ML20235D751
| ML20235D751 | |
| Person / Time | |
|---|---|
| Site: | Shoreham File:Long Island Lighting Company icon.png |
| Issue date: | 07/08/1987 |
| From: | Bolton P, Flynn C, Lindell M, Martin J, Stoetzel G BATTELLE HUMAN AFFAIRS RESEARCH CENTERS, Battelle Memorial Institute, PACIFIC NORTHWEST NATION, SOCIAL IMPACT RESEARCH, INC. |
| To: | |
| References | |
| AIF-NESP-031, AIF-NESP-31, OL-3-A-005, OL-3-A-5, NUDOCS 8709250328 | |
| Download: ML20235D751 (25) | |
Text
Ljfgg
\\
National
//g AIF/NESP-031 Environmental Studies a s a z oc 3
' '? "
Project 7/ffg7
'87 SEP 17 All :41 ff - 5 A.s.
s l
Z Planning C.cncepts and Decision;. Criteria for Sheltering and Evacuation in a Nuclear Power Plant-i Emergency l
'l
/
4
/
!E E
?
?f i!
!$0'"ES$c"!$$000!
O PD
]
Atomic industrial Forum,Inc.
I ~
__ f eI
RJ i
e I
4 i
Og 1
' 'b 7
u,ss,ge\\%
ys..
- $, f4
- gS
AIF/NESP TASK F@RCE Peter A. Moeller Melinda S. Renner Public Service Electric & Gas Company Task Force Chairman Atomic Industrial Forum, Inc.
Task Force Secretary Robert J. Catlin Electric Power Research Institute Theodore J. Myers The Toledo Edison Company Lawrence B. Czech New York State Department of Health Charles Raybum Radiological Emergency Preparedness Group B.altimore Gas & Electric Company John C. Golden Reginald C. Rodgers Northeast Utilities Commonwealth Edison Company Larry A Mill David S. Smith Fiorica Power Corportion institute of Nuclear Power Operations Donald F Knuth Robert Touchton
.<MC, ;nc.
Westinghouse Electric Corporation Barbara Mohrman Edward V. Waage Stone & Webster Engineering Corporation Pacific Gas and Electric Company Charles R. Wike, Jr.
Pennsylvania Power & Light Company toert Jaske i
U S. Federal Emergency Management Agency Liaison to Task Force Stephen A.McGuire U.S. Nuclear Regulatory Commission Liaison to Task Force Richard Van Niel U.S. Nuclear Regulatory Commission Liaison to Task Force
AIF/NESP-031 1
PLANNING CONCEPTS AND DECISION CRITERIA FOR SHELTERING AND EVACUATION IN A NUCLEAR POWER PLANT EMERGENCY 1
l Prepared for the National Environmental Studies Project of the Atomic Industrial Forum, Inc.
I i
by 1
1 1
BATTELLE HUMAN AFFAIRS RESEARCH CENTERS Seattle, Washington l
Michael K. Lindell Patricia A. Bolton l
Ronald W. Perry 1
\\
and BATTELLE PACIFIC NORTHWEST LABORATORIES I
Richland, Washington '
Gregory A. Stoetzel Jerome B. Martin l
and SO CIAL IMPACT RESEARCH, INC.
Seattle, Washington
~ Cynthia B. Flynn June 1985 W--_-_-_.-
t..
)
I NOTICE I
This report was prepared as an account of work sponsored by the National Environmental Studies Project
. (NESP) of the Atomic Industrial Forum, Inc. Neither the Atomic industrial Forum, Inc., nor any of its employees, members, or consultants makes any warranty, expressed or implied, or assumes legal liability or responsibility -
for the accuracy, completeness or usefulness of any information, apparatus, product or process disclosed, or.
represents that its use would not infringe privately owned rights. -
The opinions, conclusions, and recommendations set forth in this report are those of the authors and do not necessarily represent the views of the Atomic Industrial Forum, Inc., its employees, members, or consultants.
l 1
Because NESP is supported in part by Federal funds, the following notice is required by Federal regulations:
The Atomic Industrial Forum's NESP activities are subject to Title VI of the Civil Rights Act of 1964, which prohibits discrimination based on race, color, or national origin. Written complaints of exclusion, denial of '
i benefits, or other discrimination on those bases under this program may be filed with (among others) the Ten. ;
.'l j
nessee Valley Authority (TVA), Office of EEO,400 West Summit Hill Drive, Knoxville, TN 37902, and must be filednot later then 90 days from the date of the silegeddiscrimination. Applicable TVA regulations annear in part 302 of Title 18, Code of Federal Regulations. Copies of the regulations, or further information, may be -
.j obtained from the above address on request.
