ML20138D262

Testimony of J Nehnevajsa Re Eddleman Contention 57-C-3 Concerning Nighttime Emergency Siren Sys at Facility.Related Correspondence

ML20138D262
Person / Time
Site:	Harris
Issue date:	10/18/1985
From:	Nehnevajsa J AFFILIATION NOT ASSIGNED
To:
Shared Package
ML20138D190	List: ML20138D187 ML20138D197 ML20138D220 ML20138D256 ML20138D262 ML20138D276
References
OL, NUDOCS 8510230225
Download: ML20138D262 (32)

Text

I e

CELATN) COHHtWNW e

\ '

UNITED STATES OF AMERICA / ,

NUCLEAR REGULATORY COMMISSION

[ o

'2-BEFORE THE ATOMIC SAFETY AND LICENSING BOARD g' V g>?g) g J l j']

In the Matter of 1 'N 6V .U CAROLINA POWER-AND LIGHT COMPANY AND NORTH CAROLINA EASTERN PUNICIPAL Docket'Nos. 50-400 OL POWER AGENCY 50-401 OL (Shearon Harris Nuclear Power Plant, Units 1 and 2)

TESTIMONY OF JIRI NEHNEVAJSA REGARDING EDDLEMAN CONTENTION 57-C-3 l Q.1. Dr. Nehnevajsa, where do you live?

A.I. I reside at 1520 Ingomar Road, Pittsburgh, Pennsylvania, 15237.

j Q.2. What advanced degrees do you hold?

A.2. I have a PhD Jegree in Sociology, with minors in mathematical logic and journalism, from the University of Zurich, Switzerland (1953).

Q.3. What teaching positions have you held?

A.3. I have served as Professor of Sociology at the University of ,

Pittsburgh (1961 - to date), was Chairman of the Department of Sociology (1962'-1966), Director of the Research Office of Sociology *

(1961-1967), Director of the University Center for Social and Urban Research, and have held a joint appointment as Professor of Social and Economic Development in the Graduate School of_ Public an'd International Affairs. I was an Assistant Professor of Sociology at Columbia University (1956-1961) and at the' University of Colorado O'510230225 851018 PDR ADOCK 0

I .

9 (1952-1956) where I also served first as an Instructor in Sociology (1951-1952). As a Fellow of the Social Science Research Council (1953-1954), I was at the University of Washington, Seattle, Washington, and at Harvard University.

I was a visiting (Fulbright) professor at the Universities of Heidelberg and Mannheim in Germany (1967-1968), a visiting professor and acting Dean (of the Faculty of Social Science and Comerce) at the Chinese University of Hong Kong (1970-1971) and visiting professor of the Chinese Academy of Social Sciences, Beijing, China (1980 and again in 1981). I served as a visiting scientist (senior policy analyst and program associate) at the National Science Foundation in the Division of Policy Research and Analysis, my work there having focused on technology assessment and risk analysis i

(Jan 1980 - Dec 1981).

Q.4. What research have you done which is relevant to Eddleman Contention 57-C-3?  ;

A.4. For the past twenty-five years or so, I have conducted research on emergency preparedness problems, for a significant portion of this [

period dealing primarily with the possible threat of nuclear war, f and its implications for our people. For a number of years, this work has been expanded to the consideration of attitudes and behavior related to any mass emergency, be it a natural disaster or a particular technological hazard. In the past few years, I have served as principal investigator in a program to develop a major i

1

data resource having to do with existing findings as they are erpressed in the disaster research literature and also generating a data base of original surveys and other studies that have been undertaken over the years by various researchers on attitudes and behavior in anticipation of, preparedness for, and response to emergencies that do, or might, affect some significant segment of the nation's body politic if not all of us.

