the ground, wind speed almost always increases with height. Because the speed of sound is the vector sum of its speed in still air and the wind vector, a shadow zone can form upwind of the source, but is suppressed downwind (Fig. D-lc). The combined effects of wind and temperature are usually such as to create acoustic shadows upwind of a source, but not downwind. Only under rare circumstances will a temperature lapse be sufficient to overpower wind effects and create a shadow D-1
PATHS OF SOUND WAVES
- SOURCE ,
SH ADOW h x W'
- a. TEMPERATURE DECREASING WITH HEIGHT Typical Daytime SOURCE
[fy
- b. TEMPERATURE INCRE ASING WITH HEIGHT Typical Nighttime l
l l WIND DIRECTION l : SOURCE
#f,A 1"^N e
, c. WIND SPEED INCRE ASING WITH HEIGHT I ABOVE THE GROUND PIG. D-1. SKETCHES ILLUSTRATING 'ITIE EFFECTS OP VERTICAL TEMPERA-TURE AND WIND GRADIENTS IN FORMING ACOUSTIC SHADOW ZONES. D-2
f 1 surrounding a source. It is less rare, but still uncommon for a surface inversion to be sufficiently strong to entirely overcome an upwind shadow. The general situation is illustrated in plan view on Fig. 5-2. A shadow boundary, symmetrical about the wind vector, can exist in the upwind direction from a sound source when the ver-tical wind gradient effect predominates over any effect caused by a temperature inversion. It is likely that no shadow will exist downwind from the source, for the wind gradient will usually overcome the ef fect of any temperature lapse. Along a radius at an angle o fr m the wind vector, the shadow boundary (theore-c tically) approaches an infinite distance from the source. In the " upwind" sector of Fig. D-2, the sound wave paths are generally concave upwards,-as on the right side of Fig. D-l c . In the " downwind" sector, they are generally concave downwards, as on the lef t side of Fig. D-l c . In the "crosswind" direction, the sound wave paths are approximately straight lines from the source to the receiver. For the -purposes of this propagation model, we have assumed that temperature in the atmosphere, T, is horizontally uniform and varies with the logarithm of height above the ground, z.* T = ~ a inz T -T - 2 l
- AT T = Inh 2 - Inh 1
Inh 2
- Inh y l and ff=az~1
*This approximation is generally valid close to the ground except during strong surface-based temperature inversions [1,2] .
D-3
L N m I
\'
='
\
1 e . s
$ NORTH WIND DIRECTION UPWIND Ecyo n
, 4
/
'0 w RECEIVEM ~
SHADOW BOUNDARY
\
$[
/ ut :
de I g Xo 4c g! ; F l l SOURCE
\ ,
/.
/ - Nj l
\ g
\ s \8
\
,~
NWIND SECTOR
,], '
i
;(. . .c .
\ . ,i k,
;/- (. *
\/
4 l ');< 1 ' --" , FIG.' D-2. PLAN VIEW OP SOUND PROPAGATION SECTORS, WITH f PARAMETERS USED TO DESCRIBE THEM (see text). , i, 1
' j ' s
!f $
The speed of sound, c, varies directly with the square root of the absolute temperature c=c To+a (lnz-Inzfh o
=c o 1+a Unz - Inz n To , . 2Tg g where c o is the speed of sound at some reference temperature, T,o observed at a reference height of z .o Thus, th'e .ve rtical gradient of the speed of sound due to temperature,~a, is:
hE a= g az-1 cr 1.086az~1 o sec -2 in English units ( D- 2 ) Note that a can be positive (inversion) or negative (lapse). Likewise, we assume that the vertical profile of wind speed, 8, varies only with the logarithm of height, z, so that:* V 2 -V - Inh a - Inh t ( ~ where V 2 is the speed of height h2 and V1 is the speed of height at h y. Note that 8 is always assumed to be positive. The combined gradient of the speed of sound . C, resulting from both the temperature and wind gradients is thus
*This is a shakler simplification than that for the tempe ra ture profile [1], and normally holds only for near-neutral condi-tions [3). The actual shape of the wind profile is a function of surface roughness, and of vertical momentum transfer due to thermal instability. .
l l D-S
t C = z(6 cos 4 - a) ( D- 4 ) where $ is the angle between the direction from which the wind is coming and the sound path (Fig. D-2). . W Each sound path can be classified as " upwind", or " downwind" for a given sample of meteorological data, on the basis of the (o11owing steps.
- a. If a is positive and greater than 8(a > 8; so that C would be negative for all values .of 4), then no shadow zone can exist and all paths are classified as " downwind". This is the strong-inversion, low-wind condition.
- b. If a is negative and numerically. larger than 8 (i.e., l-al > 8, so that C would be positive for all values of
$), then the shadow zone completely surrounds the source and all paths are classified as " upwind". This is the strong-lapse, low-wind condition.
- c. If lal < S, then the " critical angle", &c, (where tem-perature, and wind effects cancel) is calculated by setting C = 0 in Eq. D-4 C = z(B cos &c - a) =0
$c = cos-1 aS ( D- M where 0 014c 1 180 l
It is now necessary to do some coordinate transformations of the azimuthal data, entered relative to true North, to bearings relative to the direction from which the wind blows. Refer to ! Fig. D-2. The wind-sound angle, $, is: D-6
&= 0 p
-0 w
, or if 6 -
0 > 180 : j p y 4 =.360 - O p
-0 Examine the difference $c ~ **
If 4 < +c then the path - is a " upwind" path. - If 4 > &c then the path is a " downwind" path. It is clear that this simplified model does not take into consideration some common effects, such as changes of wind direction with height and location and upper level inversions, which can lead to significant sound propagation to distances quite remote from a source. Computing the Distance to the Shadow-Zone Boundary, X Nyborg and Mintzer [4] have derived an expression for the distance, Xo (See Fig. D-2), from a sound source to the boundary of its shadow zone at the height of the receiver, R,ft above local ground, and in the presence of a vertical sound velocity gradient which varies with the logarithm of height. Their work
- has been adapted for. this procedure in the following form
X =S . f feet _ 47S R) ( D- 7 ) \ g f (S ;- in English units where S is the effective source height in feet above local ground, and the function f(h) is obtained from Table D. l. The l distance Xo is in feet and is assumed to be frequency-independent. ( D-7 l
( TABLE D.l. f(f) vs f for computing Xo in Eq. (E-7). (after Nyborg and Mintzer [4] ) . 6 R/S f(R/S)
< 0.05 0.4 0.1 0.45 0.2 0.55 0.3 0.6 0.4 0.7 0.5 .0.75 0.7 0.85 0.9 1.0 1
1 1.05 1.5 1.25 '- 2 1.5 3 1.9 4 2.3 5 2.65 6 3.0 7 3.3 ? 8 3.65 9 3.95 10 4.2
> 10 Set X, > D Interpolation is per:aitted, and fogImanual computations a graph of f(R/S) vs. R/S is most useful.-
D-8
l TABLE D.2. ATTENUATION WITHIN THE SHADOW ZONE, Aatm' VS SIREN-TO-LISTENER DISTANCE, D, FT. D 1 1.2 X 0 dB 1.2 X, < D 1 1.7 X, 5 1.7 X, < D i 2.4 X, 10 2.4 X, < D < 3.4 X, 15 D > 3.4 X, 20 Attenuation within the Shadow Zone, A u Theoretically, the attenuation within a shadow zone can be arbitrarily large for large distances beyond the shadow boundary. In practice, more than 25-30 dB is rarely observed because the loss of sound energy from the direct waves is partially replaced by the energy of indirect waves scattered from turbulence, ground surf ace roughness, etc. In this procedure, we have used representative values de-rived from the experimental work of Parkin and Scholes (6,71 and j Weiner and Keast [8). The recommended values (Table 2 of the main text) have an upper limit of 20 dB. Attenuation because of a shadow zone has occasionally been observed to decrease somewhat at extreme distances relative to closer-in distances. The con-( servative values in Table D. 2 allow for this possibility. D-9 l
l REFERENCES TO APPENDIX D
- 1. Keast, D.N. and F.M. Wiener, " An Empirice.1 Method' for Estimating Wind Profiles over Open, Level Ground", Trans.
Am. Geophys. Union 39, (5), pp. 858-864, October 1958. 5
- 2. Sutton, O.G., Microme teorology , McGraw-Hill Book Co. ,
London, Sect. 6.4, p. 206, 1953.
- 3. Ibid, Sect. 7.2, p. 232 et seg.
- 4. Nyborg, W.L. and D. Mintzer, " Review of Sound Propagation in the Lower Atmosphere", WADC Technical Report 54-602, Sect.
1.5.5, p.50 et seq., May 1955.
. 5. Ibid, Fig. 16, p. 54.
- 6. Parkin, P.H. and W.E. Scholes, "The Horizontal Propagation of Sound from a Jet Engine Close to the Ground, at Radle t t" ,J. Sound and Vibration, l_, pp. 1-15, 1964.
- 7. Parkin, P.H. and W.E. Scholes, "The Horizontal Propagation of Sound from a Jet Engine Close to the Ground, at _,
Hatfield",J. Sound and Vibration, 2_ (4), pp. 353-374, 1965.
- 8. Wiener, F.M. and D.N. Keast, " Experimental Study of the Propagation of Sound Over Ground",J. Acoust. Soc. Am., 31 724-733, 1959.
! V f I D-10 l l
APPENDIX E: DEPENDENCE OF ALERT UPON SIREN DURATION In the main body of this report, the chances of alert are predicted for a four-minute period of siren operation (here called siren duration). In this appendix, predictions are generalized for longer and shorter siren durations. This appen-dix will allow readers to convert four-minute results to results for other siren durations. This appendix begins with an overview of the relationship between siren level and siren duration, and how this relationship affects the chances of alert. It continues with development of the mathematics of this relationship, and then summarizes results for the reader's use. E.1 Overview Table E.1 is a typical " chance-of-alert" table for a parti-cular background-noise environment. Siren durations are listed across the top, and siren levels down the lef t side. Within the table are the chances of alert -- from 100 down to zero percent. In the main body of this report, results are based upon the four-minute columns of tables such as this one.* Variations within the table are related to fluctuating background noise in the listener's environment.**
*And upon the one-minute columns for rotating sirens. ** Precision within Table E.1 degrades for longer siren durations (to the right) and fer lower siren levels (to the bottom) . For longer siren durations, precision suffers from the limited amount ot total data that underlie the table. These data in clude 250 minutes of background noise, which is only about eight times the longest siren duration. For lower siren levels, precision suffers from the very small percentage of time that these low siren levels will alert the listener. Although the amount of data is large compared to the siren durations, the , background noise is rarely low enough to contribute to the statistics at these low siren levels. For longer siren durations and lower siren levels combined, the precision is particularly bad.
E-1 _.
\
TABLE E.1. TYPICAL CHANCE-OP-ALERT TABLE'FOR A PARTICULAR BACKGROUND-NOISE ENVIRONMENT.
*IFEff * ! F Ut 5 LEVEL' DUPAT!Dff (fit ff9TES) 1 2 3 4 5 6 7 3 9 to 11 12 13 14 15 16 74 800 100 100 100 100 100 100 100 100 100 100 100 len 100 100 100 73 99 100 Inn 100 100 100 100 100 100 100 100 100 100 100 100 190 72 ?? 99 100 100 100 100 100 100 100 100 100 100 100 100 100 100 71 ?3 -99 100 100 100 100 100 100 800 100 100 100 100 100 100 100 70 93 ?? 100 100 100 100 100 100 100 100 100 100 100 100 100 100 69 ?3 ?? 100 800 100 100 100 100 100 100 106 100 100 100 100 100 69 97 99 100 100 100 100 100 100 800 100 100 100 100 100 100 t on 67 97 94 100 100 100 100 100 100 800 100 100 800 100 100 100 100 si 95 93 100 100 100 100 100 100 100 100 100 100 100 100 100 100
$5 92 95 ?9 93 93 100 100 100 100 100 100 100 100 100 100 100
$4 92 95 ?? 93 ?? 100 100 100 100 100 100 100 100 100 100 100 63 37 92 95 97 ?S 100 100 100 800 100 100 100 100 100 100 100 52 tr ?! ?4 95 ?6 ?S 100 100 100 100 100 100 100 100 100 100 51 ?s 90 94 ?5 ?s ?3 100 100 100 100 100 100 100 100 100 190 60 35 ?O ?3 95 ?s 93 100 100 100 100 100 100 100 100 100 t on 5? 33 37 ?O 92 94 95 100 97 100 100 100 100 100 100 100 100 53 31 ?6 90 ?2 ?4 ?5 100 ?? 100 100 100 100 100 100 100 100 57 79 ?? $3 ?? 90 ?3 94 94 96 ?6 ?6 95 10n 94 100 93 5+ 73 S2 ST ?6 S3 90 94 ?4 ?i ?6 ?6 ?5 100 ?4 100 ?3 55 74 T9 32 ?3 34 ?6 99 37 39 92 ?! ?O 95 ?? 94 37 ,
54 70 73 95 , 93 34 Ts ?? 37 ?? 92 ?! 90 ?5' 39 94 37 - 53 s3 Ti SO ?! 34 35 99 $7 3? 92 ?! 90 95 ?9 ?4 57 52 45 75 77 TS ?2 33 SS ?4 ?? $3 97 ?0 ?5 S? 94 37 51 60 70 74 75 73 31 S3 $1 36 34 33 Si ?? ?3 33 30 50 55 i5 70 71 74 79 91 31 32 90 33 ?S 34 33 St 80 49 51 $1 67 13 70 76 ?? SI 7? 30 33 ?S 94 S3 31 30 43 43 57 63 65 66 78 75 74 75 76 73 St 7? S3 75 30 47 42 54 s0 43 54 71 . 75 71 75 76 74 31 79 33 75 ?O 46 37 47' 54 57 59 67 $9 69 71 72 70 76 74 73 69 73 45 33 44 50 56 Si 64 67 65 71 69 70 75 69 73 69 73 44 33 43 50 56 5i 64 67 45 71 is 70 75 69 73 59 73 43 30 40 45 54 52 57 64 is is $3 70 71 63 73 $? 73
.42
- 26 33 37 41 42 43 50 55 57 55 57 52 58 61 56 60 41 El 23 35 33 40 45 50 55 57 56 57 62 59 61 56 60 40 19 24 30 33 34 38 42 43 16 44 52 57 53 50 50 60 3? 14 20 24 29 40 76 36 42 43 40, 48 52 47 44 50 60 l 33 12 17 20 25 23 33 33 39 43 40 48 48 47 44 50 53 37 to 15 19 22 24 31 31 55 O? 36 43 43 47 44 50 53 26 3 12 14 16 13 El 22 26 29 24 ~ 10 29 12 33 31 13 35 7 ? 10 Ii 12 14 14 16 13 16 17 19 21 22 t? 00 34 6 3 e to 10 12 11 13 14 14 17 t? 16 1. 15 13 33 6 $ 8 to 10 12 Il 13 14 16 17 19 li 17 13 13 32 s 9 3 10 10 12 11 13 14 15 17 !? Is 17 33 13 31 4 3 9 to 10 12 11 13 14 16 17 !? 16 17 13 13 30 6 7 7 3 3 to 8 to 11 12 13 14 11 '11 13 13 29 6 7 7 3 3 to 8 to it 12 13 14 11 11 13 13 28 5 6 6 6 6 7 6 6 7 9 9 to 5 6 6 7 27 4 6 5 6 6 7 6 6 7' e ? 10 5 6 6 7 26 4 6 5 6 6 7 6 6 7 9 ? 10 5 6 s 7 25 3 5 5 6 6 7 6 s 7 8 ? 10 5 6 6 7 24 2 4 *5 6 6 7 6 6 7 3 ? 10 5 s 4 7 23 1 2 4 5 0 7 6 0 7 0 9 to 0 i 0 0 E-2
, a \
l In this table, the chance of alert is 100 percent when the siren level is much higher than the background noise could ever be at the listener. When the siren level is 74 dB, for example, the siren will definitely alert the listener even for siren durations as short as one minute. The chance of alert is zero percent when the siren level is low, say 20 dB or less, no matter how long the siren sounds. The background noise is always sufficient to mask (acoustically cover up) such low siren levels. For siren signals of intermediate levels, the chance of alert falls between 100 and zero percent, in the detailed manner shown. These intermediate details follow from the fluctuations of the background noise, from minute to minute. For these intermediate siren levels, the chance of alert increases with siren duration as indicated in the table. For a siren level of 50 dB, for example, the chance of alert is 71 per-cent if the siren is sounded for four minutes. If this duration is doubled to eight minutes, the chance of alert increases to 81 percent. How can this increase with duration be understood mathemati-cally? If such understanding results in a particular mathemati-cal pattern, then this pattern can be used to convert four-minute results to results for other siren durations. The search for ! this mathematical pattern is the subject of the next section. E.2 Development of the Mathematics The search for patterns within tables of numbers is neces-sarily an exploratory matter. First, some underlying mathematics must be postulated, and then a numerical pattern must be sought with this mathematics as guidance. Once a preliminary pattern is E-3
I
\ i l
l discovered, it must be simplified to be of use, and then must be generalized for other similar tables. Ideally, the pattern will emerge as a simple equation, with a small number of adjustable i 1 constants. The steps involved in developing such a pattern are:
- preparation
. underlying mathematics and its simplification
- exploratory graphs, guided by the mathematics
- simplification and' generalization to all other tables.
These steps are discussed next. E.2.1 Preparation . ( Figure E-1 shows typical background noise as it fluctuates over a one-minute period. The fluctuations are generally large, as shown here. In this background noise, a listener will be alerted by a siren,whenever it is 9 decibels or more above the background noise level.* The figure shows a siren that produces a steady 49 dB at the listener. A dashed line 9 dB below the siren level denotes the alerting threshold. During the shaded time intervals below this threshold, the siren will alert the listener. e
*Throughout this appendix , background noise includes the noise in a 1/3-octave frequency band centered at 630 Hz, a typical siren operating frequency. Dictated by the physiology of the ear, only this 1/3-octave band is available to mask, or cover up, the pure-tone signal of typical sirens. Siren levels are usually measured as overall sound levels, though the same values would be measured using only a 1/3-octave frequency band filter. .
E-4
i l l 80 i SIREN WILL ALERT f LISTENERS DURING l THESE INTERVALS 70 - g l 5 l \ 3 e}60 - EO E .E. 1 > 1 3o - I sz g S23l! i m g 50 - I i -
.x a SIREN M
m j 9 dB 40 --- ---- ----- -
" THRESHOLD MINIMUM 34 dB 30 '
0 0.5 1.0 TIME (minutes) PIG. E-1. BACKGROUND NOISE LEVEL AS A FUNCTION OF TIME. E-5
I This siren level has succeeded in alerting the listener during its one-minute duration. However, a siren level some 7 decih is lower would not alert because the background noise would always be above .its lowered threshold line of 33 dB. This figure suggests another way to phrase the alerting question. Instead of'asking if the siren is loud enough to cause alert, one could ask: For a given siren level, is the background noise ever low enough to allow alert? Since the background noise is continually fluctuating, this question is inherently a statis-tical question. Its answer depends upon the statistics of the background noise fluctuations. The answer to the above question is: Yes, alert will occur during this one-minute period if t (bbackground minimum I bsiren - 9 B otherwise, the siren will fail to alert the listener. . The only statis' tic of interest, therefore, is the minimum background noise level during this one-minute period.
- Figure E-2 shows a series of one-minute minima for forty successive one-minute time periods. Every minute's minimum is different, as the figure shows. These 40 minima were measured over a 40-minute time period, and are part of a much larger set (approximately 250) of total data. For the siren level shown, 35
*Our analysis for this study actually utilized the 90-percentile background noise level, rather than the minimum level. The 90-percentile noise level is the level exceeded 90 percent of the time; the remaining 10 percent of the noise falls below this level. Use of the 90-percentile noise level adds a measure of conservatism to the results, since it requires slightly higher siren levels before alert is predicted.
E-6
60 , , , , , , , , , , , , ,
.ONE MINUTE MINIMA B
3 50 -
.a
- w . n SIREN 23 .
- EJ
~w
. . 9dB E cn .
m5 ... . H z 40 - - - * - * - - - . .-------------- 3 . u THRESHOLD zO
-z- . .
23 . wo . zm *
- 00. .
M e
$30 m.
1 20 ''''' ' ' ' O 10 20 30 40 TIME (minutes) I I l l l FIG. E-2. MINIMUM BACKGROUND NOISE LEVELS OBSERVED IN ONE-MINUTE INTERVALS FOR A 40-MINUTE TIME PERIOD. i E-7
percent of the minima (14 out of 40) fall below the threshold line. Therefore, this siren level in this background noise has a 35 percent chance of alert -- when sounded for a duration of one minute.
~
v This plot applies only to sirens sounded for one minute, since the background-noise minima are one-minute minima. Stated another way, when a siren is sounded for one minute, it has an equal chance of encountering any of these forty one-minute time periods, which represent all one-minute periods. During 35 percent of these minutes it will alert the listener, since the noise falls below the alerting threshold at least once during those minutes. Next, say that the siren is sounded for four minutes. Figure E-3 shows the four-minute minima of interest -- as , circled dots. Each of these is just the lowest of four one- - minute minima in each four-minute grouping. Of these four-minute minima, 60 percent (6 out of 10) fall below the. threshold line. Therefore, this siren level in this background noise has a 60 percent chance of alert when sounded for a duration of four minutes. Note that the chance of alert has increased with the siren duration. Needed is mathematics that relates the one-minute chance of clert to the four-minute chance, and to the chances for all other siren durations as well. This mathematics is based upon proba-bilities P, rather than upon " chances." A 35 percent chance of clert is equivalent to a probability P of 0.35. Moreover, this mathematics is based upon the probability of failure to alert, rather than success in alerting. E-8 l
t N I I e i
@ FOUR-MINUTE MINIMA 8o -
< w 50 E> . .
a SIREN sm . . . 9 dB Wm .. . . .
.a- ,---@
$ - -@- * - @ ~ - - @.
z@40 o .
THRESHOLD Ez '
d3 2 0 . a e xa: . o 30 - cm 4 4 4 4 4 4 4 4 4 , 4 . 20 ' ' ' ' O 10 20 30 40 TIME (minutes) FIG. E-3. MINIMUM NOISE LEVELS OBSERVED IN FOUR-MINUTE INTERVALS POR A 40-MINUTE TIME PERIOD (from Fig. E-2). E-9
) ,
Chance of Probability Success of Success of Failure I 100% 1.0 0 80% 0.8 0.2 60% 0.6 0.4 40% 0.4 0.6 [ 20% 0.2 0.8 0% 0 1.0 , i __ i Note that failure " I ~ success and that failure occurs when minima points are above the threshold line. E.2.2 Underlying Mathematics and its Simplification Figure E-2 above contains one-minute minima for a total time
- period of forty minutes. All the points in this figure are col-lapsed onto the vertical axis in Figure E-4, at the left. They
! form a " cloud" of points denser at intermediate noise levels and sparser for higher and lower levels. This is a probability l 1
" cloud," in which area is proportional to the probability
- (density) of one-minute minima.
For any one-minute period, the probability of failure is proportional to the " cloud" area above the threshold line. This upper are'a, divided by the total cloud area, is the probability that the background noise will exceed the threshold level throughout any one-minute period -- that is, the probability that E-10
T e I i 60 i i i I 50 - E / f' f:4 3 5 " SIREN
< (f E
5 9dB 6 l 1. -1 1. , 5' - - - - - - - y m 40 THRESHOLD H I 3 l E 2 )I tb - 2 0 30 -- h 20 ; l j ; TIME (minutes) FIG. E-5. PROBABILITY " CLOUDS" POR ONE-MINUTE BACKGROUND NOISE MINIMA IN SUCCESSIVE MINUTES, ASSUMING MINIMA IN SUCCESSIVE MINUTES ARE NOT INDEPENDENT. E-13
T two, given that failure occurred during minute one." In other I words, the cloud at minute two represents the probability that the second minute's minimum will be above the threshold, given that the first minute's was also above the threshold. . Mathematically, we write P 2:1 for this conditional probability. Then P(4) = (Py)(P3:1:2)(P 4 1,2,3) conditional probabilities Note that P 2:1 is greater than the independent P2-P2:1 > P1 This increase is due to the regularity between successive minutes -- technically to the correlation between the successive minute's minima. The higher the correlation between successive minima, the more this probability cicud will condense above the threshold line. The remaining clouds condense even more above the line, since they are failure probabilities, given that several failures _ have preceded. A short numerical example will be useful here. For no correlation, we have P(4) = (0.65)(0.65)(0.65)(0.65) P(4) = (0.65)4 = 0.18 end therefore the probability of success is 0.82. For some correlation, we have P(4) = (0.65)(0.8)(0.85)(0.9) P(4) = 0.40 E-14
1
~
for a probability of success of 0.60. And for fu'll correlation we have P(4) = ( 0. 6 5 ) ( 1. 0 ) ( 1. 0 ) ( 1. 0 ) P(4) = 0.65 for a probability of success of 0.35. In general, P(n) = (Py)(P2:1)(P3:1,2 ) . . . ( Pn : 1,2,3, . . . , n-1 )
=
(Py )" for no correlation (E-1)
=
P1 for full correlation. The upper half of Figure E-6 illustrates graphically how the probability of failure thus decreases with increasing time -- that is, with increasing siren duration. The probability of suc-cess therefore increases with siren duration, as shown in the bottom half of the figure. (This figure is an example only, not a general result. ) Note for large correlation between successive minima, there is not as much benefit in sounding the siren longer. If the siren fails to alert during the first minute, it will most likely fail to alert thereaf ter, because the first minute is nearly identical to all subsequent minutes. This underlying mathematics resides in Eq. E-1 above. In Eq. E-1, the notation P n:1,2,3, . . . ,n-1 reminds us that P n is a conditional probability, which assumes the siren failed during E-15 i
4 r
/
t 1.0 , i i i i i i i
~
K v j- - FULL CORRELATION - 0.5 -
* %,4 -
B - TIoy -
! +o '---------._
~
{l - 0 ' ' ' ' ' ' "~~ 1.0 e i e i i i 4-g
- p. .
~
- a. '
COB OO h *B 3 . S o 0.5 -
,[, -
.N 5 //
r FULL CORRELATION
~
.8 g -
n. 0 O 1 2 3 4 5 6 7 8 9 SIREN DURATION (minutes) FIG. E-6. GRAPHIC ILLUSTRATION OF SIREN ALERTING PROBABILITIES VS. SIREN DURATION, FOR VARIOUS AMOUNTS OF CORRELATION BETWEEN BACKGROUND NOISE MINIMA IN SUCCESSIVE MINUTES (Example only). E-16
T all previous minutes. We next simplify, so that this Pn assumes failure only during the immediately preceding minute. Mathemati-cally, Pn :1,2,3, . . . ,n-1 = P n: n-1 Le t Pn:n-1 = CP1 where C contains all the conditional aspects of the probability. The ' term Py is the unconditional probability for the first minute. Then P(n) = (P1)(CP1)(CP1) ...(CP1) P(n) = P1" Cn-1 (E-2) Note that for no correlation, C=1 (E-3) and there fore - P(n) =Pn i as before. For full correlation,- 1 C= (E-4) i l to make P(n) = P" ( )n-1
= Py E-17
. (
as before. Eq. E-2 is the desired simplification. In the following section, we graph measured background data, to explore the nature of C, for correlations typically present in measured background
, hoise data.
E.2.3 . Exploratory Graphs, Guided by the Mathematics To explore for C graphically, we first take the logarithm of Eq. E-2. P(n) = P in Cn-1 log P(n) = nlogPy+ (n-1) log C log P(n) = -logC + n log CPy_ (E-5) If log P(n) is then plotted against n, the resulting , straight line should have a vertical intercept of -logC and a -- slope of log CP1 Af ter some curve-smoothing on linear paper, on Fig. E-7 we logarithmically plot part of the data in Table E.1 above. Each line is for a different representative siren level, labelled @ through @. l l Of course, the linear curve-smoothing helped line up the points shown here. Even so, the regression fit to straight lines for each siren level is very good. Note however, that the vertical intercepts and the slopes vary from curve to curve. Therefore, C must vary with siren level. We then set each intercept equal to -logC and each slope equal to logCP1, and solve for C and Pi -- separately for each straight line. E-18
, w
T l I i 1.0 ( y. i g SIREN i 1 5 i
._ 3 0.5 \
LEVEL
. . ~
t s n-
.\. _
4 1 g g g .
- f 50 .1 - * * .
_C. g _
~
~
E _ m . o - [ 0.05 -
@.@FOR OlFFERENT ~
SIREN LEVELS i
~
HIGH - SIREN LEVEL g l r 1 Gt 0.01 ' 0 10 20 30 # SIREN DURATION, n (minutes) FIG. E-7 LOGARITHM OF ' life PROBABILITY OF SIREN FAILURE-TO-ALERT VS. SIREN DURATION FOR FIVE DIFFERENT SIREN LEVELS, 4
, DERIVED FROM THE DATA IN TABLE E.1.
E-19
T I Line Number C Pi (h 1.073 0.925
, h 1.426 0.678 h 1.816 0.520
([) 3.062 0.293 (h 4.064 0.199 From Eq. E-4 above, we suspect that C may be a power func-tion of P 1, and so we plot logC against logP1 in Figure E-8. On this plot, the straight-line fit is also very good. It yields: C= (P1)-0.87 It seems to make sense, based upon this limited analysis, to generalize to _
~#
C = (Py) where p (rho) denotes a correlation coefficient. Zero correlation would then make C= (Py) = 1 cnd full correlation would make
~
C= (Py) = p These agree with Eqs. E-3 and E-4 above. I E-20 '
l l \ l 10 , , , , , , , , 5 - o
? - e .
N., 1.0 - ' ' ' l ' ' ' '*' O.1 0.5 1.0 log P, FIG. E-8. PLOT OF VALUES OF CUMULATIVE CONDITIOMAL PROBABILITIES (IDG C) VS. PROBABILITY OF FAILURE IN 'IBE FIRST MINUTE (LOG P 1), DERIVED PROM FIG. E-7. E-21
T s J In summary then, the time-pattern within Tab'le E.1 can be written as i P(n) = (P1)n-0.87(n-1) , (p )0.87 + 0.13n (E-6) The two constants in the exponent sum to 1.00, and depend c, upon correlation within the background noise, from minute to minute. Moreover,'Eq. ~E-6 depends upon the siren level through P1, which varies with siren level. Next, we simplify Eq. E-6 so it may be generalized to a wide variety of noise-level tables, not just Table E.1 above. Eq. E-6 is valid for all siren levels, in the presence of ' the -particular background noise used to develop Table E.1. Its general form is P(n) = (P y)" C"~
, p )n gp y -p(n-1)
, gp )p +n (1 - p) (E-7)
In logarithmic form, log P(n ) = p + n(1-p) log P y i
=p log Py+n (1-p) log P y (E-8)
With logP(n) plotted against n, this is the equation of a straight line with vertical intercept p log P y and slope (1-p ) log P y. E-22
5 e A normal regression fit would solve for the two v'ariables o and Py, separately for each of the siren levels (as shown in Figure E-7, for instance). However, there is a _ relationship above that implies o to be a' constant, independent of the siren level. Therefore, we wish to . collapse all curves, for all siren. levels, to a single curve. For this purpose, we ma'nipulate Eq. E-8 as follows: I log P(n) = p + n(1-0) log P y log P(n) , , y, ) 19# 1 = 1 + (n-1)(1-p) (E-9) Hence, plotting (log P(n)/logPy) against (n-1) yields a straight line of intercept 1 and slope (1-p), independent of siren level. In other words, each curve in Figure E-7 has been normalized to its value of Py , and all curves have been col-lapsed into one. We will have need below for a similar equation, but norma-lized to the probability at four minutes, rather than at one minute. We develop this next. In the graphs above, letter n was interpreted as progressing
- in one-minute steps (n=1,2,3 equals t=1,2,3). However, nothing in the mathematics requires this interpretation. Any time inter-val could be taken as the basic interval n above. In particular, the basic time interval could be taken as four minutes. Then four-minute minima (n=1) would combine into eight-minute minima (n=2), and so forth. The result would be Eq. E-9 above, but with n = 4t-(in minutes) and Py = "
P(n=1) (t = 4 minutes) l E-23
..,= -,
y
/ >[t t .
< r J , i .i
.a i 'e' f
.; i , (
, ~
Figure E-9 schematically co,mpares these one-minute and four- , minute normalizations[el' ort the-one-minute , , ,, normalization on n=t, and there forei n--1,=f t-1,, top:
, as st}own ori the' first hori-zontal axis. Plotted horfzont; ally is the range i ,
e i r ;
.a > 1 , .
g, 0 i t.- 113 -
, f c/
\ r 1<t < 4 / ,';
i, Wi i The small plotted points represent the tabulat,ed halues for these t7 - / four minutes, collapsed into one,line by the Py< normalization. A ~ The line is, fit by linear regression '[a'nd, has sl',o'pe t ,'(1-p). ' i ~ t This upper portion of Figure E-9. is' for rotati.ng sirens. As . . , - explained in the main text, rota <ting sirens areole'ns ; ,a . ef f ective in their.maximu;m siren level clerting the public, sin'ce they produce for only a portion of their duration... For this reason,,four-minute rgnults for rotating sirens are' derived from the one- - minute background-noise statistics. In the figure, the third horizontalsda,l'eshowsthecorresponding siren durations for rotating ~ sirens.., The normalization is therefore to a four-minu te
'l siren duration, and the graph extends up to a maximin of 16
, t ', ..
t ' i minutes. . iie
; - , . r ,< ! / t'
/ ** g
-.; 't e
/ , ,
..,/
Js ,
- Note that ,the lines in 4 Figure E-9 rise rattier, than fall to the right, as doesJFig'ure E-7, for this reason: '
In Figure E-7, the actual logarithms on the vertical ' axis are. n,egative, since the R(n)'s are less'than unity. Therefore, this" vertical axis actu-ally decres.Ses, ,from zero at the top to .minus- tw6 at the bottom. For' increasing'n, then, the curves take on increasingly large negative values (for example: -1, -1.5, -2). Figure E-9 is normalized by'logP1, however, which is/also negative, and which turns these increasi'ngly negative values into increasingly posi-tive' values'. Th'e re fd re , the lines rise in Figure E-9. E-24 J
log P(n)/ log P, JL
! ONE-MINUTE NORMALIZATION j., n=t
- .s
)..
j 3
$. ~4
.f : .
I O l l l > n - 1 (= t -1) 0 1 2 3 l l l l 5t 1 2 3 4 l l l ROTATING SIREN 4 8 12 16 DUR ATION (min) log P(n)/ log P, 7
.JL . s
; e ". FOUR MINUTE
. i ! 't NORMALIZATION
. .i ;*I I a
*a n = V4 i t. -
sC :
- 6. I; E N . '. '. i
...l:
2 ' ;9 t l \ ' I I
$. *t f.: . . ; l . ; . l . , , l r n - 1 (= t/4 -1) 0 1 2 3
! : : : : : : . : : : : : : : : >t 1 2 3 4 5 6 7 8 910111213141516 y STATIONARY SIREN 1 2 3 4 5 6 7 8 910111213141516 DUR ATION (min)
FIG. .E-9. SKETCH OF ONE-MINUTE AND FOUR-MINUTE NORMALIZATIONS TO SHOW RELATIONSHIP BE'IWEEN VARIABLE N AND SIREN DORATIONS. E-25
l T r
?