Included in these data banks are also numerous studies which were undertaken with respect to public views concerning nuclear power plants that bear on perceptions of. risk, benefits, acceptability and the like. Most of the empirical research that was carried out in the aftermath of the Three Mile Island incident is, of course, also incorporated into the data resource not merely in the form of reported findings but also in terms of the original records of responses by the individuals who were sampled and questioned.

Q.5. Do you have any other pertinent experience?

A.S. I served as chairman of the section of public acceptance problems in Project Harbor (1964) which examined, under the overall Chairmanship of Professor Eugene Wigner, a Nobel prize laureate, at the request of the Department of Defense, the state of national civil defense preparedness. I served as a member of the civil defense committee of the National Academy of Sciences (1966-1969) and have been an associate member of the Federal Emergency Manage-

, ment Agency's Advisory Board for the past several years. I also t

L

serve on the editorial board of the International Journal of Emergency and Disaster Medicine, and was a co-editor of the Journal of Mass Emergencies (which ceased publication a few yea s ago). ,

Q.6. What work have you been asked to.do in connection with Eddlemen Contention 57-C-3?

A.6. On a consulting basis, I have been asked by International Energy Associates Limited to address some of the key issues which arise in th'e context of behavioral implications of possible nighttime emergencies at the Shearon Harris Nuclear Power Plant.

Q.7. What materials did you consider in doing this work?

A.7. I had access to, and reviewed, all documents that my colleagues, Dr. Van Lee and Dr. Kryter, perused and I also participated in several conferences in Washington D.C. pertaining to the issues at hand. Furthermore, I have been able to study the drafts of Dr. Kryter's testimony, since some of the major results he points to become an essential input into the kinds of conclusions I have reached and which I will present here today.

As a baseline, I use data on the Shearon Harris Emergency Planning Zone as provided by the Donnelley Marketing Information Services in an "American Profile" printout dated June 26, 1985 which gives some basic socio-demographic statistics for the approximate area in question in tabular form.

I <

. There is no reason to assume that these data, for instance on the exact numbers of households or the age composition of residents would be exactly right. However, based on Bureau of the Census materials as these data are, these are by far the best estimates possible, and only dramatic changes, none of which appear to have occurred around Shearon Harris, could render my conclusions sensi-tive to differences. I am entirely confident that such differences-as might exist between the current situation in the area and that to which the Donne 11ey materials refer, or the changes which may have taken place, in no way affect the major findings discussed here.

INTRODUCTION Q.8. Dr. Nehnevajsa, how would you summarize your conclusions?

A.8. I open the substantive part of my testimony with three altogether important statements.

1. Between 1 AM and 6 AM, not all people are asleep. Some are at work. Some are watching. television, listening to the radio, some are in transit, some are visiting neighbors and so on (see Table 3).

i

2. An immediate cascading, or intrafamily " networking" process starts for those who are aroused by an alerting signal: this means, that an alerted member of the household will alert the l

l

remaining members of the household when there exists a credible threat of any kind.

3. Many people will contact, or attempt to contact, officials, relatives, neighbors'and friends, and in some instances such contacts will arouse and alert some persons, and through them households, previously not aroused.

I therefore conclude that:

- People who are not asleep do not need to be aroused so they present no special or additional nighttime alerting problem.

People who are already awake and become alerted, as well as people who will be aroused by the siren system will actually alert all other members of the household. It is not plausible that they would somehow ascertain a danger to their family and their community and let the other members of the household who may not have been aroused initially simply keep on sleeping.

No human logic could justify such a conclusion.

- Those who are awake and not at home, whether at work or for whatever other reasons will certainly seek to contact their household / family members by trying to get home as fast as possible and/or phoning, thus alerting those household members who may be at home asleep.

- Once aroused, the desire to confirm the warning along with the desire to be sure that relatives, friends and neighbors know that something serious may be " amiss", will cause people to attempt to get hold of others - even were they only seeking confirmation of the warning message, if they contact someone previously not aroused, the interaction will have the effect of arousing, and alerting, such others. This is the broader, and so well documented, cascading or networking process which is triggered by any initial warning that is interpreted, in fact, as a warning.