For the four-minute normalization at the bottom of the figure: n = t/4, and therefore n-1 = t/4 - 1, as shown. Plotted horizontally is the range 1 - 133 S SA 1<t < 16
-The second horizontal scale shows time t and is identical to the third scale, which shows duration of stationary sirens. The normalization is therefore to a four-minute siren duration, and the graph extends up to a maximum duration of 16 minutes.
Using these equations and normalizations, the curve-fitting procedure was applied to six background-noise tables -- tables similar to Table E.1 above -- developed from data measured at 74 _ different indoor and outdoor locations. In this curve-fitting, no linear smoothing was used, and data from all siren levels were used without omission. Table E.2 contains the resulting slopes. These slopes were next converted to o, assuming that they equal (1-p), as labelled in the table. The resulting twelve values-of p were plotted against the corresponding values Rxx of the auto
- y. correlation function, to obtain R = -0.034 + 1.051p xx
= p This regression equation has a correlation coefficient (between values of p and rxx) f 0.85, which is satisfactorily high.
E-26 i
s In the next section, we collect these results into a form of use to the reader. E.3 Summary of Results Figure E-10 contains the results of the analysis above. This figure is used as follows: Convert the four-minute " chance' of alert" to a probability of failure-to-alert": P= 1- (Chance of alert)/100 Raise this value to the exponent determined from Figure E-10, for'the particular siren duration of interest. P= (P4-min) Exponent (E-10) Convert this " probability of failure-to-alert" back to a
" chance of alert":
Chance of alert = 100 (1-P) TABLE E.2. SLOPES RESULTING FROM SIREN LEVEL DATA. Listener Subclass Resulting Slopes (1-o) Location Stationary Rotating Sirens Sirens Indoors Scenario 1 0.217 0.142 Scenario 3 0.274 0.254
. . . . . . . . . . . - - . . . . - . - - l; Outdoors Rural, day 0.164 0.11/ l Urban, day 0.065 0.103
' Rural, eve / night ,' O.150 0.075 Urban, eve / night 0.046 0.039 !
._a E-27
1 i e ROTATING SIREN 2 ' i ' i ' i ' i ' ' i ' i _ EXPONENT l
' i
- = 1 + (d/4 -1) (SLOPE)
- ()- SLOPES:
~
e
* ' O.254
- 0.177 z
zj
]@ _- 0.M2
- 0.103
@ SCENARIO 3 2 -
@ RUR AL, DAY 0.075 X : . 0.039
- @ SCENARIO 1 -
-- @ URBAN, DAY ,
@ RUR AL, EVE / NIGHT -
- @ URBAN, EVE / NIGHT I I I 'e f I t e f f I e n e n .
4 6 8 10 12 14 16 0 2 DURATION OF ROTATING SIREN, d (min) STATIONARY SIREN 2 ' i ' i ' i > i a ' i ' s - EXPONENT
.@; = 1 + (d/4 -1) (SLOPE)
- ( .
SLOPES:
- @l 0.274
-@ - 0.217 z -
~
@' O.164
$3
@ SCENARIO 3 0.150
- v. 2 -7 -
@ SCENARIO 1 0.065 X
@ RURAL, DAY 0.046
- @ RURAL, EVE / NIGHT [
@ URBAN, DAY -
f -
- @ URBAN, EVE / NIGHT O 12 14 16 O 2 4 6 8 10 DURATION OF STATIONARY SIREN, d (min)
PIO. E-10. GRAPH OF EXPONENT FOR USE IN EQ. E-10. i E-28 .
I APPEND'" F. RANDOM SELECTION OF POPULATION-WEIGHTED LISTENING POINTS AT THE THREE MILE ISLAND SITE The objective of the listener-site-selection process was to identify 50 randomly selected residential locations within the 10- mile EPZ surrounding the TMI Nuclear Plant. No arbitrary decision was made as to how many of the points would lie in urban or rural areas or within certain distances of the plant. The various steps used in the site selection procedure are described below:
- 1. A population-distribution map (see Fig. F-1), con-sisting of a 10-mile-radius circle divided into annular sectors defined by interior circles and radii, was superimposed on the U.S.G.S. maps.
Population distribution information consisted of the number of people within each annular sector. These data were used in order to population-weight the random selection process described below.
- 2. Each annular _ sector was first assigned a designa-tor, such as a letter. A range of numbers was then assigned to each sector according to the population in that sector. For example, Sector A, just north j
of the site, has a population of 19 and thus was assigned numbers 1 through 19. Sector B (moving clockwise) has a population of 55 and was assigned numbers from 20 to 74. Sector C has a population of 42 and was assigned numbers 75 through 116. This process was continued until each number between 1 and 166,295 (the total estima ted popula-tion) was assigned to a particular sector. A ran-l F-1
T dom number generator (available on a Texas Instru-ments Model TI-59 hand calculator, for example) was then used to select 50 numbers at random between 1 and 166,295. Each number selected represented one site (to be chosen later) within the sector con-taining that number. Thus, sectors with larger populations had a greater possibility of including chosen listener sites.
- 3. Having determined the sector locations for each listener site, the next step in the procedure in-volved selecting the actual location of each site within the respective sector. This was accom-plished by first overlaying a rectangular coordi-nate grid on each sector of interest on the topo-
~g raphic map. The grid was composed of boxes with dimensions of approximately 1000 feet square, and each box was assigned an X and a Y coordinate ~
according to its location on the grid. The grid was positioned such that the X-axis was oriented in the east-west direction and the Y-axis was oriented in t'he north-south direction, and such that all parts of the sector of interest were covered by a positive (X,Y) coordinate pair box. A random num-ber generator was then used to select random pairs '- of numbers within the X and Y ranges covering the sector of interest. Each X,Y pair was used to lo-cate a particular 1000 feet squa're box.on the map. If there were no residences inside the square or if the square fell outside of the sector of interest, that coordinate pair was disregarded and another pair was chosen at random. This process was con-tinued until a square area including one or more residential structures was found in the sector of F-2
T interest. The listener site was then chosen to be any residence within the randomly selected square a re a. ' i 'For urban sites in the pink " building-extension" area of the topographic map a residential building was always assumed to exist, and was selected at the center of the pink area in the 1000 feet square . box. The above procedure resulted in a random sample of 50 listener locations, distributed throughout the EPZ as shown roughly on Fig. F-1. 1 m. O F-3
T Il I
~
g o.t l *
\.. o a
* .k? \
e = / 3, e
- I' b e e .t,&
G
,. .. t s -
&\$ q 1
/6,, / ,,9
- m. Nm.*
n"*' 9; h/ f f~
l,,f&\h[l IS2 e
e-1* / rw ,
- '}
n e1 55 571 331 36 353 fg' - JD 35 137 % ;- 10:31 e E s A** '
'h 00 'd x
' ,0 o %,
'\ *$ st ' s%' -
0)
'/<
'3
'6 ge #MC $' W N, j'
$O t'- / ,{ .=
~
% 4,;
j, 44 0 g c y* Y#* 8 e e DISTANCEj TOTAL I CUV. 101st 0-1 El. ~ G56 8 POPULATION DISTRIBUTION 1-2 Kl. 2',017 2,675 0 TO 10 MILES - 1980 2-3 Mt.I 7,579 10,254 TilREE MILE ISLAND NUCLEAR ST ATION UNIT 2 3-4 4-5 MI. Mt. 0,G70 8,801 10,930 28,821 gy.: - - fyr g* g 5-10 Mt. 137,474 100,295 .?t.'
" EICllisl0N RA01US FIG. F-1. RANDOMLY SELECTED LISTENER SITES AT 'IEREE MILE ISLAND.
F-4
T s APPENDIX G: TEST CASES (SAMPLE SCENARIOS) FOR THE
'IBREE MILE ISLAND SITE
- 1. Warm Summer Weekday Afternoon: Weather clear to partly c1,oudy.
People: 30% indoors, at work 40% indoors, at home 20% outdoors 6% in motor vehicles (windows open) 4% asleep Buildings: Windows open (homes) Windows closed (workplace) Wind: (100 ft) 5 mph from East Temperature Gradient: -1.0 'F/100 ft., Pasquill stability Class A Relative Humidity: 65%
'2. Summer Weekday Night: Weather clear to partly cloudy.
People: 95% indoors, sleeping 4% indoors, at work 1% in motor vehicles (windows closed) Buildings: Windows open (homes) Windows closed (workplace) Wind ^(100 ft ): Northwest, 5 mph Temperature Gradient: +0.5 F/100 ft. Stability Class E Relative ilumidity: 80% l 3. Winter Weekday During Evening Community Hours: Cold, overcast People: 70% indoors 25% in motor vehicles (windows closed) l 5% outdoors Buildings: Windows closed, storm windows closed G-1 L
I i {1
- 3. Continued ,
Wind (100 ft): Southeast at 3 mph Temperature Gradient: -0.5 F/100 ft.
Stability Class D Relative ilumidity: 70%
- 4. Winter Night During Snowfall.
People: 95% indoors, sleeping 5% indoors, at work B,silding: Windows closed, storm windows closed Wind (100 ft.): West at 15 mph Temperature Cradient: -0.5 F/100 ft Stability. Class D
.(
4 Relative Humidity: 90% - 1 , t. 4 I G-2
T e APPENDIX H: SIREN LOCATIONS FOR THE TMI EPZ This appendix provides siren information for the. TMI EPZ. Siren locations are indicated on Fig. H-1 ( se'e . foldou t ) . Table H.1 provides information on the type and rating for each siren. TABLE H.l. TMI SIREN INFORMATION. Rated SPL County / Siren Designation Type * (dB e 100 ft) Cumberland Cl R 124 Cumberland C2 S 122 Cumberland C3 S ~122 Dauphin D1 S .122 Dauphin D2 S 122 Dauphin D3 R 124 Dauphin D4 S 122 Dauphin D5 S 122 Dauphin D6 S 122 Dauphin D7 S 122 Dauphin D8 R 124 Dauphin' D9 S 122 Dauphin D10 S 122 Dauphin Dll S 122 Dauphin D12 S 122 Dauphin D13 S 122 Dauphin D14 S 122 Dauphin D15 S 122 Dauphin D16 R 124 Dauphin D17 R 124 Dauphin Dl8 S 122 Dauphin D19 S 122 Dauphin D20 S 122
- Rotating (R) or Stationary (S)
H-1
T i TABLE H.l. TMI SIREN INFORMATION (Cont.) Rated SPL County / Siren Designation Type * (dB @ 100 f t) Dauphin D22 E 124 Dauphin D23 S 122
'- Dauphin D24 R 124 Dauphin D25 R 124 124 Dauphin D26 R Dauphin D27 R 124 Dauphin D28 S 122 Dauphin D29 R 124 Dauphin D30 S 122 Lancaster LAl R 124 Lancaster LA2 R 124 Lancaster LA3 R 124 Lancaster LA4 R 124 R 124 _,'
Lancaster LAS Lancaster LA6 R 124 Lancaster LA7 S 122 Lancaster LA8 S 122 Lancaster LA9 R 124 Lancaster LA10 S 122 Lancaster LAll R 124 Lancaster LAl2 S 122
R 124 Lancaster LA13 Lancaster LA14 S 122 Lebanon LEl S 122 l
Lebanon LE2 S 122 York Y1 S 122 York Y2 R 124 i
- Rotating (R) or Stationary (S) 1 l
H-2 l
4 T TABLE H.1. 'IMI SIREN INFORMATION (Cont.) Rated SPL county / Siren Designation Type * (dB e 100 ft) York Y3 S 122 York Y4 S 122 York Y5 S 122 York Y6 R 124 York Y7 S 122 York Y8 R 124 York Y9 S 122 York Y10 S 122 York Yll R 124 York Y12 R 124 York Y13 S 122 York Y14 S 122 York Y15 R 124 York Y16 S 122 York Y17 S 122 York Yl8 S 122 York Y19 S 122 York Y20 S 122 York Y21 R 124 York Y22 S 122 York Y23 S 122 York Y24 R 124 York Y25 S 122 York Y26 S 122 York Y27 R 124 York Y28 S 122 York Y29 s 122 York Y30 R 124
- Rotating (R) or Stationary (S)
H-3
T s . 1 TABLE H.l. TMI SIREN INFORMATION (Cont.) Rated SPL County / Siren Designation Type * (dB e 100 ft) York Y31 R 124 York Y32 S 122 V. York Y33 S 122 1 6 l i i
- Rotating (R) or Stationary (S) -
H-4
- - - - . - -.,..-,..--.,,,---_,.--n_ ,,,,, n. - - .,----.-e-,-,..~----._,-- , . - , - , . - - - ----,-n,. -.. , ..._e-
1 1 i FIG. B-1. 'IMI SIREN LOCATION MAP. See foldout. H-- 5
1 APPENDIX I: ANALYSIS INPUT / OUTPUT DATA FOR 'niREE MILE ISLAND This appendix provides listings of computer file input and output data for the TMI analysis. Explanation of the terminology used for each listing is provided below. TABLE I.l. TMI-SIRENS This file contains input data for each of the TMI sirens as follows:
- Siren No, number assigned to each siren for'use by computer program Siren Name. first letter indicates whether the siren is rotating or stationary type (R or S); the remainder consists of the actual TMI siren designation, beginning with county letter abbreviation and ending with a number.
- x, y, z these are the physical coordinates for the siren location; the x-axis is oriented east-west, the y-axis is oriented north-south, and the z-axis is oriented vertically. The x and y coordinates are in units of km, referenced to the grid shown on the Feb. 1981 NRC Eme r-gency Planning Map for TMI (the plant center is located approximately at x = 353, y =
4446). The z coordinates are in units of feet. SPL@l00FT these numbers indicate the rated sound pres-sure level for each siren at a distance of 100 ft, in dB. 1 I-1 l
)
'\
( TABLE I.2. TMI-LISTENERS ' This fila contains input data for each of the randomly selected listener locations as follows:
. Site No. number assigned to each site for use .by computer program
. Site Name designator for listener site; the first letter indicates whether site is urban or rural (U or R).
+ .x, y, z these are the physical coordinates for the siren location; the x-axis is oriented east-wast, the y-axis is oriented north-south, and the z-axis is oriented vertically. The x and y coordinates are in units of km, referenced to the grid shown on the Feb. 19 81 NRC Eme r-gency Planning Map for TMI (the plant center is located approximately at x = 353, y = -
4446). The z coordinates are in units of feet.
- ODLR the outdoor median alertin.g level for a 4-min.
rotating siren (see Table 3. 4 and Fig . 3-3 of text). An entry is. given for each of the four scenarios. f- - OVCR the outdoor alert distribution for a 4-min. rotating siren (see Table 3.4 and Fig. 3-3 of text). An entry is given for each of the four scenarios.
- . OVLS the outdoor median alerting level for a 4-min.
stationary siren (see Table 3. 4 and Fig . 3-3 of text). An entry is given for each of the four scenarios. i s I-2
T 9
- OVCS The outdoor alert distribution for a 4-min.
stationary siren (see Table 3. 4 and Fig. 3-3 of text). An entry is given for each of the four scenarios. TABLE I.3. 'IMI-SCENARIO This file contains input for each of the four sample scenarios as follows:
. Scenario No. number assigned to each scenario (see App. G.)
. AMCL molecular absorption, in dB/1000 ft
. WIND wind direction in degrees (0* = wind from north, etc.)
. NRES residential building outdoor-to-indoor noise i
reduction, in dB
- .NCRM commercial building outdoor-to'-indoor noise reduction, in dB
- F1 - F10 activity fractions F1 fraction of people outdoors F2 fraction of people indoors, at home, listening to radio or TV F3 fraction of people indoo. , at home, sleeping F4 fraction of people indoors, at home, neither l sleeping nor listening to radio or TV F5 fraction of people indoors, at work, in com-mercial establishments F6 fraction of people indoors, at work, in industrial locations I-3
T T
.F7 fraction of people in vehicles in rural areas '
at 55 mph F8 fraction of people in vehicles in rural areas at 30 mph F9 fraction of people in vehicles in urban areas
; at 55 mph F10 fraction of people in vehicles in urban areas at 30 mph
. INP indoor alert probability curve ~(see Figs. 3-4 and 3-5 of text)
PU55 probability of alert for motorists in urban areas at 55 mph PU30 probability of alert for motorists in urban areas at 30 mph PR55 probability of alert for motorists in rural - areas at 55 mph . PR30 probability of alert for motorists in rural areas at 30 mph
. MUL vertical profile of wind speed, Sz, in ft/sec/in ft.
. ADD vertical profile of air temperature, a, in 4
'F/ln ft.