Some people, indeed, will attempt more than one such contact, whether in person (of the knock-on-the-door variety) or by phone.

- Such intrafamily cascading (alerting other _ household members) will occur immediately once a threat is perceived ("immediately" means with a time delay of perhaps not more than a minute or so).

Such social cascading or networking, that is efforts to reach officials or others, will also occur very rapidly, most generally right after the other household members _will have been aroused and alerted - hardly anyone should be expected to wait and sit around in the face of a threat if they seek to either confirm the nature of the evolving emergency or to share views with others as to what is to be done (or seek official advice on how to act) or, for that matter, to make sure that relatives, friends and neighbors'in the area are also aware of the danger.

i i

l

l I

Q.9. In general, what role does intrafamily social networking play in alerting the public of an emergency?

A.9. I am entirely certain that the intrafamily networking process has to be fully recognized as part of the direct warning even though it must of necessity rely on interpersonal communication within the household itself.

This is so simply because no one can assume that a 2-year-old or a 6-year-old or even maybe a 10-year-old, if awakened from sleep by the alerting signal, could properly interpret the meaning of the signal and thus recognize that a threatening situation /an emergency is in the making. Thus any concept of " direct" warning message delivery simply cannot be grounded in the premise that every single human being, or almost all of them, can or must be warned by the signal system. I repeat; at least children up to a certain age cannot be alerted in any other manner than by some adult who will have been aroused and alerted.

If such intrafamily cascading is, as I think it must be, an integral aspect of an alerting system, then it is not unreasonable to suggest that the social networking and cascading system, too, ought to be viewed as the equivalent of " direct" warning message delivery.

ASSUPPTIONS Q.10. What assumptions have you made in conducting your study?

l

q A.10. In presenting some basic results, in terms of the implications of the three processes specified previously (that is, that some people remain awake during the_ nighttime period, that particular members of a household, if alerted themselves, will arouse and alert all other members of the household and that social networking will also, to some extent, take place), I make several important assumptions which need to be made quite explicit at the very outset.

I assume that the Donnelley Census-based data provide a good approximation to the distribution of households, by size, in the EPZ of Shearon Harris. This distribution is shown in Table 1 of the Appendix.

I assume that the distribution of households by peak dBC signal delivery throughout the EPZ is that which is indicated in Table 2 of the Appendix. This distribution is one which was, to my knowledge, developed by International Energy Associates Limited.

I assume, implicitly, Dr. Kryter's specifications regarding the

" behavior" of sirens over a 15 minute interval, and his specifica-tions regarding the kinds of sirens that are in place, i

I take, as initial parametric values, Dr. Kryter's estimates of arousal probabilities for different sound levels and for different magnitudes of attenuation. My Table 4, which shows this distribu-tion, was derived from Figure 7A of Dr. Kryter's testimony.

I assume that only those who are 18 years of age and older serve as " message recipients" or " message targets" - in that they are individuals who can not only " hear" an alerting signal (if aroused) but also interpret its meaning (and thus trigger the intrafamily aYousal-alerting cascade). This is a very conservative assumption because it would seem altogether reasonable to set the " age" of such primary message recipients at "16" or even "13+". But since Dr. Kryter's probability estimates deal with only those who are 18 years of age and older, and since he also asserts from his expert knowledge that younger people are harder to arouse than are others, I use this "18 years or older" benchmark. But if we were to admit to the' reasonable proposition that, say, 16-year-olds are old enough to recognize a warning, and even if fewer of them than in other (older) age brackets might be initially aroused (for which I do not have any probabilistic estimate on hand), it follows that all results I present here are slanted toward the minimum estimates

, of what would actually happen rather than in some more optimistic direction.

I further assume that the Shearon Harris area people are not very

~different from other American households and families. I thus use national data (from my own research in a probability sample in late 1978) to estimate the composition of the EPZ households with respect to age: thus I use national data to give a good approximation to

! households of a given size in which there are 0,1, 2, 3, or more l

l

m -- _.

- II -

" adult's" (those who are 18 years of age and older by the prior assumption). This distribution is shown as Table 5 in the Appendix.

I have also assumed, in the absence of more detailed information which would not alter the findings in any significant manner, that the households by size-and-composition are proportionately distributed-in the various " zones" defined by the peak dBC's.

Table 6 in the Appendix shows this distribution.

EFFECT OF THOSE WHO ARE AWAKE Q.11. You stated earlier that some people are awake at this time of night. What effect does this have?

A.11. Based on a national sample of the University of Michigan study which is used here to provide estimates for the Shearon-Harris area on the premise that these residents are not dramatically different from other Americans, we note the following:

Hours of night Probability of being awake Midnight to 2 AM .092 2 AM to 4 AM .058 4 AM to 6 AM .058 This, of course, also means that about .908 represents the proba-bility of being asleep between midnight and 2 AM, and .942 the likelihood of being asleep between the hours 2 to 4 AM and 4 to 6 AM.

I l It is not unreasonable to postulate that these basic patterns would be descriptive of the behavior of individuals in households regard-less of size or composition.

Now by my prior premises, only households where everyone is asleep require special attention as far as nighttime alerting is concerned.

With a focus on households as well as implied numbers of persons, let the basic result be highlighted here by showing the percentages of households and persons respectively likely to be awake during specified nighttime intervals anyway, and thus not being, in principle, part of the nighttime alerting arousal " problem".

Hours of the night: Percent with at least one adult awake so that household " arousal" by sirens is not required.

Midnight to 2 AM: 17.0 percent of households 18.7 percent of people 2 AM to 4 AM 11.0 percent of household 12.2 percent of people 4 AM to 6 AM Same as between 2 AM and 4 AM.

In terms of the issue under consideration here, this result is of l

utmost importance. It means that to achieve alert notification for

l' essentially all households in the area, the problem of sleep arousal is salient for about 90 percent of households during the " worst" hours of the night, after 2 AM, and such a notification, by sirens or whatever means, is a problem with regard to some 88 percent of the total population.

This robust conclusion is driven by a very simple assumption which I would find it most difficult to disregard or view as lacking in face validity: it assumes that people who are awake, the adult recipients of an alerting message (those who are 18 years of age and older), would alert the remaining members of their family /-

household and that they would do so even if the warning " reaches them" away from home (at work or otherwise).

Only very few people are "in transit" during the particular periods considered: the probability is only .006 between midnight and 2 AM, it drops to .002 between 2 AM and 4 AM, to increase again to .004 between 4 and 6 AM. This small fuzziness in the interpretation of the data can certainly be ignored for any and all practical purposes.

I mention this explicitly since it could be properly argued that people in transit are unlikely, or unable, to alert members of the at-home family / household.

AROUSAL ESTIMATES Q.12. What do Dr. Kryter's and Dr. Lee's findings imply?

A.12. According to the analyses of Drs. Van Lee and Karl Kryter, if all were to sleep with bedroom windows open,the expected average attenuation would be of the order of 15 dB.

If half of the residents kept their windows open, the attenuation would remain 15 dB for'these households, but would turn out to be about 27.5 dB for the other half of households whose windows are closed.

If all were to sleep with their bedroom windows closed, the expected attenuation would be about 27.5 dB for all households.