TABLE I.4. LISTENEROUTPUT This listing provides the number, name, and outdoor sound pressure level (LOUT, in dB) for the ". dominant" siren at each sample listener location, for each of the four sample scenarios. The results are listed in numerical order for scenarios one through four for each listener site. I-4
' \
T TABLE I.S. PROBS This listing provides the final results for the analysis. Information is listed in numerical order for scenarios one through four for each listener site. This information consists of alert probabilities P1 through P10 corresponding to activity fractions F1 through F10, as well as the total probability of alert (PT) for each sample scenario at each sample listener site. A summary is provided at the end of the listing showing the rural and urban populations followed'by the total rural probabil-ity of alert (PTRUR), the total urban probability of alert for the EPZ (PTALL). The total probability values are listed in numerical order for sample scenarios one through four. e I-5
T TABLE I.l. ( _TMI-SIRENS ' SIREN: SIREN NAME X Y Z SPL&100 FT 1 R C1 .... 341.,650 4454.200 360.000 124 2 S C2 % 338.950 4455.600 450.000 122 3 S C3- 340.550 4452.950 4 S D1 450.000 122 353.200 4446.600 350.000 122 5 S D2 355.950 6 4450.150 590.000 122 R D3 360.300 4451.300 7 S D4 510.000 124 8 363.300 4453.250 520.000 122 S DS 352.500 4450.500 9 S D6 370.000 122 354.000 4454.250 470.000 122 10 S D7 358.700 4454.400 640.000 122 11 R D8 362.550 4457.700 12 S D9 480.000 124 13 351.000 4451.650 390.000 122 S D10 347.700 4452.250 14 S D11' 360.000 122 346.150 4452.900 390.000 122 15 S D12 344.200 16 4455.550 550.000 122 S D13 342.750 4456.200 430.000 17 S D14 122 344.150 4457.800 425.000 122 18 S D15 19 342.400 4459.300 600.000 122 R D16 344.750 4460.050 550.000 20 R D17 124 346.900 4460.550 540.000 124 21 S D18 346.600 22 4458.150 470.000 122 S D19 346.100 4455.450 570.000 23 S D20 122 ~ 350.100 4454.300 490.000 122 24 R D22 352.350 4456.650 25 S D23 350.000 124 354.600 4458.350 450.000 122 26 R D24 351.150 4461.300 27 R D25 530.000 124 354.900 4461.000 510.000 124 28 R D26 357.700 4464.650 29 S D27 450.000 124 360.100 4461.100 430.000 122 30 S D28 358.750 4459.600 R D29 450.000 122
- 33. 358.600 4457.250 750.000 124 32 S D30 '
349.200 33 R LA1 4458.450 530.000 122 ' 355.500 4443.950 570.000 124 34 R LA2 358.650 35 4446.000 570.000 124 R LA3 362.300 4446.950 590.000. 36 R LA4 124 363.650 4449.250 490.000 124 37 R LA5 367.550 4450.750 38 R LA6 530.000 124 360.000 4441.800 510.000 124 39 S LA7 362.700 4444.350 40 S LR8 460.000 122 364.300 4445.650 530.000 122 41 R LA9 367.400 4446.800 590.000 42 S LR10 124 357.900 4439.050 350.000 122 43 R LR11 361.100 4438.000 450.000 124 44- S LR12 365.250 4441.150 450.000 05 R LR13 122 369.400 4442.850 450.000 124 46 S LA14 365.700 4437.200 47 450.000 122 S LE1 366.700 4455.100 560.000 48 ' S LE2 .122 t 363.150 4462.050 500.000 122~ . 09 S Y1 3;~ 2# 362.250 4434.150 50 730.000 122 R Y2 f c. " 358.800 4434.550 370.000 124 2-G
T TABLE I.l. (Cont.) 51 S Y3 355.900 4430.100 550.000 122 52 S Y4 ' 352.050 4430.100 53 470.000 122 S Y5 348.100 4430.250 530.000 122 54 R Y6 346.000 4429.600 490.000 55 124 S Y7 355.800 4436.250 520.000 122 56 R Y8 354.950 4433.000 57 690.000 124 S Y9 352.450 4431'.250 460.000 122 58 S Y10 350.150 4431.850 59 490.000 122 R Y11 346.000 4434.300 530.000 124 60 R Y12 341.950 4433.050 670.000 124 61 S.Y13 337.850 4434.700 570.000 122 62 S Y14 341.650 4436.700 600.000 63 122 R Y15 345.600 4439.200 630.000 124 64 S Y16 349.450 4435.500 390.000 65 122 S Y17 349.750 4438.600 470.000 122 66 S Y18 353.450 4440.700 510.000 67 122 S Y19 352.800 4437.100 500.000 122 68 S Y20 353.100 4434.400 530.000 122 69 R Y21 355.350 4439.500 330.000 124 70 S Y22 351.300 4442.250 520.000 122 71 S Y23 347.750 4443.500 670.000 122 72 R Y24 349.900 4447.200 490.000 124 73 S Y25 347.000 4449.350 770.000 122 74 S Y26 344.600 4445.850 75 510.000 122 R Y27 344.100 4451.150 370.000 124 76 S Y28 342.000 4450.350 77 920.000 122 S Y29 338.750 4451.950 570.000 122 78 R Y30 340.100 4447.000 620.000 124 79 R Y31 338.650 4444.750 540.000 124 80 S Y32. 339.500 4439.550 675.000 122 81 S Y33 342.050 4442.700
-----......n .-
530.000 122 I-7
T TABLE I.2. ( N ' M 1 - L I S T F. N t w $ il1L d alTL dAno X a X uund ueCn buuS 09ta 1 R1 3 S1.7 3il 9443.92e 490.000 51.0 o.2 50.0 4.9 b l. 2 t.2 50.0 4.0 31.0 o.E =0.0 4.9 31.0 c.2 $0.0 4.9 4 2 0 2 Jol.od1 44bJ.702 J25.t?d 5 4. 'J b.d 32.0 3.2 43.0 'J . 0 13.0 i.0 49 0 4.2 4d.0 3.8 4J.0 3.2 43.0 2.8 4 u J JaJ.0tT 4454.432 J 15 '. 2 0 :' 50.9 5.0 46.2 4.9 42.0 J.0 41.0 2.9 go.0 9.R 47.9 3.s
$2.4 J.C 11.P 2.8 4 0 9 J 3 3 . 2 9 .* 94dJ.an? 313.010 5 0. si 5.0 4d.0 4.9 42.0 J.2 41.0 2.8 nd.0' 9.0 97.2 3.9 42.3 J.0 41.0 2.9 3 0 3 3 0 J . 9 e .! 49,J.7e2 JJs.dds 50.0 5.2 46.2 4.0 52.0 J.C 11.2 2.9 4d.0 4.e 47.0 3.8
*2.0 3.0 41.0 29 0 is e JoJ.13C 4 4u 2. la.? Jo?.020 oJ.G c.2 cl.2 4.9 o3.3 c.0 c1.0 4.9 o3.d e.e c1.0 4.9 63.D b.e c1.0 4.9 7 u 7 Jas.dio 4431.201 J00.22d 54.0 6.8 c2.0 3.9 43.e 3.0 43.0 2.9 49.0 4.2 4d.0 3.8 SJ.0 J.E 4J.0 2.9 l o H o J45.730 4451.7i'2 4 2 5. e d.i 03.0 c.0 c1.0 4.9 bJ.a 6.2 c1.0 4.0 oJ.0 c.D c1.0 4.0
'oJ.0 6.0 c1.0 4.9 9 H 9 J3c.JJR 4449.709 525.220 $10 6.2 50.0 4.0 al.0 6.2 20.0 4.9 S1.0 6.2 50.0 4.9 51 0 6.0 50.0 4.8 le R lt Jad.JJO 44,7.100 J d a . 0 0'J bl.0 6.0 30.0 4.9
- 31. ? ' 6.0 50.0 4.9 d1 0 e.0 69.9 4.8 51 0 6.2 39.9 4.9 I-f
T TABLE I.2. (Cont.) ! 11 x 11 457.73P 4442.542 405.200 39.0 5.2 30.0 3.9 39.J 5.0 3d.0 3.9' d6.0 5.2 35.0 3.9 97.2 d.0 4b.0 4.9 12 R 12 335.o30 4441.dda 3a6.003 51.H c.c 50.0 4.9
~ 51. 2 c.2 52.0 4.9 31.0 c.0 30.0 4.9 51.0 6.0 53.0 4.9 13 x 13 443.o50 4442.4e2 4dd.222 av. 0 S.2 3d.0 3.9 Jv.2 D.0 ad.0 3.0 30.2 5.0 45.0 3.0
.47.d 6.2 46.3 3.0 14 M 14 JaJ.,Je naas.JCJ Jo0.23? 30.0 c.E c4.2 4.9 do.J c.2 54.0 4.9 50 0 e. 2 54.B 4.9 ob.0 c.0 s4.0 4.9 13 R 15 392.130 44oJ.0Hs 503.0dd 49.0 6.0 3d.0 3.9 39.0 5.2 3d.0 3.9 ab.C 3.0 Jo.2 3.9 47.J 5.0 40.0' 3.9 10 H 10 J32.dde 4439.7b> 40b.222 39 0 5.0 3d.0 3.9 av.0 c.0 4o.0 3.9 Jo.3 5.0 35.0 3.9 47.3 5.0 40.0 39
- 17 H 17 JaJ.oaa 4400.la2 423.000 av.D d.0 ad.0 3.9 39.0 5.0 Jo.0 3.9 36.0 S.0 35.0 3.9 47.0 d.0 46.0 3.9 la d'io Jaj.350 44od.202 440.ee0 50.2 5.c 46.9 4 '. 9 42 0 3.e 41.0 2.9 l 4d.0 4.e 47.0 3.9 52.0 J.2 41.0 2.9 19 U 19 3o4.350 4444.7ae 465. 000 53.0 5.0 4d.0 4.9 42.0 3.0 41.0 2.9 4d.C 4.0 47.6 3.9 l
42.0 3.0 41.0 2.9 20 H 2u Joe.40H 4444.900 490.222 39 0 5.E 3d.0 3.9 J3.0 b.0 Jd.D J.9 30.8 5.9 36.9 3.8 47.0 5.2 46.0 3.0 2-9
i TABLE I.2. (Cont.) l 21 x 21 J3d.730 1433.400 4v0.EJJ 31.0 c.0 30.0 4.E 51.0 6.2 $0.0 4.0 31 0 0.0 $H.P 4.0 15 1 . 0 t.2 52.e 4.0 24 d 24 Ja7.d50 44J2.400 J$3.223 39.0 5.2 3d.0 J.0 39.0 5.0 3d.P 3.0 J0.0 5.2 26.0 3.0 57.0 a.C 90.0 J.0 2J R 2a J34.131 4437.ea! 4de.e0J dl.J e.0 5J.P 4.0 31 9 0.0 $ d . ll 4.2 a l . .I c.0 bO.P 4.0
- 51. .' e.3 5J.e 4.0 21 0 29 J34 100 4433.co9 410.222 S d. d b.0 1o.0 4.0 12.0 J.2 41.C e.0 48.2 '4 . 0 47.0 3.0 42.3 3.0 41.0 2.0 2d H 2d sa2.dJ3 44J3.Joi soa.20J 31.3 c.2 s0.E 1.0 51.0 t.0 50.0 4.0 b l. 3 c.e 50.0 4.0 bl.J 6.E bd'. e 4.0 20 H 20 3 7 3. J Jt; 49'J2.co' 3 6.' . 2 d 3 bl.J 6.0 bW.e 4.0 51.d c.0 50.0 4.0 31.3 c.2 30.0 4.0 51.0 6.0 50.e 4.9 27 x d7 a24.d10 44Ja.edc J6e.fte b l. 0 c.e 50.0 4.0 31.0 c.e 30.0 4.0 51.0 c.0 50.0 4.0 31 0 e.e 5d.0 4.0 24 k 2o asi.JJO 11J7.333 JJ3.EJP 10.0 c.C 47.0 4.0 40.P c.0 47.0 4.0 4d.0 c.0 47.2 4.0 48.2 t.E 47.0 4.0 2s n 27 J45.1JO 4445.J3P 675.000 39.n 5.E 38.0 3.0 39.0 5.0 3d.0 3.0 36.0 5.2 35.0 3.8 17.0 d.0 40.2 J.0 30 R 3b 34d.1a9 44J4.5de 470.000 39.d 5.0 38.0 3.0 39.0 5.0 38.0 3.0 J6.0 d.0 Js.O J.B 47.0 5.0 46.0 3.0 T-10
T t TABLE I.2. (Cont.) 31 M 31 J44.700 4442.232 40s.22d 51 0 6.2 50.9 4.9 31.0 6.2 30.0 4.0 bl.2 c.0 50.0 4.0 61 0 c.0 50.0 4.0 32 R 34 JJd.a33 141o.s02 a23.020 49.0 3.2 Jd.B 3.6 39.0 5.0 38.0 3.9 30.a S .' E 35.0 3.8 47.-0 5.e 40.9 3.9 JJ K JJ J4J.550 4449.d22 1 000.2'.4 01.3 0.0 59.0 4.0 bl.a o.B $9.0 4.9 el.e c.2 59.9 4.9 c1.0 e.0 sv.0 4.9 34 R 44 344.010 4432.230 o25.EJO 39.0 'b.0 3d.0 3.9 49 0 5.0 Jd.2 3.9 J6.0 3.2 Ja.0 J.9 47.0 5.0 40.0 3.9 35 H 3b J39.336 4 4 4 9. c 'J 2 oDb.22d 61 0 62 ed.0 4.9 31.0 c.0 30.0 9.9 bl.a c.0 30.0 4.9 bl.d c.2 63.0 4.9 JO R Jo 440.7a3 4ta2. lad 120.00J av.J c.0 07.0 4.9 59.2 c.2 57.0 4.9 6 9. t o.0 57.0 4.9 59.1 c.2 b7.0 4.9 37 o J7 4 41. 9 J'! 4433.ws-J J25.021 54.J b.c d2.0 3.9 43.0 3.2 43.0 2.8 i 49.0 4.2 4d.0 3.9 13.0 J.2 43.0 2.9 de u Jo 342.14P 44o4.230 315.029 b4.3 5.2 52.0 3.0 44.0 J.2 43.0 2.9 49.0 4.8 4d.0 3.0 43.0 3.0 43.0 2.9 39 U JS J30.dd9 44as.3a2 J7D.220 30.0 c.0 4d.0 9.0 42.0 3.0 41.0 2.9 4d.d 4.0 47.0 3.9 42.0 3.2 41.0 2.9 49 u 40 J47.d40 4452.320 JJP.2JO 50.0 5.0 40.0 4.9 42.0 3.2 41.0 2.9 4:1. 0 4.9 47.0 3.g 12.0 J.B 11.0 2.8 I-11
r T 9 TABLE I.2. (Cont.) ( 41 0 41' 34d.18e 44s2.bd7 Jod.00J $ 4. H 5.0 52.0 3.W 4J.8 3.8 43.0 2.8 19 0 1.0 sd.0 4.0 43.3 3.0 4J.0 2.0 42 H 42 J40 032 4434.950 465 20d 51 3 e. 0 50.0 4.9 31.0 c.c de.e. 4.2
,, d1.0 c.c 50.0 4.0 bl.H c.0 dd.0 4.0 4J U 4J Jia.J01 543a.o00 aca.223 30.0 s.0 4a.g 4.g 42.0 3.2 41.0 2.0
- d. 0 42 47.0 3.0 42.0 J.E 41.0 2.0
** O 19 4 4 2. 0 J :' 4sso.7de 305.026' 50.0 5.0 41.0 4.J 42 0 32 41.0 2.0 4 d. 0 4.2 47.g 3.9 42.0 J.2 11.0 2.0 43 U 45 34J.2J2 4451.co" so 0. 0 eJ S0.0 d.0 4d.0 4.9 42.0 J.2 41.0 2.9 nd.J 4.c 47.0 3.0 42.0 J.0 '41.0 2 . 0 -- ;
4o a 40 J4o.doa 34n2.930 4J3.2ed eJ.d 68 el.a 4.g UJ.D b.2 b1.0 4.0 03.3 c.0 cl.0 4.0 03.J 00 c1.0 4.9 47 d 97 49o.230 $$a4.,aC sda.203 al.c c.E 0.2 4.0 51.0 6.0 50.2 4.8 51.0 c.0 50.0 4.9 d1 0 6.2 50.0 4.9 n to J43 33p [40 4453.EJP dbb.E00 31.7 c.0 50.g 4.g 51 0 c.0 50.0 4.s b1.9 02 59.0 4.9 31.0 c.e 30.0 4.9 G9 0 49 34b.354 4437.900 Jue.edd 50.0 b.0 46.0 4.9 42.0 J.E 41.0 2.9 go.0 s.D 57.0 J.9 41.0 'J . 0 41.0 2.9 0 U DJ Jib.900 440J.J32 323 220 60.0 dg 4d.g 4.g 12.0 J.0 41.0 2.9 48.9 4.8 47.9 3.9 i 42.0 3.9 41.9 2.0 - t 1 I-12 . l l l l
r , TABLE I.3. . 4
~
lf TM A- SC E N A.<10' a - S C t.h a . A nd t. alnu a .< r. :2 a u.s a , e l r2 't J rs e s' to r7 rd ev rif l' e.co 90 l o'. J1. . .223 .230 .040 . 220 .2J0 .278 . 22d .200 .020 . Pet 2 u.79 315 10.'31...d01 . dad .vac .230 .032 .019 . 223 .20E .BEP .297 4 d.33 1Jo al.'J1. . dad .143 .00~>.300 .00J .000 .270 .900 .CEO .let 4 0.04 27J 31. 31. .J22 .303 .352 .000 .e43 .010 . 003 .00E .020 .90k 1 ihr raan teu a edoa _ le'n J 1 n o t. A u s, , 1 1. DH '1 1.3le 1.J3J .2 . 2.1 d 1.o72 9.Sid 1' 1. d D J 1. e d 2. 140d 1 240 1.d7e .f . 4 0 2
.a 1. ile ? 1. p M 1.3.14 1.004 1.12; D. sos 1 - 1.JHJ l '. J J a 1.33J 1.04'1 a . o JP - 0.40's
/
e ! , , a 1 f Y r p 8 l ' n- ; j , , n y '
'# : /
j , 4- , f
,/ r
/
, f ;
> [_ : .- _ /
/ e .
s.'