With the arousal probabilities estimated by Dr. Kryter for one-person households, and applying them to households varying in size and with a variable number of household members who are 18 years of age and older (the primary recipients of an alerting message) a summary of the basic results for an alerting signal replicated 3 times during a l 15-minute-interval (with 3 minutes "on" and 3 minutes "off") are of the following kind:

Siren Time Percent aroused or awake Attenuation Activation Period Households People All with windows First 12-2 73.9 78.4 open 2-6 72.6 76.9 Second 12-2 77.9 81.3-i 2-6 .75.5 79.8 l

l Third 12-2 78.9 83.2 l 2-6 7/.6 81.8 L

Half with First 12-2 66.6 71.4 3

windows open 2-6 64.5 69.0 Second 12-2 71.2 74.7 '

2-6 69.3 72.8 Third 12-2 72.3 76.7 '

2-6 70.5 75.0 All with windows First 12-2 59.3 62.8 closed 2-6 55.4 59.8 Second 12-2 62.2 66.8 2-6 59.9 64.4 Third 12-2 64.7 69.3 [

2-6 62.6 66.8 l

Q.13. What other research do you consider pertinent?

A.13. In a June,1985, Pittsburgh area survey in the aftermath of the tornado experience, some 83.6 percent of the respondents said that i

they typically sleep with their bedroom windows open. If this is also approximately-so in the Shearon Harris area, the typical arousal rate expectation would be closer to the estimates based on I the 15dB attenuation than to the results predicated on 50% of  ;

1 i l windows being open. I Q.14- What did you take into account in generating your results?

A.14. In generating these results, I took into account an estimated I distribution of people who are awake during specific nighttime

[

periods so that the " arousal requirement" was evaluated for that I j part of the household population that can be expected to be asleep  ;

f rather than otherwise active.

I l

t _

I took into account household sizes as well as, for each household size, the number of likely individuals who are at least 18 years of age beyond 2 person households - that is, three person households and those with four or more individuals, the latter averaging 4.9 persons per household, with an average of 2.42 individuals 18 years of age and older.

d The estimated arousal rates for people rather than households yield consistently somewhat higher percentages than do the results for percentages of households: this is as it should be once it is essumed, as I have done, that any primary message recipient, any individual 18 or more years of age, if aroused, will alert the remaining members of the household. In other words: the intrafamily/ '

intrahousehold cascading effect is explicitly taken into account in my estimates.

EFFECT OF SOCIAL NETWORKING Q.15. What effect does social networking have on the number of people alerted?

A.15. The literature abounds with data pertaining to the fact that many I

people, if alerted to a danger, will seek to contact others either in person or by phone, with the latter being the most common form of reaction. .

In the Pittsburgh arer study, we find, for instance, that ,

i I

I

\

87.5 percent of these respondents expect that people in their area wculd be contacting others to make them aware of an impending danger, and just as many respondents claim that they wculd expect that someone would try to contact them under such circumstances.

Thus I am likely to err on the most conservative side if I assume, as I have done, that only half of the alerted households would contact others, whether to confirm that an emergency is evolving or to warn them. In effect then, any contact with households not aroused by the alerting signal itself will amount to alert notifi-cation, whatever the reason may have been for initiating the contact on the part of the contact initiator.

To assess the implications for this process of interfamily networking, I focus on a realistic though conservative situation. I assume:

Half of the people sleep with all windows closed, and half with at least one window in at least one bedroom open, so that the attenuation is 15 dB for half of the residents and 27.5 dB for the remaining hal f.

An alerting requirement between 2 AM and 6 AM, is the " worst case" with respect to the proportions of people being asleep.

Since the nonarousal proportion is .355 for this situation once the initial warning will have sounded, the probability is .232 that a household making a contact (whether on a door-to-door basis or by phone) with some other household would " reach" some family not yet aroused: and this, of course, if all awake and aroused households made an effort at one such contact.

As I have stated: I assume that only one half of such households might do so, a conservative premise in the light of such data as those from the most recent Pittsburgh area survey.

This then implies that the probability of an alerted household actually making contact with an unalerted household is about .116.