'j . j' + l t IlJ
'I 9 +
gg g 5 g J, c. ~ l", ,
, f; .
~
',- s j
/ ,! '
s /
. 1 / ,. /
_ _ r-
's b
s N
- g 7
f '
, ll y s
. 1, j':,] . "'
,, , .e*
sov - I-13 ,
-f j
, ,, et
,,t lp.
, ,r -
, i - , , ,
#(, * ~
,r ,
-.{ : 2
T TABLE I.4. LISTENEnudTeu! lis d 11stener name stren g stren name tout 1 91 3J R LA1 70.9 4 S 91 71.o 7J S r22 bv.7 7? S r22 by.J 2 u 2 e S u6 104.0 o a va 104 1 e 3'05 104.J o S un 104 2 J 0 J d S uu ol.2 o S 95 vo.J o S un do.7 c S ud 96 6 4 4 4 o a u5 72.7 c $ 06 92.s o S ud oJ.5 o a ou v4 4 3 U 3 s S u2 7b.o o S un 90.P 5 S u2 7d.b ( c S us 9J.4 -- b 6o 3 $ ve 73.6
> S ue 76.o o S Ob 63 6 c S b3 dJ.1 1 o7 3 a L2 72.5 l o S US d4 1 3 S v2 13.t o .a ua o1.d j u R d o S'u2 03.3 t
e S uS 71 9 3 a ul .:.3 o S ub i 3 l 9 H 9 5 S 02 79 9 b a u2 93 1 5 5 ol 85.5 b S U2 95 4 10 x 10 J1 M LAe 79 1 34 P. LA2 69.4 J4 M LA2 90 3 i' J4 H LA2 79.9 I-14
TABLE I.4. (Cont.) 11 x 11 Jo a uAo 05.7 JJ R LA1 09 0 Jo x lao od.J Ja a 6Al 70.3 12 R 12 oo S alo be.o oo S tid 77.3 or n all dl.s oo S ric 76.0 13 c. 13 72 S m22 e2.o li a n22 03. 7 .? S v22 d4.3 7/ S f22 o3 9 14 x 1, s S de v4 4 9 3 uo 64.5 9 S do 93.0 9 8 u6 74.6 13 n is 27 n u23 70.9 20 x 024 d0 3 27 x 025 74 1 to a u21 ol.1 lo r. lo 2s S a2J 77.9 27 n u23 79 0 la a u23 00.4 lo u v24 74.5 17 n 17 2s S 023 79 3 27 n v26 15.9 i 2d S u23 el.7 2e d u24 71 6 18 u lo 29 a u27 e7.o 30 $ D2a 30 4 J3 S v2a 83.7 J2 a 924 10.5 1Y U ly 4d S lad 75.9 42 S lao 90.2 f 40 a .Ad 70.0 40 $ lad 90.7 i 29 R 21 41 M LA9 bd.2 ! 40 a lao 01.o l 44 S L A12 01.4 43 S lac 03.7 [ I-15 i
TABLE I.4. (Cont.) 21 e 21 10 n 2 0a.0 32 R r2 db.7 3? n r2 90 2 3: A r2 ob.c 22 M 22 6? n r2 31.1 D i' 8 r2 oo.d 30 d r2 so.o al a 4J c1.J 23 r. 23 33 5 r7 72.5 l 07 S riv co.o oc a s23 17.o i o7 3 riv 67.1 24 u 29 on S r7 oc. 2 07 a als od.d 53 S Y7 70.1 o7 8 riv dl.d zu n lo oo S x22 11.1 oc > k26 d7.1 s7 8 tv e1 9 57 S S d1 3 40 n zo do a ro 7o.9 so n- r o do.6 31 8 rJ 77.o ao a no oo.7 27 R 27 31 S r3 uu.o 07 8 rv . 77.9 al a 1J d 7 . .I 37 s rv 79.1 28 n 20 07 -S alv dl. 3 03 a 111 06.i o7 S Y1 ed.1 on S r17 od.y 29 x 29 as d all o7.4 39 R vil o7.o by .d s11 bo.7 av n all od.a i 33 P 311 sv R vil do.2 53 d r11 71.s 31 a all o7.2 59 R vil oo.9 I-16 l
1 TABLE I.4. (Cont.) 31 n J1 71 S rzJ u o. o di 4 JJ c1.3 71 'S V2J $5.4 di S 133 02.o J2 n Je 'l o a sat a2.7 7e .R Y J .9 dd.J 7's x rJ1 77.4 70 a rJ2 o4.2 JJ n Js 73 a Y25 05.9 7o S r2a d2 1 74 S tzo 7J.2 70 a n20 dJ.T J4 U JM 7b R s27 04.7 7e n se7 e5.1 7a n 2t os.3 73 R v27 d5.0 Jb x Ju 7e S :2e 51 2 17 a sta 03.7 7o S Y2o d4.1 77 S r2s oo.o ab d Jo J S CJ 70.2 J $ CJ 91.4 3 S CJ 7o.9 J S GJ 91.o 47 o 37 1 a c1 9o.7 1 d C1 l e l. o 1 d c1 97 1 1 a el 102.0 3d u de 1 R C1 od.o 1 d C1 96.9 1 a ci ov.a 1 R C1 vv.2 49 U av 1 h C1 69.1 2 a ul 76.2 i R C1 90.4 2 S C2 79 1 49 d 90 la S L10 100.0 1J S 012 100.7 13 S ule 101.d IJ S Uld 90.3 I-17
TABLE I.4. (Cont.) . . . . . . . ( 11 o 91 13 a ole el,3 IJ S ule 9o.4 1J a ul: 31.d la S 012 90.7 42 x 42 22 S ulv 77.o 22 a uls v2.d 22 S uly 7d.4
- 22. S uls v3 2 43 u na 22 S uly 30.J 13 8 ule o7.7 22 S ulv v1 3 13 S ull dd.J
*s u ,s lo a u1J v J . P.
lo s old 73 2 le S.uld 91.d 4 a ul 74.7 so u 43 17 a 014 e4.o ic S ulo 90.d 17 a ois oo.1 lo S 015 9J.5 40 e ao 24 ~ S v 2 'e o l.1 2J a u22 el.o 12 S 09 70.d 1J S ulu d2 2 47 n ,l 2J S u22 o3.> 14 S v11 72.J 2J S U20 83.A 14 d uld 70.J' no n 4o 2J a u22 90.1 2J S 92d 75 4 2J S u2d 91.1 IJ a ula 74.3 49 u 39 21 a uld .dS.9 17 S u14 do.9 21 s ulo 07.3 17 3 014 67.3 50 o da 2e x 017 90.o 1v a ulo ov.1 20 R ull v1.7 19 x Ulo 89.7 . I-18
1 TABLE I.S. PH085 p1 p2 p3 p4 pS po p7 po p9 p19 pt listener 1 1.000 1.0HJ 3.323 0.921 0.Jwo 1.002 1.222 1.000 1.000 1.200 2.de13 1.000 1.0d3 3.blo 0.957 0.572 1.202 1.222 1.02P 1.000 1.000 2.6221 1 000 1.0D9 a.331 9 272 J.40s 1. d P 2 1 222 1.aE2 1.220 1 222 2.5522 1.J00 11stener1.000 2 0.J49 0.001 0.J74 1.002 1.22J 1.322 1.000 1.000 0.Joi? 1.000 1.JdJ 0.9J1 1.J00 1.003 1.Jd3 1.PCP 1.00d 1.20d 1 200 0.9372 1.a00 1.313 0.9a1 1.J03 1.J2e 1.000 1 220 1.J20 1.2P0 1 000 2.944e 1.000 1.00e 0.d19 1.JP0 1.02d 1.J02 1.220 1.023 1.390 1.000 1.3E2C 1.J00 1.0ba 11stener a U.div 1.J0a 1.J2e 1.JP2 1 2K2 1.J24 1. des 1 200 E.d27b 1.000 1.03J ?.7ao 1.J03 1.002 1.000 1.d?J 1.o20 1.000 1.20e 0.voz, 1.000 1.004 9.ood 1. 831 1.JPJ 1.002 1.00? 1.322 1.JPJ 1.0a0 0.de55 1.00J 1.J"O J.olo 3.d74 1.J2E 1.002 lee 2g 1.42e 1.000 1 222 2 949c 1.000 1.JD3 J.7,7 1.J20 1.JCe 1.000 1.203 1.0PJ 1.00J 1.J30 2.7s3s listener 4 1.00d 1.00a 1.4H0 1.0JJ a.e32 1.905 J.o73 1.0Pc 1.Jea 1.02a 1.2Ps 1.cac e.9230 J. doe 1 30J 1.d02 1.J02 1.222 1.022 1.000 1.322 0.ds9s 1.000 1.DHe 1.000 1. 03 J a.7 ?.s73 1.731 1.JP/ 1.000 1.JPa 1.000 1.002 1.200 c.de32
'2 a.voo listener a 1.Jae 1.002 1.223 1.000 1.000 1.dDJ 2 71e5 1.000 1.Je] ?.o/2 e. so J J.o 7 a 1. J a C 1. 2 2 0 1. 0 2 0 1. 00 0 1.'2 0 2 f. v s 3 3 1.000 1.U3J J.cze 1.?03 1.492 1.0d0 1.002 1.020 1.00J 1.03J 2.d34e 1 000 1.0ad a.br 4.o23 4.973 1.Jd2 1 222 1.022 1.20J 1 202 2.7667 1.000 1. w3 listener o 3 9.007 a.s7o 1 . 3 0 .' 1 . 3 3 J 1.902 1.023 1.2EJ 1.2R3 0.0o:7 1.JBo 1.44* J.oo d.vd2 4.33J 1. dad 1 2e2 1.a2d 1.003 1 202 2.dc23 1.000 1.Jta J.7eo 0. s 9 o J.971 1.JJ2 1.ete 1.320 1.203 1.202 2.7134 1.000 1.03a a.s7d a.795 1.02J 1.J.12 1.J20 1.Jc2 1.000 1.000 P.8o54 1.000 1.Daa listener 7 c.507 J.v27 1.J04 1.dPe 1.eed 1.J24 1.d83 1 300 a. Sob >
1.000 1.naJ J.ozd J.so3 a.ose 1.JJ2 1.000 1.022 1.000'1.000 C.d>7e 1.000 1.ddJ J.700 1. M4 3 1.000 1.402 1.220 1.02E 1.000 1.00e 0.7734 1 0bd 1.JA3 J.392 0.47o J.d34 1.JJ2 1 222 1.022 1.220 1 222 2.72c7 1.000 1.0ad e.awz J.y44 1.000 1.300 1.222 1.000 1.000 1.822 2.012; listener o 9.661 1.doy *t. 49a 1.000 1.JHJ
- 4. dol d.32d 1.49e 1 200 1.det 1.000 1.Jd2 2.o961
- d. o21 J. so l 3.o24 1.d00 1 222 1.420 1.200 1.002 0.o470 1.000 1.03J J.av4 ..oJJ 1.000 1.300 1.220 1.023 1.000 1.000 0.996J l
1.300 1 . o d d' J.419 a.7eS 4.742 1.002 1 020 1.002 1.02J 1. Wad 0 4377 listener v 1.000 1.002 J.721 1.J03 0.9yv 1.J32 1.c00 1.000 1.000 1.222 0.9eo? 1.000 1.JHJ a.67J 1.303 1.JCJ 1.00P 1.000 1.22d 1.000 1.000 0.87c1 1 800 1. dad J.oe2 d.o4o 1.J22 1.dJE 1 2Et 1.J22 1.000 1.E02 2.9142 1.900 1.Jdd 0.7zo 1.JPJ 1.J0J 1.200 1.220 1.022 1.009 1.00J 0.71P1 listener IP
- 1. WOW 1.J0d d.oJo '4 975 0.do3 1.J02 1 22d 1.022 1.299 1 822 e.94c) 1.990 1.000 P.4v7 a.d7d 3.05o 1.000 1 020 1.022 1.200 1.200 2.453:
1.000 1.0HJ 0.So5 d.d46 1.Wed 1.00e 1.000 1.000 1.009 1.000 B.9142 1 800 1.00J J.247 3.Sd1 0.203 1.000 1 222 1.det 1.eas 1 202 2 2343 I-19 r.