It is important to note, even though this is rather self-evident, that when it comes to the effects of interfamily networking, it does not matter in which subarea of the EPZ people live with respect to peak signal strength or whether or not windows are open or closed. It only matters that one member of the household,18 years of age and older, be awakened by the ringing of the telephone or, for that matter, by neighbors rir.ging the doorbell or knocking on the door.

The 1985 Pittsburgh area study shows that 93.0 percent of the.

respondents asserted that they, or at least someone in the house-hold, would be awakened by the telephone were it to ring during i

I . .

7 nighttime hours. This may be because a few people are simply such deep sleepers that they would not be awakened by the phone, or because a few might even disconnect the phone, unplug it, at night.

Using this probability of .93 (of being aroused by the phone) as an additional conservative factor to degrade the 116 probability of an unarcused household being aroused due to the interfamily networking pattern (under the assumptions already specified), I end up with .108 as the likelihood instead of .116.

Under these simple premises, 75.3 percent of the households would end up being aroused before the second sounding of the sirens rather than 64.5 percent, and 79.8 percent of people, rather than 69.0 percent of them, would be aroused and alerted.

But this, of course, has immediate and direct implications.for the i

l arousal outcome as a consequence of the second, or subsequent, sounding of the sirens. There are, to be sure, fewer people in ner d of arousal given interfamily~ networking, as rather modestly assumed here with regard to its magnitude, than there would be were it not for this empirically well established pattern of behavior.

l

Given the arousal probabilities upon the second sounding of the sirens, the results now show that 77.8 percent of households would be alerted, and 82.2 percent of residents will have become alerted.

This, of course, compares favorably with the 68.3 percent of house-holds and 72.8 percent of residents who would be alerted by the end of the second 3-minute warning signal if there were no interfamily cascading of the alerting message.

It is reasonable to argue that some people may make more than one phonecall or otherwise seek to contact other residents of the area.

Suppose only one third of them were to do so, rather than one half as had been postulated for the initial warning. This means, of course, that some people may call more than once, while others, the newly aroused residents, will perhaps make only one contact. In any event: it is assumed that the " average" contact probability will have declined from .5 to .33 (though an increased probability could be theoretically rather well justified).

Then again: if a telephone has a .93 chance of arousing someone in a sleeping household, the overall likelihood is .053 that an already alerted family will alert a previously unaroused household somewhere I

in the period after-the onset of the second warning signal and before the initiation of the third replication.

This reduces the pool of unaroused households that need to be alerted by the third signal replication.

Q.16. What conclusions have you drawn?

A.16. The result then, after three signals, is as follows:

83.6 percent of households will have been aroused and alerted, and 87.8 percent of the individuals in the EPZ will have become aware of the impending emergency.-

A

SUMMARY

ACCOUNT

~Q.17. Explain how you arrived at these conclusions.

A.17. The findings, in a cumulative manner, incorporated three major considerations:

The effect of nighttime activities, so that some percentage of.

people are awake Between midnight and 2 AM, Between 2 AM and 4 AM, and i

L_

.= _ _ -

22 - )

Between 4 AM and 6 AM, this latter percentage being essentially the same as that for activities between 2 AM and 4 AM so that [

~

no separate evaluation is necessary, and the results are i I

applicable, as an estimate, to the whole period between 2 AM i and 6 AM.) }

L i

The effect of intrafamily networking given household size and. j j

composition, household distribution by peak dBC estimates, and j arousal probabilities for attenuations of 15 dB, an average of 21.25 dB, and 27.5 dB and for midnight to 2 AM and 2 AM to 6 AM periods.  !

t The effect of of social, interfamily networking, assuming that t

about half of those awake or aroused by the. initial' alerting signal will make one single contact with some other household so that some ,

of these contact arouse and alert households, and people in them, that may not have been alerted by the sirens and in which no member ,

i i

is awake during .the period in which the alerting signals are sounded.

i \

And, furthermore, that one third of the people who will have been i aroused by the end of the second siren. signal will attempt to contact others . ,