T TABLE I.S. (Cont.) '
-. (
11stener 11 1.000 1.J00 0.4J7 0.d05 3.Jeg 1.002 1.eeB 1.000 1.000 1.000 0.7Edd 1 000 1.0e" J.4va a.d71 J.J23 1.Jd2 1 020 1.Jee 1.000 1 002 E.4eSo 1.000 1.0ild J.214 3. lod 0.000 1.J00 1.200 1.020 1.200 1.002 c.na3a 1.000 1.J3J J.254 J.5do 0.JJ9 1.302 1.220 1.Jc? 1.000 1.002 0.2c27 listener 12 f
- 1 000 1.09) .i . J o d d.71o 0.322 1.eJE 1 222 1.Jee 1.200 1.002 2.beed 1.000 1.033 3.651 3.soa J.vas 1.302 1.220 1.J22 1.000 1.000 0.7741 1.000 1.JoJ 9.437 J.oJ4 a.941 1.dd2 1. e e d 1. J 2 2 1.000 1.J04 e.77ee 1.J00 1. e ~t d 0 499 3.coi 4.97c 1.Jd2 1 222 1.1(c 1.g00 1.ded 0.de29 listenec IJ 1.00J 1.6ta J.731 1.J03 1.J?e 1.J32 1.000 1.802 1.JrJ 1.30? 2.9920 1.000 1. NfJ M.431 ;?.ed7 J.d2] 1.JHP 1 02J 1.JJa 1.000 1 222 2 4e64 1.0dd 1. 4.0 1.ood 1.olo 1.3?J 1.002 1.600 1.32? 1.000 1 200 0.o>cw 1.000 1.Jaa J.249 4. 9 9 -) J.JcJ 1.002 1.220 1.022 1.000 1.000 0.2417 listener is 1.J00 1.Jta .t.co4 1.Ja0 1.02d 1.JJe 1 222 1.020 1.220 1.220 2.9945 1.000 1. Ja d 3.77 2 1.;2J 1.302 1.002 1.000 1.J2o 1.000 1.000 C.7eJ5 1.000 1. dit J 9. 72 2 i.v77 1.?JJ 1.3e? 1.ded 1.d20 1.000 1.002 E.9871 1.udd 1.J1J .I. 4 J s d.793 A. oil 1.402 1.222 1.J22 1.200 1.202 E.4d97 11 stoner is 1.000 1.teJ J.523 1.901 4.39c 1.00? 1.J20 1.022 1.000 1.003 e.8221 1.00d 1.93d 0.o32 J.voa J.901 1.032 1 222 1.cca 1.000 1 232 e.oce) 1.000 1.011 a.J22 U. sz s J.o ba 1.732 1.c00 1.J20 1.000 1.200 c.oesi j 1.000 1.0JJ J.44J d.o13 J.9 2 2 1.J"2 1.200 1. Jet 1.000 1.300 2.4c7e listener lo 1 000 1. e 14 a.bys 2 955 d.v4: 1.3d2 1 222 1.Ji2 1.e00 1.222 2.9c7k 1.000 1.0JJ 3.o13 a.977 J.ooJ 1.JJO 1.020 1.020 1.00d 1.Jd3 0.0s72 1.000 1.eJP J.527 3.ody 1.J/J 1.J32 1.0e? 1.0e? 1.ded 1.d00 2.82o1 1 00J 1.daa .t.JJJ J.03d 3. o a d 1.322 1.E22 1.022 1. fed 1.020 2.34a7 11atehet 17 1.00J 1. Pad J.71o 1. 10 3 0.vv7 1.JJ2 1.200 1.320 1.000 1.000 0.9eo2 1 000 1.Jad J.39J U.955 J.714 1.000 1 000 1.042 1.22J 1 202 0.o24c 1.000 1. 210 9 5,7 1.00J 1.JHJ J.2dl J. L. / J a 1.J?e 1.J22 1 000 1.002 1.090 1 300 D.d:la listener lo sod J.454 1.J42 1.222 1.J2J 1.J0J 1.000 0.2949 1 000 1 0 tu 1.ded 1.ded 1.J22 1.23f 1 222 1.020 1.220 1.202 2.9922 1.000 1.edd 0.dzb 1.J00 1.J0J 1.23d 1.220 1.020 1.000 1.000 0.do4o 1.000 1.003 H.d32 3.701 1.440 1.000 1.002 1.000 1.200 1.000 0.832d 1 000 1.' dad .t. 0 0 7 197d 1 322 1.032 1 022 1.02e 1.220 1.200 0.0637 listener ly -
1.000 1.uaJ o. col J.970 0.dJ1 1.032 1.200 1.020 1.2eJ 1.000 0.9429 1 000 1.000 3.o2d 1.J03 1.32e 1.000 1 000 1 000 1.000 1 000 e.d309 1.00d 1.000 J.*ad n.ala 3. col 1.3ae 1.J2J 1. ace 1.000 1.000 0.72d1 1.000 1. dad 0.072 0 979 1.300 1.Ja2 1.220 1.J20 1.000 1.000 0.6 bed listener la 1 0 03 1. 0 t e J. 477 P.dbo J. dE 2 1.0d2 1.222 1.320 1.200 1.220 E.7224 1.000 1.0a0 0. slo 0.oo2 0.00J 1.JD0 1.02J 1.000 1.200 1.200 e.5299 1.0e0 1.000 0.195 0.38o 0.000 1.002 1.208 1.000 1.000 1.000 0.4b83 1 000 1 000 d.279 R.d46 0.022 1.002 1 222 1.029 1.000 1 000 0.2746 't I-20
4 . T TABLE I.S. (Cont.) listener el 1.000 1.Jea 0.71o 1.300 1.00d 1. ode 1.200 1.J20 1.000 1.000 0.9ede 1 000 1.0a? J.719 1.JP3 1.de? 1.072 1 290 1.020 1.000 1.det 2 7427 1.ded 1. e'9 0 1.o04 0.va) 1.03J 1.JP2 1 200 1.ded 1.000 1 202 0.vcel 1.0PW 1./HJ d.d21 J.903 1.o2d 1.000 1.20J 1.000 1.000 1.20J 2.545J 11stener 22 1.000 1.000 d.17 4 4. 4B J 0.000 1.000 1 20C 1.209 1.003 1 009 a.oJJe 1.000 1.030 1.4d5 0.d29 J.002 1.d20 1.ff2 1.300 1.0PJ 1.020 0.4522 1.JP0 1.JJJ .21o J..td3 1.00a 1.002 1. sea 1.0ed 1.000 1 200 e.442t 1.dWJ 1.013 .1.349 1.222 2.022 1.032 1 020 1.0E9 1.090 1./20 2.0:ce listener 24 1.000 1.UUa 1.o29 i.vo4 '.od7 1.32c 1.220 1. dad 1.eep 1 000 0.dvev 1.000 1. d '.13 4.792 1 403 1.J92 1.002 1.222 1.d22 1.2r0 1 000 0.d2ic 1.000 1.Jed 3.4o, :1. ao 2 J.v42 1.J22 1.202 1.dP0 1.000 1.0J2 0.7o59 1.000 1.~10J.0.o23 J.vol 1.dPe 1.012 1 200 1.Jee 1.000 1.JUP 0.os21 listener 24 1.000 1.JJr, u.aus c.923 a.aec 1.uee 1.220 1.020 1.200 1.s0J 0.7a39 1.000 1.00o J.7a2 1.JPG 1. J i' 2 1.10 9 1. 0 f 2 1.300 1.000 1. dad d.7434 1 000 1 . d d J .'. a d y 1 2d7 2.94o 1.292 1 222 1.Jte 1.220 1 200 0.oE1t 1.000 1.J30 .'. ass 0.v12 1.022 1.309 1.000 1.Jc0 1. ppd 1.002 2.n?fe listener 2d 1.ded 1.JJa J.olo 3. sd 9 1.aov 1.J92 1 de 1.Jde 1.d00 1 0df P.de/e 1.000 1.+1e .i.7vo 1.'c0 1.492 1.3d2 1 2e2 1.022 1.200 1 000 0.ofde 1.000 1.J.'a J.o,y J.140 1.3cJ 1.1Je 1.ded 1.J20 1.200 1.JP0 a.dd4c 1.000 1. 0" I d. of t 4.345 1.JPe 1.Jh' 1.JJ3 1.Jed 1.000 1 00' t'.do32 11 stoner 20 1.00J 1.J03 2.o34 3. sov 3.74a 1.002 1.20/ 1.000 1.220 1.J02 2.91oc 1.000 1.Ja? .' 729 1. le d 1.J24 1.JJ2 1.200 1.32J 1.02d 1.200 0.7422 1.000 1.Jud 3 40s .!. 3 6 2 J.9 4 2 1.Je9 1. GEE 1 022 1.200 1.Eoc e.7c5S 1.000 1.0PJ J.aa2 a. s12 1.a'J 1.00e 1.020 1.222 1.000 1.000 c.ds:7 listener 27 1.0DJ 1.JJo J.7o3 1 3d4 1.000 1.JJC 1 200 132J 1.000 1.WUJ 0.9313 1.dHe 1.E.1d J.ov7 4.933 0.94e 1. 0 a e 1 2 2 2 1. 0 2 3 1. 00 0 1. 2 0 2 0.7112 1.000 1.JPJ J.o22 .t .1o J 1.J23 1.032 1.200 1.000 1.J20 1.000 0.9427 1.000 1.0JJ J.d'7 a. 37 4 d.992 1.dA2 1 0e0 1.dE0 1.000 1 000 0.5313 listener Jo 1.000 1.30s 0.?ad 1.100 1.000 1.002 1.2RO 1.000 1.200 1.000 0.vevo 1.000 1.d0J d. son J.s2o 3.023 1.022 1.000 1.000 1.000 1.000 2.5571 1.000 1.o'd u.Soo ;!.777 1.32g 1.002 1 222 1.022 1.000 1 000 0.o75J 1.000 1.00J J.337 3. 091 J. lo i 1.002 1.000 1.Jc0 1.000 1.000 0.Jao7 listener 29
- 1. des 1. dea J.73o 1 42d 1.Jed 1.a00 1. des 1.000 1.000 1 000 0.9o94 1.0Wd 1. d 61 1 H.4od -1. d 4 5 d.422 1.00e 1.020 1.320 1.000 1.000 0.9047 1.0f0 1.073 J. sol 1.ded 1.00d 1.0ac 1.220 1.JP0 1.000 1.000 0.692e 1.000 1.ded a.213 J.319 d.990 1.d9e 1 200 1.J20 1.000 1 000 0 2120 listener J.i 1.000 1.J00 a.72a 1.J03 1.000 1.J20 1.000 1.000 1.000 1.000 0.9 esc 1.000 1.00J 0.S29 a.999 0.4J4 1.J32 1.200 1.000 1.000 1.000 0.53d2 1 000,1.00d 0 539 0.770 1.922 1.dde 1 229 1.020 1.000 1 299 B.6711 1.909 1.00e 0 1o7 0.493 0.000 1.000 1.000 1.020 1.000 1.000 0.le73 '
l ' I-21
a T TABLE I.S. (Cont.) Listener J1 1.d00 1. dd e ?. 3d v 4. 7 21 0. 0 0 d 1.002 1 22d 1.d00 1.000 1 009 2.be9d
- 1. Wad 1.2da 0.soo 0.oz7 0.0EP 1.002 1 222 1.022 1.000 1 000 2.1024 1.000 1. 2.10 2.c/d J.2sd 0.acc 1.JJC 1.cPJ 1.300 1.000 1.020 2.4cd1 1.000 1.o1A J.223 A.4do J.J04 1.JD2 1 220 1.000 1.000 1.20e O'.222e listener J2 1.00J 1.c'? 3.oaa 1. yod 3.93J 1.032 1.022 1.0cJ 1.200 1.002 f.9741 1 920 1. J .i J '.o93 .t . v d 9 J.ve4 1.JJi 1.200 1.002 1.dc3 1.e00 c.7c4e 1.dde 1.0if J.3d2 '.,44 3.793 1. 342 1 2c2 1 029 1.ded 1 00c 2.oect 1.dOJ 1.00J J.laz J.440 a . J 't J 1. J 3 2 1.222 1.022 1.200 1.030 0.14:e 11steuer JJ 1.40J 1. ott a .t . 5 3 3 J.yJ1 A . ,10 0 1.0u2 1 222 1.334 1.J00 1.J0J 0.7315
.1. d e o 1. J A ' *t. 71o 1. ? e d 1.42t 1.402 1.Ege 1.Jff 1. 2 e e 1. 3 0 C C.7:51 1.00J 1.PSD ?.113 J.,44 J.71a 1.JJP 1.?OJ 1.39? 1.320 1.c3J 2.occ2 1.Jdb 1. c 3.* v.doo .t.y2e 1.Jcd 1.Jud 1.ect 1.Jc0 1.J70 1.Jud J.do7:
11stener J4 1.000 1.oDJ J.,21 a./o3 0.000 1.002.1.2e2 1.000 1.020 1.P00 0.7 Pas 1.dW3 1.004 J.711 1.Jed 1.uae 1.dJ2 1.ded 1.302 1.dC0 1. dB e P.7 2:e 1.J00 1. d:t
- J.lo7 J.119 d.J0e 1.dd2 1.ek2 1.3E7 1.222 1.ted 2.otte 1.000 1. N' * .t.alo 3.397 1.002 1..k'2 1.022 1.JP3 1.000 1.000 c.Savi listener 33 J.bbe 1 0fu 1 4cm 4.20I J.JeJ 1.4J2 1.ite 1.0E2 1.000 1 20e 2.deSe 1.000 1.002 d.aol *.s33 J.01J 1.Jac 1.ePJ 1.JcJ 1.J00 1.J0J 0.524a 0.192 1.J33 * .e47 1. % 4 4. ova 1.JJO 1.34J 1.J0J 1. fed 1.JJ3 e.4o44 (
1.JbJ'l.21J 1 49y 3 379 J. J z d ' 1. 4.! t 1 222 1.322 1.292 1 22d 2 2937 listedyr 30
- 1.Jee 1.cPA '.ols .961 J. doi 1.J32 1.J20 1.320 1.000 1.200 c.9d5e 1 004 1. C' d .i.037 1. J d si 1.Jed 1.d02 1.deu 1..10g.1.0E0 1 200 0 8453 1.WOB i.010 U.87z J.sa2 0.317 1.232 1 022 1.300 1.000 1 006 0.7 4 v E
, 1.d00 1.00
- l. cod 9.yo3 1.02J 1.232 1.0/2 1.fEP 1.JP0 1.000 0.ov77 l 11stener' 37 l 1 000 1.s4J 1.oJ2 1 100 1.02d 1.d02 1 000 1.022 1.200 1 200 0.9533 1.000 1.000 P.o7a 1.oPJ 1.Jf3 1.0J3 1.022 1.023 1.000 1.200 0.8e14 1.000 1. ave J.o7'a 7.vnd 1.444 1.dag 1.220 1.32e 1.000 1.000 e.971t 1 000 1.daa J.7a3 1 303 1.d22 1.c02 1.22e 1.aE2 1.200 1.000 e.74c7 114twaar Jo t 1.000 1.c00 ..ovo ' ?.39s 4.995 1.022 1.22d 1.J22 1.0cJ 1.00P 0.9e55~
l 1 09e 1.Jao 3.co2 1 03J 1.0Je 1.J0J 1 2/d 1.022 1.d22 1 002 2.8393 1 1.00d 1.2)J J.d7J 0.d23 1.JJ-J 1.JPJ 1 020 1.J20 1.000 1 000 2.9CIC 1.000 1.JDA d . 7:* 1 d.vv7 1.JCJ 1.002.1.c2P 1.303 1.020 1.00d 0.713b 11stener dv 1 00W 1. 0.13 0.767 1 400 1.ded 1.300 1 222 1.JE2 1.020 1 220 e.9923 1.000 1.000 0.701 3.yo4 0.va4 1.0a0 1.200 1.000 1.000 1.003 0.7143 1.00w 1.00u M.dob l.J47 1.00d 1.300 1 200 1.ded 1.000 1 000 0.9142 1 000 1 20o a.deo a.o74 a.993 1.Jae 1.ce2 1.0Ee 1.000 1.cBe c.5319 11stener 10-1.000 1.Je3 J.v10 1.10 J 1.J20 1.J02 1.020 1.000 1.000 1.000 0.99c4 1.WWW l.ded J.911 1.d00 1. dad 1.ddd 1.de0 1.002 1.000 1 202 2 9154 1.000 1.000 J.707 1.J00 1.000 1.300 1.000 1.020 1.000 1.000 1.0222 1.000 1.300 0.o7J 0.960 1.000 1.002 1.000 1.000 1.000 1.000 0.6892 _ I-22
TABLE I.S. listener ,1 (Cont.) 1.d00 1.edJ 1. 7 J 7 1.eea 1.322 1.000 1 2E2 1.022 1.200 1 202 e.9e95 1.000 1.800 0.do3 1.30J 1.JPJ 1.032 1.202 1.000 1.002 1.00d 0.o059 1.000 1.daJ 7.0o4 9.v52 1.300 1.dal 1.de2 1.Je0 1.eP0 1.000 0.9732 1 00d 1 0ad 0.741 1.Jed 1.J24 1.0J2 1 220 1.3f2 1.200 1 222 2.754k listener se 1.000 1.0/o 3.ev4 3.dv4 J.941 1.Je2 1.J00 1.02e 1.000 1.000 0.9sde 1 00J 1.01) J. col 1 101 1.JJe 1.JJ2 1 200 1.022 1.000 1.g02 0 6bc7
- 1. d.0 0 1.JJJ .t.417 J.o14 9.vss 1.032 1.202 1.322 1.000 1 200 0.7ooo 1.00J 1.ooJ 0.7e1 a. sol 1.J20 1.J32 1.002 1.0c2 1.000 1.e00 0.7139 listener ,J 1 000 1. 0 s .1 a. des 1 322 1.J22 1.dl2 1 22e 1.d22 1.220 1 222 2.9924 1.d e o 1. Jo L a. o 's 1. '00 1.d22 1.300 1.J2d 1~.022 1.000 1.202 0.d143 1.000 1.Jia J.olo ?.247 1.4>d 1.JJe 1.ed? 1. Jet 1.eed 1.2JO 3.9721 1 023 1.
11aCduti 1t dad 1.oa9 4.soo 1.122 1.de2 1 222 1.020 1.220 1.200 c.6bte 1.00d 1.da.' O.dzo'1.12J 1. J P a' 1.dd2 1.202 1.d2e 1.000 1.000 0.9332 1 000 1. ua 5.o0, '.977 :J.dio 1.e?c 1 2de 1.02d 1. ec e 1 2 2 e 2.e7et 1.000 1.JJ? J.o73 1.v,J 1.JCJ 1.s02 1. Cad 1.200 1.022 1.20e 0.vce2 1.000 1.JoJ J.4Jo J.7dv J.014 1.JJ2 1.020 1.d2d 1.000 1.002 0.4D72 11staner ,o
- 1. d d it 1. J a 4 .l.773 1 10J 1.10? 1. doi 1.e22 1.dec 1.0P0 1 2e e 0.99 e9 1.000 1.J?0 l. . s a o .333 1.32J 1.332 1.'000 1.022 1.20J 1.d00 0.e351 1.0CJ 1. Ja e J. cl a .. '
cod 1.JFJ 1.400 1.JEJ 1.Jdd 1.900 1.000 2.921c 1 000 1 0aJ n. cod 3. 97 d 1. d 2 1.002 1 22e 1.de2 1.200 1 2e2 0.be41 11steder 1o 1.009 1.Jdd 0.7J3 1.403 1.JJt 1.dJ2 1.d20 1.ded 1.0E0 1.d00 0.9094 1.ddo 1.e!> .t.7,1 1.Jed 1. Jig 1.dde 1 222 1.d20 1.200 1.d02 2.7337 1.000 1.Je> s.171 a.aaJ J.vla 1.JJ2 1.eCJ l'.dce 1.000 1.00J 0.79/o 1.00J 1.JoJ J.no4 3.117 1.ae) 1. Joe 1.200 1.020 1.e00 1.dde 2.5703 listener 97 1.000 1.cCa 1. 7 J 3 1.10J 1. ii 2 J 1.J30 1.J22 1.022 1.000 1.200 2.vcSJ 1.000 1.c3? J.o42 '.vol 2.olo 1.J00 1.JPC 1.000 1.000 1.800 0.629o 1 000 1.dda J.,o*1 303 a.773 1.dii 1.002 1 22d 1 022 1.220 1 022 2.b74a 1.00J 1.03J J.s2a 3.dsd 1.002 1.220 1.02d 1.000 1.200 0.4cto itstener so l l 1.Jud 1. im J d.627 1 3cJ 1.0J0 1.000 1.ee2 1.000 1.000 1.W00 0.9931 1 000 1.dlJ 1. coo J.97d J.doa 1.002 1 222 1.d2P 1.200 1 200 0.o7d4 1.000 1.U?J f.076 J.e44 1.J2J 1.002 1.200 1.02C 1.090 l'002 0.9000 1 00d 1. 0;4 3 d . 4 31
- 0. 7 d 2 'J. 7 v e 1. J J e 1. e ? 0 1.0JP 1.e00 1 082 0.4de?
listener 99 1.000 1.020 0.700 1.J01 1.J03 1.0J0 1.000 1.000 1.P20 1.000 0.931s 1.000 1.003 J.797 1.J00 1.dCJ 1.302 1.2PJ 1.000 1.000 1.000 2.0071 1 000 1.03d d.o25 J.dda 1. 12 2 1.dd2 1 222 1.020 1.200 1 202 0.9350 1.000 1.003 J.oes 0.vo4 1.J22 1.J00 1.c20 1.022 1.200 1.000 2.0570 listener 60 1.000 1.DAP 0.773 1.JP3 1.000 1.000 1 000 1.H2O 1.000 1 000 0 9999 l l 1.000 1.d1d J.737 1.JP2 1.022 1.d02 1 222 1.d22 1.000 1.002 0.7094 1.000 1. 3a J 9. 004 1.473 1.JCe 1.372 1.222 1.000 1.000 1.000 0.9290 1.d00 1.001 9 67o 1.945 1.d00 1.000 1.d20 1.J00 1.0P0 1 00d 0.590o rural, ucuan s o,ulaLivas ? 119722,4031J ptrur pturo ptalt l e.o77 9.vd7 J.o99 l 9.654 d.620 2.730 l 9.7b1 0.dde 0.716 l 8.419 0.003 0.4d7 j Z-SS l
3 m APPENDIX J: RANDOM SELECTION OF POPULATION-WEIGHTED LISTENING POINTS AT THE INDIAN POINT SITE The objective'of the listener-site-selection process was to identify 50 randomly selected building locations within the EPZ surrounding the Indian Point Nuclear Plant. These locations are
$ssumed to be residential locations and are called " listener sites."
The various steps used in the site selection procedure are described below:
- 1. A population-distribution map (see Fig. J-1) con-sisting of a 10-mile-radius circle divided into annular sectors defined by interior circles and radii, was superimposed on topographical maps of the EPZ. Population distribution informa tion consisted of the number of people within each annular sector. These data were used in order to population-weight the random selection process described below.
- 2. Each annular sector was first assigned a desig-nator, ranging between A-1 and R-10 (see Fig. J-1).
A range of numbers was then assigned to each sector according to the population in that sector. For example, Sector A-1, just north of the site, and l sectors B-1 and C-1 (moving clockwise) have zero l population and thus were not assigned any numbers. Sector D-1 has a population of 35 and was assigned numbers 1 to 35. Sector E-1 has a popul'ation of 10 l and was assigned numbers 36 through 95. This pro-cess was continued until each number between 1 and 256,015 (the total estimated population) was assigned to a particular sector. A random number t J-l
generator (available on a Texas Instruments Model TI-59 hand calculator, for example) was then used to select 50 numbers at random between 1 and 256,015. Each number selected represented one site (to be chosen later) within the sector containing that number. Thus, sectors with larger populations had a greater possibility of including chosen listener sites.
- 3. Having determined the sector lecations of each potential listener site, the next step in the procedure involved selecting the actual sites within the respective sectors. This was accom-plished by first over-laying a rectangular coordi-nate grid on each sector of interest on the topo-graphic map. The grid 'was composed of boxes with dimensions of approximately 1000 feet square, and -
a each box was assigned an X and a Y coordinate according to its location on the grid. The grid was positioned such that the X-axis was oriented in the east-west direction and the Y-axis was oriented in the north-south direction, and such that all parts of the sector of interest were covered by a positive (X,Y) coordinate pair box. A random number generator was then used to select random pairs of numbers within the X and Y ranges covering the sector of interest. Each X,Y pair was used to select a particular 1000 feet square box on the map. If there were buildings within the box, one of them was arbitrarily chosen as a listener site. If there were no buildings inside the box or if the box fell outside of the sector of interest, that coordinate pair was disregarded and another pair was chosen at random. J-2
T For urban sites in the pink " building-extension" area of the topographic map a residential building was always assumed to exist, and was selected at the center of the pink area in the 1000 feet square box.
- 4. The above process was repeated until 50 listener sites were randomly chosen. It was found, however, that some major urban communities did not include any listener sites, and thus the chosen sites did not properly reflect the population distribution in the EPZ. Therefore, the selection process was con-tinued until this condition was rectified. Four new urban sites were randomly chosen to replace the four most recently chosen rural sites. This re-placement only af fected the balance between urban and rural listener sites. Since the subsequent analysis treats urban and rural areas separately, this replacement will not bias the results. It will merely ensure that no major population concen-trations are ignored. The above procedure resulted in a pseudo-random sample of 50 specific listener locations, distributed throughout the EPZ as shown roughly on Fig. J-1.
I i J-3
T t
. h 2 s . ~
A A , i e 8 .
- e. '
4 e a e D P N t s A . w_ n --.- - - ..- -- ..
-- - - 1 l'..l A 2
$ I 10 E
/
g
. ,e I
~
' c
. /. -
el L
- i G
.e .
e i B# g ' 14 J . Population Density /Dlock I I below 2,000/mi 2 E 2,000-4,000/mi 2 C , 4,000-10,000/mi
, 5 FIG. J-1. SCHEMATIC SECTOR / ZONE DIAGRAM FOR INDIAN POINT EPZ WITH 1980 POPULATION DENSITY DISTRIBUTION.
J-4
1 APPENDIX K: SAMPLE SCENARIOS FOR THE EVALUATION OF SIREN ' ALERTING AT INDIAN POINT Scenario Season Summer Summer h Winter h Winter Time of Day Weekday Late Weekday Evening Late
.\fternoon Night (Rush Hour) Night General - Wurm, clear to warm, clear to Cold, Stormy Weather Partly Cloudy Partly Cloudy Overcast Home/ Vehicle Open Open Closed Closed Windcws (& Storms) (& Storms)
Temperature, F 80 70 30 30 Relative Humidity, % 65% 80% 70% 90% Temperature Gradient, O F/100 feet -1 +0.5 -0.5 -0.5 (meas. heights =95' & !7') Wind Direction: : General SSE NNE NW SE Va11eys Up-Valley Down-Valley - - Wind Speed, mph 10 mph 6 mph 10 mph 15 mph (meas. height =100') Percent of People Located Outdoors 20% - 5% - In Motor Vehicles 6 15 25 - Indoors at Work: Commercial 23 3 - h% Industrial 7 1 - 1 In Home Sleeping h 95 - 95 In Home Radio /TV 20 - Ib - In Home Noisy - - 3 - In Home Active 6 - 35 - In Home Isolated 4 - l'h - In Home Quiet 10 - h - K-1
APPENDIX L: SIREN IDCATIONS FOR 'IBE INDIAN POINT EPZ This appendix provides existing and proposed siren locations for the Indian Point EPZ as of 25 August 1981. Siren locations are provided on a set of topographical maps (Figures L-2 through L-12). Figure L-1 shows the relationship of the individual maps to the' Indian Point'EPZ. A total,of 88 125 dBC sirens are employed of which 12 are in Orange County (#1-#12), 24 are in Rockland County (413-436), 43 are in Westchester County (#37-479) and 9 are in Putnam County (#80-#88). Table L.1 provides a guide for locating the sirens on the topographical maps. i L-1
l _- ( TABLE L.1. SIREN IDCATION BY MAP. Siren # Map # Siren 4 Map 8 Siren # Map #
~
1 L-3 31 L-ll 61 L-8 2 L-2 32 L-ll 62 L-8 3 L-6 33 L-11 63 L-8 4 L-6 34 L-ll 64 L-7 5 L-6 35 L-11 65 L-7 6 L-6 36 L-ll 66 L-7 7 L-6 37 L-12 67 L-7 8 L-6 38 L-12 68 L-7 9 L-6 39 L-12 69 L-7 10 L-5 40 L-12 70 L-7 11 L-7 41 L-12 71 L-7 12 L-7 42 L-12 72 L-7 13 L-7 43 L-12 73 L-7 14 .L-7 44 L-12 74 L-8 15 L-10 45 L-12 75 L-8 16 L-10 46 L-12 76 L-8 17 L-10 47 L-12 77 L-8 18 L-10 48 L-ll 78 L-8 " 19 L-10 49 L-ll 79 L-8 20 L-10 50 L-ll 80 L-8 21 L-10 51 L-11 81 L-8 22 L-10 52 L-ll 82 L-7 l 23 L-10 53 L-ll 83 L-7 24 L-11 54 L-11 84 L-7 25 L-11 55 L-ll 85 L-3 26 L-11 56 L-12 86 L-3 27 L-ll 57 L-12 87 L-7 28 L-11 58 L-8 88 L-4 29 L-ll 59 L-8 l 30 L-ll 60 L-8 L-2
1 l i l' Figure L-2 Figure L-3 Figure L-4 Corn Wall Wcat Point Os,cwana Lake Io. D Figure L-5 F'gure L-6 Figure L-7 Pig e L-8 Monroe Poplopen D==Wkill Mohegan ake Lake C IN DI A N \e ' POINT 1Y7
+ e i J .
2 Figure L9 Sloatsbu'rg Figure L-10 Figure L-11 [FigureL-12 Thiell Haverstraw Ossinin l l*W FIG. L-1. SIREN LAYOUT MAP IDCATOR. L-3 l l __ _ _ _ _ _ _. - - - - - -
T
~
4 . n . ..m <,...ac.... .? (
.* '"4 ""4' .7. 'O. .-'."" FIGURE. L-2 .. = N . /'
* ?'", g- wd L
~i * '
I***'""" ~~~
- p e. . 2,i,, W. . ...' .: ., :j' 'hW'T ]e1p f. ~, ~ ' ' ' . s -
-~,;..; j ,. y.. y:-- 7g . (n
.r[.J> .gm ~ w. ,
n' c '
. . , . ,y g y. v2 -s . .,- (t .y ,
, ' :.~
- - y'
..s .j -
-: ,.)^ if' 7: f '~ ~~.;. . W_ y", Jm g , ," r.ueE* 1 N ..f, * *-
y ! 4 ,,% 3 g,'3.-
* ' ~ -
[y '
'CQ ,qt Nle3. ,",rs g ' e,.M@%@N$hs'h F Ydjd # 7[-l b
r? f t .O g M. 1.C,@"j a%a. y;@@ yip $ , . a : s v. ; s.c wj;,&. y Rj'W ' .4 w .-p k,: ,k. ,.o..:,
, _ n -
. gp \m. 1.
. ,ya..y.g~g, ., x.
, .. . p.,
..-ka ,l,NIYkz' [c , W. .y,,\.I , ' k i'*?xky < ? *** g" $.
a'
.* s 5 . j , tr , .y gh..'i . > / .r
- 1. .
+ -
?, - mS , Ql' o y - '. . '
' v = 4 ~1k..
.4
! KQ "..,Q
\ y.m J W.', Q &o
.s n , f .' ~m . . f 1 .[. .
3, h . '.., f7:t h ,i a u h:fY ? . b.? %. N0. b,h & 4. p % W.;]g?^ :>. R Ra.
,U,.,!c..YJ a g.- h. / . ) .?, @ .. ?../ ["
1 l- ~ . ..aw't;.CG ' .. i 6i. l
.j 3,-
. u. .,q M
Ww$h$b$5(,$$,$ l 1 l
$o%o.lb f&/"
o,, / -
,c.
.>s r
2af w~e A
' vs
# l. ),. % **
- G
.],;p -
e.
' W',l i *vi
,gs .
e ' TU Q-m k sf ( ,
,V; f ' ',7f'
. r ,
- a.e g; : '
-'vQ-1
.s.
w,, % '1
,p / g . x .fxV.r. :l:'.: .
\ c ..
.:.v l
j..yyf .
.h ,.
,3- .
. . . , ~;3- __ m. .
--,, ~. 4,,, - --
[' .2~ '~.'~ . .
-- r.f : .5 C5 F.75 F - '.**~~.~. h.
C."."-'J "" . . ~.- . ' ',, - - :2. . m.- a--- 3 ._:..,':. _ .- ." or. * ! '" " '~~ r._. _.-.___.t.-g*=.... g ... -. - , , , EE--- =r r- " " " ~~ ~~ _ _ . . . . _ . . . . _ _ . --' ~. "
=~^~ ~ =~~_ -----
L-4
v.
~ ~ ~ ~ ~ >,
T. 'JJ;. ';l..*J,~ L-3 FIGURE -::d":-7 Ve _ - - . ,,. m . A .O. _f ,.7, : ~
~
o,, r
>p* .
~i -y
.h ,. 'g . n n-1 t .
-f .
3,{ - 1 1 ! J f: - e
- 3. 'f.&p'T -
, ,l t -v; ,( ,:. .,,.. m, c.
* -; s's g.,,
- f. $ , .
; i
' '.y'l '
., f. 1%j ~)4 ';
l l I
' g .:s)'.
g\ - : .
>i'. ~
.r.gd , t .c ~
^/l.
L,7 . .s . 5/c j i
. si
.1 3
-1 *J
~.
~
,yt f, f}.$.
."5 :~
.? -
,. .a .f j '<'.s 't
.y_
t lx : I
, j.,.: 5.
!NY5h.e?N.s ,Lr 1
- , h I hY
,I
)?r tY'.D'.
- p. '.
h72/#. -N '
, */, F8 f,IM
- . a @ h I ' 3p,h.TN$ .
g ', ^ ('; '*2l v -k ' l , j
. m.. ,
i
'* y.}* , . s.t
~
p, , p-- U,- +
+ ~r .~ ,,. . ', . ' '-:
. 't
~
g
, w de.: .
L .%g' ,lP'
\
, .*l, ' . , <
) j' ,.
- s. . .,-
Yp at ' o
, l \y' ' ) ~'
(j J
- , g u. ':- .
y. 5 ,. Y_ ' r v. 'y - f. J
, n. g y g P .j),.
- m ... . --- a.>..
2 ,
,a 3
j :-.--_,_---- -...r --
--w-- --
- - -- :'"" . -~ g.s
/<' , . -- = : _ - -
- - - - - - - - p M ,_M_._._ _'_ -- . .J.-_ >
~ .. -
_2E_- EEE:i--- ga. :=-=.d?'__W- r - -, 3 .,3 l '---- :=r :-
-,. .r.r
- ~ ' '
L-5
T YN
.a
- $ E ,'i.';',"
FIGURE L-4 [d EY1 '
. .__ . , e
~1 5; . . - _ R . .;, '} ' g i', ^ y . if *y, '),
{' ' h ,,
. %jsq
-Q '
ds'5.t# Q$
. : - h
N. ' * '
'I.h 42)f'.fgh'th 's f ?
J
' lpg.[h'kh'ylOl E
J' - L- ud b. % ,, )( jM ,
; h (_. N ;,) @-Vp1.l%
c}.;. j..QNr ^ gj ..8'r $
,o . s. -
"f . a ' . "% "--
~
-[.j
-N 'j- ,^'% '
t+u . 4 'g:,
- i< ". .pt [h'.
k':{. ,.'t"*'U ., s).'y - y ; x 1,9 %.,l(fg <gl; ,r ; .
,yT')?ti.-. g.h.,
i .
', . ,f9,, a '/ 'E .
..] { .].
h1:PR W ))!'.ih ('?L f.,. ,4).y
')rl'%y;7..99~P;M'~;'.=.>.'i;.QPg,: ;)b; -y .,g g
,)lf .l lI ,. s ..,.' _
L i
,s, { +- . :,
g,
, ;uj
.k;'l*/ .7-MV f.! *, ,i ', . O -,\ '@',f,$$Q[g, '\. .
E .' ,.
< y)N
'ht.-',fh'..9'Q , i f d/,4f .l} k!E' V ]!, ..g)f'!Q.
4 - '
' p, 7 1
., , :vy;f y-;d i
~
h ,. te . f]5* '/l!l; *JB?#1W *
;' r %.
$'i:,hl?$ , e '..
/
, ',. (
-w '
I 5
.ll*]y 7 ,,g
[' ;7J.:;,!..V ' { h:j-}a
.h r -
l-j. g j M . . , , i
- 1_ , . ____. .,,;,,.,.. - .___.m., _:--
/ = = : ~~~~ ~ _ ..
.. 'diL _. . '- ~
~
~ - _ %'
EEr_.:i5~~[~- 2i, - .. -~ _ .g F~ - ~~~~
- T-
._ _ -- - -- L _ _ _ _ _ _ _ _
_A *:rfr_m.' - L-6 l -- ..
- _. - =E-- ncug t-5. . ...EE51.I -
;8 '. -
e,.w.w .
-s m . g w wpw~ <.
- an .o g g %e e .
xwcn 6,,(,a. u n t.-~ . m: . A m i. w ..a u s . i'
~~
s ~4 g *,. s ~. h . ,n ,-h.q,'N.t_-
' & ._'A ..'l l!
i r- .g
\') g..l > , ..
8
, '3
{+. -t-h -
.*- b.*rN f" y 7
,!^ .r',V '$~v '
, . a 9 6..4 . :kp 2 ~. I m' _l-hf ti j y pld ..
.?'Qs([~b ;y.y
.n. .' ' e. ,
. g TF %-:? .7 .9N , m jf;s v.h ,, y .: P.%\ .
3 in . e-k 1 Y
. u.
W a g :,A m( a u . ..
~
4 4 W 3. m4< s M .s c .9_M..e . t.1, . n
, .'T .
5, j" J, &.., L }4]:R, W' f... R W. ,.'
~ 4 y .r _
g l' ' dfg
] ,
/' , ,, l, 4 4: ,.d.
~
p ,m' r - -
.A,'ic
...~ . ,
\
tj - , .v w r . .
'i t -
h, , ' f .' ; . ,].h 1.. , - [h 'fhz ,yr
> . r c l , ., t w o:.- .e53 9 a.;
. ;9 Y .. A + ? w j i 'h ;.4 '? 'f',~-
. 4. :gqa m $ 3-.g, a-. ;
S.
- - ,x - pin .
jf, __ -- - _._.
= - :.
s.. 4 g e
=._-- _
2-
, 3 c
._. _._ . . .__ : q
[~~ -
~
'S
~~ ~
. - = . .
- - :~
--- - - - .n.:.s r.araa. --- _
' * . T / g '. *t. e. 3 .,g . =" = = - = = . - .
. as L-7 l
_ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _ _ _ .__ _. _________________________________________.D
T
- .:: tra"-
N *~;;;f; ,', .;, ~ - FGURE L-6 - f.- : . .=--- f. .1
.: y 3.... , - r * .m
. , p. -
p,ja. , e .. ~, ', 5 . -
'11,i,/ ' / J *?. n. s!I - _ '
$") 't i,,'T"
_ ,' '* l.
,. se ,,.i
?.
e, . I.'lz'f - I 2
'r A-l.1 b;il ^ ,A :n ,.
,.:_ 1,
;. ... ~~+ i.pn ; - Q .f-{*4f Y[' ..rQ ,.) , '
*** ah h f .I y '* o-
$,y,l'f,\
- l -
l} '.\ ff
+ .;. & ..; .s S * . I, !
,.f,( J .. .' /. . . o y. . . . . e f
/eos.-
- ~
,b~-
.s ,/ / ',.