Finally, that such attempts at contacts have a .93 probability of success, that'is, in actually awakening someone in a-household i where all members,18 years of age and older, remain unaroused. j f

, -. , . .e - - - , . - - - - - ,

. ,.n-,._- , - -

. i I

The interfamily ripple effect was applied here only to a s'ituation t

in which half of the residents sleep with all bedroom windows

{

closed, and half keep at least one window (in at least one bedroom) open, that is, for attenuations of 15 dB for 50 percent of the i households,and 27.5 dB for the other half of the households, s

The " worst" night hours were assumed, between 2 AM and 6 AM, that l is hours when the potential alerting " payoff" due to people who are awake anyway is at a minimum. h t

i Some 87.8 percent of the people in the EPZ can then be expected to i be aroused and alertea.

Quite a few assumptions had to be made, as is always the case when one has to apply more general bodies of insight to a concrete, but not'in every detail investigated, situation. But these were assumptions which are mistaken, if at all, on the side of caution. l r

l l

One of the key conservative limitations of the results, perhaps, l needs to be repeated because relaxing this particular assumption would-actually increase all percentages of expected alerting system  ;

effectiveness. It has to do with the fact that only individuals 18 years of age and older are considered as potential sources of the intrafamily cascade.

l  ;

l

l This is not merely conservative, but somewhat unreasonable: certainly, young people who are 13 years of age and older would be, for the most part, in a position to interpret an alerting message, if aroused, so as to make sure that other family members are awakened as well.

True enough, the arousal probabilities for younger people are lower than they are for the age-groups (18 and over) explicitly taken l into account in this analysis. But even relatively small probabili-

~1 ties of arousal of those who are in the 13-18 age bracket would have highly relevant effects on the overall patterns of arousal and alerting, so that more households, and even more persons as such would end up warned after each subsequent emission of the alerting signal.

I i

i I

i l

1

Table 1 Distribution of Households by Size

[Shearon Harris EPZ]

Household Size Percent

• One person 18.5 Two persons 29.5 Three persons 19.6 Four or more persons 32.4 Total households 6,884**

• Donnelley Marketing Information Services data, American Profile, June 26, 1985.

• • Number of households based on IEAL count of residences in the 10 mile EPZ. The Donnelley percentage distribution by size assumed to be the same for the 10 mile estimate. The Donnelley data involved 23,161 persons in the households. Adjusted for the 6,884 households

'in the EPZ, 20,339 persons were estimated, with an average of 4.9 in households with four or more persons.

i i

l s-1 I

I 1

Table 2 Distribution of Households by Peak dBC Peak dBC Households Percent 105 511 7.4 95 1,049 15.2 85 2,314 33.6 75 2,742 39.9 65 268- 3.9 TOTAL 6,884 l

1 L_

Table 3 Some hightime Activities Activity Midnight to 2 2 to 4 4 to 6 Asleep 90.8 94.2 94.2 At work 2.0 1.7 1.8 TV, radio. 2.5 0.4 0.2 In transit 0.6 0.2 0.2 All other 4.1 3.5 3.6 Data from University of Michigan study, "1975-1981 Time Use Logitudinal Study" (F. Thomas Juster and others). Alsc, Norman-3 P. Hummon, Linda Mauro and George Rogers, " Time Budget Analysis and i Risk Management", University of Pittsburgh, August, 1985. '

The original data were obtained in minutes for each two hour interval for each of the activities. Transformed here into probabilities = minutes per activity /120. The Table above, of course, presents percentages.

I

/

e 4

i l

l

l r

Table 4 i Individual Arousal Probabilities First Activation f

House attenuation Indoor DBA SEL EPZ Population l Average i 65 dBC Outdoor Sound level  !

t 15 dB 60.79 .25  ;

i 15-27.5 dB 54.54 .16 27.5 dB 48.29 .08  !