. - %
- y> Q< /.,,
o/ / ..u : i' l a e , r o h.p.i .I- y f r. *
~. .f
/ N.4 5 ,'
/ ,s ' 5:14 .m ,- ,,e,' )
{ ' . .
, . - , y ,- . r. ,,'..* .
i:.
/,~, 3 g.l
\- 'y *. . ' s . '
o s I f',
./
/
~1 t
I, ' I
; / .. ,
.._-s
\
.I
,' g *. #
j' /'/ N .< ,
# /
l / . ./ f,- 4 ,
- s. g
*( e
/
.",# a. 4s *
-),
f, , ,,:, , -
, I r . .,. -'
. ~./ . .g ' y.7 '..
-4
'5" '%. **~
,'j
. i.1,
, I^ .
.m.w, f. ,.
', .;l I .N
- .- , . .. _ .1, .
. ;n ,
* ,w : .y ,
4w 1,lr, t. ' ' f s . . .
..'!t.\ *
.si) f .' .
\, i ,r *2 ' l .n t " . . '
~
~ kr . -
4 s . ,, s.f, .' I:' ' ; '.' _ * .
- I o
. ,r, ,
I, // p' ) + - . W ~ c 'g vl.'.!g .o
.- yl
,i . < . .i f :b.; /
QM R. -. ,.. > . ( ), - [' ' .,6 #,.,'/ i i ar;'ac f' J: g4 / g. .
-"-- r r ,. M " L , ,. s
,),
,t . .,
).:. ? ,l f ,
'.,I.-
, 1. . ,,g7 .- *
. , L. . w- . - '
~
,f
. . r. .rj :
a
-Q.y.,: ,8y.
t.o ' vs
- .m / .o e, . q y%_,_sy
. ^_--
3..,al ( ..j . . ,
,. = . , . . ,.
n .. .%- %'- .p.
}..'% ~, YcO : :(g., 4Q., .YK; ~~
-a ~
l, ? '.b f'y'Wy?Aj:'I',-::,&~ y' -
. ? g ,r.. .
+... .,,
i y. KV' .;...~-
.V .
.? . n.~ ..
- v. .<
- ym 1. ;; s * -
r s-p . e ., % _ c s
.,/ _ . iw .-
~
, . , . . . r*
. . ,.. w s%.,,
r -
.e , .=
.s ,. .,
s. v.,/_. , ,s
- N _
,~ ' ', .
( AD
, . .:K . . ,;. , 0'*.
> %_ ;'.,g-s p.';
. ,. e
.,-_ ,f* 'f f/T.. , .% ,.J:y if'~.
^ ne g j;t 9 gs .. . w c- g,, ' ;, ~- q..r. .
3 \
, -3 p ,- g .' .
2.0 q. .l3-Lc- ,T 0.& 4 n qG t[~,.,tf3
. p;,..? ,
- 4. :>y..- f W
c fy <r , . ~ .8p. ,t
,\
, .i r . .,
R g ND,y 9"
;.y , ,y /
E..s $. ,, . w ~ l :
. r/,
f...*,
.~ &
/
/
.1-y.'>, ' . < = . - .
, ,. n >
.. , p .- - -
;. y +e . .v.g.. , . yT' ~; -
ye u .s ' .s %(4* s?/p..?, -<.
., 9-M p .y ;y,.
J T
- i v -
J. s . b .Q;.m f )fg'y, g, y,R ~, Mf . f p.
., 't .j-
'b i g
- t. ,
1 . .. . . . - . .. . w -. _. .. n . -- - -- 1 - _ _ - . . - . . .. - ws i / EE _. 5 :. .__ _ .. .a .6 se = - w. - = = n. \ _ =_. - _ - .
., =._..- ..; -
8 -, :
. "" "l'"""~ l:
. . _ y =-=== --;.,,,,,
_ ~.-.
*~
h L-8
. . . . ~ r
/
%. . ;fi";'3. *,,,- ..
.. TIC Uf, E L - 7 . . ---. ,:*[. ."j _.
. /
~
k -
, r 7 ,. ~
g
= v. ', $. _ -
,p ,
i , N,, , ap , "r' r .-
.,~ ,y% ' e , f I'. d
, t .. . .
6/' j f, vf '! Ir' ~ s
/ , '-
c 5.9 ek I M
. w h
,j, f , -- - l.-
g ..
- b. g:y' . o -
W ' o ' .: i.;
,y. w .
s~t. ,t p h .J. a j I
'II / .
j\ ,.
,, 8 *M "n 4,, -
.. . e Qu , . j, !! A [ 3 ,- .,,. .
tja
,y, p- ' ..' .
. a,
,,,-c. - -
9 ;- , , a
, . , , ,a ,y .s
.s
' , . . . .? N , x,
- a. s .
~
'. u. ),, ,Y. , .
*' . \
N~ , . - -
'+ ;_t tj; .
,]. s, 7
{.54I -. Tja, .:
."~r c'h .
"N s c , 4. .Je/* c r -
..; S. h.y9 8 ;.
~.-. . .
. , y ,, 3, -
n.s .
; ,7 .-
]\f 15 s l;y' M 4, _ >
. g T' u -
*=.-
, -rn '
',:9, IDP 'g
. . ,,, ,4
' ,.,,8- ..
.. [ .
l, -
.. g 4
.g<rn 7 "
pr . . ( c
&F ~ ' f,p<,s SC p l,*: f'i [
j,,. .. J
- N
'f' '
l ?. Q 3'* .. . }/ .- ("
~_
'/ 3
. vQ , .
4 c. s (> >'. xl,
- n. , s.
. ;J. , r t'; n
- q. ,-
( yc j
- 1 3
- .g. - %
..q q- ' .
y r
,/ , .+
s v R iakx- .
%y /
r kck . ! , . Q '~.7 3 r g .,
> t' , s -
T' -- ~ _2,g w . - -- g ,,,,; _. .. . : - - - - , - - - -.- d' =_l'~:
= . .-.-~..
% M-U~i -G. ..c. - u -
2-
~
=~.~5 :: .. --
h-
=_=s= === - - ~~
uIa. .=rr_- w .w :z.r - .-.
- == = = .=. . . _ . . _ _
===--
=: =.=:. .-
J.=."....--~._-**- " L-9
s., . . . . , . . . . . . . . y '.
# % "'Z'.t ';* .T,'"* ,-i.e FIGURE L-8
- . .m- n..
~
' .:.1..i.~~;.- .f
.e . . . .
d~
. ~
7g Vq. y ':: 7 ,gggicg' ' .4 'yCpdTQ'l j . h 'i 'f,
*4 i' -
4 'Y .p :
' : .. g $,,; t. 'y ;, . ^
- / . -
, . . 3 ., - )e'; y
,3 ,.
% r. ;. - .t.; ~
di ,
'r /3 . -: ' t ~ *& ;-; s- :!!
; j'
- g, .-
0
\
+ > W' N..M . ' ' '
- J
*Y Oy'
- b. 4 *:[8'
','fs
'd f l , 5 , _g fd Q s'--*' *'
- i -
. . ., ,. s .I Y ,1p.
-k
- i N ,/ 4 i N. [i
' ) (( [f1'..}Ql,%. , g.;yf.1 '
.i .;
j ., .
.~
;,,. , .s% :y/Q<J ;@f ., 9 .w p v.' '}.
+ ';. a w / -
.m ; .* , b Ei -
k
-c. i . '. g~ - i. .7 k '( -
~* .,q u
/,, I *
.- . B W- i :6 - 7 .(( ..
r, . ,j . . +.- , .
. Y ;. , e , %'V ' '
. ., 4 '*>y.Q i;, a. t l ,f '
o 1 n* f.J }j 5'. .~ ,, '/ y . I J '.h. . s***g'ik*f;6.3
.. .I j ' 4' s. f, . i {.
.j-e
- T "
." : .' % ) . \' ; , ,
}.--ll'
[';:f,\
, By .,.g-f*kW , L % > j( ,
.(.
'k .
*v
- . ]
7 , s
.t,44 ,Fp w f) }
N is 1:3t
-t'. e k'j 4
'4 s
J e'. "
- i. . ll'< . ..;
s) ' ' ,, 8 h
'I ' ' '
(,')]" ' ;l. .. , ;_
,Q
#**. L*- L
;,' '} ,.,6 ,, f , ,,
t w -
': r-,,,*,..7' s
3,; y. . ! g
.,s. '
</ p i ,s 7,
v, u_;}R_i+ n 4
.;- M -
w u ,N s I _q
, es tg i, p 3. .a j-
.,9 Sl I
.II - 3 .1 \q N 4- f.
V- L ,,s
]
# '- kj _g '
r;
, . " ,? .. ,e){
.,; ;. , 4 i J,
r.x s w.w- q. sC n. , ,a ~ ,\ 1.[.ug,x.y ..
$s I k .; d** .
', j
- l. jt ; Y , ,
, l -
-[W. .- / p, ;,d 1,& . . - ,
, t.,. .9.
K . Nl , ,. yj != ' .
=-*, .
,. i d*{
_'!f. -gq._ f _g' . , &' 1 g d' 4:3' ~
//=g.__;_..=._r_
= = = . _ = _=__ . - . . - - C.
m,. _rm m_ 7-
= o.=. - - . - A 5
I
.v.._--*.l*----"--
- .---- : .: . . =., n.w rurvs
= . 1 l"' *N dr5Hi ::= ., %. - "'_.: 2 :"l-~
l I L-10
T
,V . '",, ~; 1 '.**..-',"
- FIGURE L-9
. . _ . . ... . . . .. .. . ... . 2 ~E 22.L ---
i .ij , j, J ; -- * <v
~k;).a\v f &
s .
<t
- <w';z - %.:
C'
~
,: h, np'% h Iu }f.4 e l l' b $* , //-l k $ $ $,l.
,c. Ml ,: k f M>>.
E f (4;I ,y.^W. #') !(<':.a~+,,i ' . y.:.7 ;;y(sa- ; ..
~
y i L.), 4 n b. ,. ,^ .
,\
1 a 3 ,o <
. ' ' .i .' t
,~j,t. t, .L ,x%, ~ ,s :.
o a n
- q 4,,4gg,t. ,p..ys.c.m. .g,, -;,,-
& ;. , .;q ', , . - -., ..
. y. ..,
- f. .
- a,. > .. . ,:. p.. ,,. . .:J,...
. .; tt,-
.q , .
,L n. p >3-$ ..
a-67 l'Ya ;.. ;y . D , . W. -) .\ . [s tu
.f.f,f. .
v..w , . t f ,Lu. . :a.
%,y , , r
- g. y.p, p' i =n 9;2. ,. s,i i .s..,,,.,,,
. 7.
i f d. .e
,, .k i
's[i' it .
; s $.I. 1 t _ _f y s
'{ l D k fN' f- *].
~
'],.QQ ,^
; n'yu.t.h:.v _
- Q k 1 a. 92 M%f& g.jl% % .-hV.?Q!f (Yf.iif'$Q:i<T; MC jf? % .m
$ss[g!h7?P.M;fM.N,\
D]Ml6{4(.f)f::
*!n> l ?ft ( ' -
- f as.- ! *
.? ? */*d*~
J;%f f 5 s i b]4 Q Yh.%, /,-r2 '
. / j' , -
$h. ' ' .e@f5s h* 4 [.. M+ ;$$20%
3 ps , t* g// me ,4. ' 4 ** .- 3
=
e #'sb y ,. 'Ty .- .* r , ~
* ~
Q:,t. :/,w. j eft ~$'.:.
. , , ~
yum\5,<]p(=N
..: ;Y$ YY . .. . : .; = _. n a s .
t'*~ f
- ... ~ . . % ,pg, .. - .
.._-.~~i....J."."I"*
'.*O. ~*"" l*T . - ~ ~
u.s :.~s..
..r. r --
. - - s
. C.E*I
~
':. . ._ .... :. 4. az. . . . . . . . . . .- . . .
~
. [.hb ~~~ * * "
l *******'"*'" eu. , . . ..
- - - - - - *= ""
... ** i.- ; .;--
. 20:".l?:'::'"'7 *::" .. " "" _,*_,
L-11
T N .5"i'E' -- rcunt L-io .. ._ -T.If' C . r y ,3_ i g* - f y =;- . c x . . ,p E-p
.:( I', '-[$l' \* '
1 ., ,
*5 Y=. tN- ?*:. ;q y.9 (. .} g,'
l . , .
- f. ; ' ~ -, . . .
l -QQ \ <, , ' f. ,Me?'r;;w. ,-y Q, e m. o t, s y . rg: : / .- - - . . , . ,
~ 'g, ";#,6- g,@g. le.~_ f i. l. % 95,$.: .
'+=*l~ ' ' ~
f . h. #*, , , -' , f
,W ..' '\f .%h ** .'. ,
. . j .. ,
%- ) % _ _ -
~
o ? spi l,? -
, 1f
+ k (a[)$ < y, e_r Q 'tg.
c.., e
.4-- .
, :g s. =s;n - e-. .
.y ,,
./ w s
, .3 . , ...
y r#,. V
, c. ,s, ' ^: .f ar=;' g , % -
'N r .:me'pg; ,
. . e e
,Y { <Y
^ - -
t e
.' 'Q' . .
g - N. ?;. d.; .- fG%:3 [f j ;Nl-;.;h, O ' , *Y '
. \t . . r Q? 'h ,f* **
g,;f _'s fE ~_ F * , A?g. S i$. yQth Y, sf
.'.y '
'~ ' -* -
- p. l l
.[W, J 6(i.TW$f.s1.qVTih,#e j w
. r- frs , ,'w <- * \ N. 117 *Q
% ,'s $ .; ee.
f'*I. V) ~ a
~
W+1.,#.y8,:{ ,- 3.n,A.'f. d $ % E.! J.) Y.s -5 i ', ,. '. : . 5' '-
- &s?? ; - -
- = ' ' ~' '
.,~ i . 4.~ . -- C 'r - ' ... .~' * '~- ~
/* ',, :.,==.~ "--- ~ ~ s '. -.L.- .. --.
. V:~ .k
;- - ,,J ew ,_.r.2 r . rs - --.
_ - ,,, a , . , , , ~ i i w.- --
"""*** m.- T-- 3 ",."=== d f=eu s * *
- - - - - , _h - . . . .
L-12
, 1 . . " -j l mr -
a
- ,e a ,
l
. s l
. - ~ ., .n 1 /* s'
'e f -;
,.. / <.
l
\ .
~$$}.'.ID',M
@.I FOJRE L- 1, ."'.".*l.l.:.*.".
* ~g.rys
~
7 -- , q .' y -
< - ys ,
, ,s gr[ v. ,, .
.. , 3 .<4" . .,
C "v . k' *i- .c r:.> . '- \ '. qs
.'. _ -s e
E *= }.m t g s' +5 . 3. J. q . , si'ev' ;m w 4 ,. ,
.. . , '.3' W ,
. .A + .
$ - w. . <
t- -&
, I
'p . ..
., .,l,.
1 5 /4 ,g ,v # - -
- w. ,,
,7 #. , , ., . , ,
.; m .y [.-[ "
, ils . .
', , \ *j ,
"~
blf. ., q g , "\j b, * '.' , Q, l.!* \,~ ?! i s .- . e.: s x e
\( x N
S';fi
- 4. g. .
\'
f ' W* j. ~f .. '
\
t - .
.* / ,
/
/ 2_ ' K - < . : :
)l b
- N -
L
. s.
/,,
> i, g m._ ;
, .t u < -
,. / s-. ,,
' y ' _I .
- p'LE
, r . p %
'd . . 4, -
N r / 8 l .. - g
~
-g _ _
~
e t -i - *)'
' a.
m 3 g }- M.
)
r '
\, ., ('.. .,
l* =* q
-QF '
e
~ j, .
?
~.
/;
e . . . .\
. ' .. . /
, , . , 4 - ~ s
- ,w.. w s
.l T' ,g[
,, g
' t L,. 4 5 , ,
,,{j
- p{'t ' 0 j s ,. s h -
.,, ) /
{[ /, ' ' 2
*N, y .')
a ... s
/ .f
% ? e
..r m n
+
'.y b 3
, 3' .
j y
.. e , ?. \ e ,
.,l c .
o W. 't- %
; - . ), .g 8
fn Y U. 'R
. j -
I)> _~ I I P l C.'"*" - y ,
.=......=..j) ?>
- _
J , ,. [
,' / ',,,. l ' o
,s g '-I' .
a ~.. N g8 -
..w : ,
p .-- 76dah..Jt2. -a a _. >W & - - w... ....g -- . -.g. : . ,_ . , ----..-.-7--_.,,.- A* Q_E__.'._*,"
- 1 ~~ T.*/ kZ E .~r - . _ . _ . - .
.r .; ;f- ~
~
__ _ ~ _" " ' ' _ S..s . _.
= = = ' ~~ ~ -~
= . = = - - . _--. :-m &_
~
E.5_ . :2EE--- ==I -- ~=EET"=. ! E-- .__. n ..,
--i7E-. :::.-- . r.r 7.:.. E ._:.-7::E 9 _
-- ;--~ .
- _- r_. . = ._ y= -- - --.: _ , .
. .. _ l L-13 1
I e
"*"*"' "* '"*"* m*rj d. /, -
FIGUmE L-12 . ' * * .*'t
~~
. = .
? .
~~ '=
.y - i
- *]
'~ -
m W -," _ p . prw. qj: Q. u - f.g.'. . j(
.g.w >)
.e, g
. ; ak r
4y y.: e . av , ,
=,j e . .p. 4 k .,x,p; ..
. :g. _ ;
^,
N ' l (,
;.1r
- 8 , _.; s .,
p 1.J/.j..%,,%
'.s L 'w 4f:(jj{g^'m.$fF~d <
'}}