75 dBC Outdoor Sound Level f 15 dB 70.79. .38 [

15-27.5 dB 64.54 .30 -

27.5 dB- 58.29 .21 -

85 dBC Outdoor Sound Level ,

15 dB 80.89 .52 15-27.5 dB 74.64 .43 i 27.5 dB 68.39 .35  :

95 dBC Cutdoor Sound Level f 3 i 15 dB 91.19 .65  ;

15-27.5' 84.94 .57 t 27.5 dB 78.69 .49 i

105 dBC Outdoor Sound Level  !

15.dB 101.39 .79 15-27.5 dB 95.14 .71 27.5 dB 83.89 .62  !

f f

I l

Table 4 (Continued)

First and Second Activations House attenuation Indoor DBA SEL EPZ f Population Average 65 'BC d Outdoor Sound Level 15 dB 62.80 .29 15-27.5 dB 57.55 .20 27.5 dB 51.30 .12 75 dBC Outdoor-Sound Level 15 dB 73.80 .42 15-27.d dB 67.55 .34 27.5 dB 61.30 .25 85~dBC Outdoor Sound Level 15 dB 83.90 .56 15-27.5 dB 77.65 .47 27.5 dB 71.40 .39 95 dBC Outdnor Sound Level 15 dB 94.2 .69 15-27.5 dB 87.95 .61 27.5 dB 81.70 .53 105 dBC Outdoor Sound Level 15 dB 104.40 .83 15-27.5 dB 98.15 .74 27.5 dB 91.90 .66 L

Table 4 (Continued)

I All Three Activatio'ns House attenuation _ Indoor DBA SEL EPZ f Population Average 65 dBC Outdoor Sound Level 15 dB 65.56 .31 15'-27.5 dB 59.31 .23 27.5 dB 53.06 .14 75 dBC Outdoor Sound Level 15 dB 75.56 .45 15-27.5 dB 69.31 .36 27.5 dB 63.06 .28 ,

85 dBC Outdoor Sound Level 15 dB 85.66 .58 15-27.5 dB 79.41 .50 f 27.5 dB 73.16 .41 95 dBC Outdoor Sound Level 15 dB- 95.96 .72 15-27.5 dB 89.71 .63 27.5 dB 83.46 .55 105 dBC Outdoor Sound Level 15 dB 106.16 .85 15-27.5 dB 99.91 .77 27.5 dB 93.66 .67 t

i I

y w - g - . , , , ' - . ' -+

• 4 4 --+ ,.

Table 5 Household Size and Composition Household size 18 and Older Percent One person One 18.5 Two persons Two 26.9 One 1.8 None 0.8 Three persons Three 6.1 Two 11.3 One 2.0 None 0.2 Four or more 2.42 average 32.4 Data based on a national probability sample of late 1978 (1,620 respondents in 48 contiguous states, 18 years of age and older, noninstitutionalized population). The research, sponsored by the then Office of Civil Defense, was directed by this-author.

With 4.9 being the average number of persons in households with four or more people, the average of those who are at least 18 years of age was 2.42. This was used in estimating arousal probabilities for these larger households (if only those 18 and older are considered

" primary message recipients" and thus. capable of triggering the alert:ng intrafamily cascade.

I l

\

O Table 6 Distribution by Size and Composition and by Peak dBC's Size-Composition

• 105 95 85 75 65 All One-One 95 194 428 507 50 1274 Two-Two 137 282 622 738 72 1852

-0ne 9 18 42 49 5 124 ,

-None 4 8 19 22 2 55 Three-Three 31 64 141 167 16 419 l

-Two 58 119 261 310 30 778

-One 10 21 46 56 5 138

-None~ 1 2 5 5 1 14 Four-2.42 average 166 340 750 888 87 2230 Totals 511 1049 2314 2742 268 6884

• As in the preceding Table. " Composition" refers to numbers of individuals within each household size who are 18 and older.

l i

i

--..,,_~.w ,-.+rr , ,-y e y<