.)
i A. seett Los er, NESP Protect Menseer e
Melinda Renner, Manager. Specol NESP Projects susanne Rk heedeon, NESP Secretary slien sweeney, Staff Artet Copyright @ 1985 by Atomic industrial Forum,Inc.
7101 Wisconsin Avenue Bethesda, MD 20814-4805 sas.oe seeneers/s7s.oe asen..,eneer.
All rights reserved.
r
.___.__m m _._ _ _
PREFACE' Emergency planning, particularly associated with technological accidents, has received a great deal of attention in recent years from federal and state agencies, the Congress, state-level and county-level officials, the nuclear power community, and the press. Emergency preparedness is presently one of the principal areas of corporate concern for certain nuclear utilities as well as for other in-dustrial segments of the nation's economy. The commitment of nuclear utility licensees to protect -
the health and safety of the public, coupled with increased regulatory activity in emergency planning, has promoted greater involvement between licensees and the federal sector, represented primarily by NRC and FEMA. The publication of NUREG-0396 (1978) and Nt; REG 0654 (1980) and the promulgation of Appendix E of 10 CFR 50 (1980; amended 1981) attest to increased federalinvolvement in emergency planning regulation.
Under current regulations, the management of a nuclear power plant.is responsible for notifying cognizant "offsite" personnel (county, state and federal officials and agencies responsible for coping with emergencies) in the event of an accident "onsite" (within the boundaries of the plant site) which has a potential for affecting the public. Emergency plans based upon 10 CFR 50, Apk pendix E. already exist for all operating and soon-to-operate nuclear power plants; these plans un-dergo testing periodically through exercises and drills that point out areas of exemplary perfor-mance and areas needing additional attention, whether onsite or offsite. Auch exercises are cooperative efforts calling for responsible actions on the part of local, county, state, federal and '
utility personnal. When all of these elements work well together, exercises have been shown to be an effective means of building mutual confidence on the part of allinvolved. Further, successful ex-ercises help instill in the public a belief that their health and safety are being protected.
in 1983 the National Environmental Studies Project (NESP) committed to sponsoring a study to offer guidance to offsite emergency planners on how to evaluate the options of sheltering or eva-cuating the public in the event of an acci: lent. Up until now, there has been no document giving advice to offsite decision makers - governors and county executives - and to their technical sup-port staffs (including state or county health officers and emergency planning managers) about.
what to consider when required to make a timely decision regarding which option to choose fol-lowing an emergency at a nuclear power plant. The study first examines each of the options, sheltering and evacuation, individually for its effectiveness historically, its risks, costs, and consequences. The two options are then compared to each other to discem the advantages and disadvantages of selecting one over the other for a given situation at a given time. Much attention has been focused on the merits of evacuating large numbers'of persons from an area where a nuclear power plant accident has occurred. Considerably less attention has been paid to the advan-tages of sheltering some or all of those persons. Virtually no guidance has been forthcoming to familiarize the decision maker with the extenuating and sometimes uncontrollable factors, including.
" political" questions, which enter into his or her decision to publicly announce a protective action recommendation and then ask constituents to take certain actions. More must be considered than purely technicM matters when making a protective action recommendation with potentially far-reaching health, social, economic, and political consequences.
A second purpose in sponsoring a report of this type was to provide planning criteria for onsite and
-1 offsite personnel to use wellin advance of any accident to ensure that appropriate consideration is given to every f acet of planning for protection of the public. The questions posed in the checklists in Chapter 2 offor new insight for such planning purposes.
This NESP report does not address th6 probability that any given accident will or will not o cur, nor l
does it address the radiological consequences of such an accident. These topics have bewn exam-i ined over the past several years by NRC, DOE's national laboratories, and the nuclear industry 4
i
j under the auspices of IDCOR (the Industry Degraded Core Rulemaking Program). Findings from re-search about the probability and consequences of severe accidents have been and are being pub-lished in the public domain.
The Task Force listed inside the front cover determined the scope of the contractor's effort and monitored the study's progress. The Task Force Chairman would like to acknowledge the persis-tence of many members without whose hard work and long hours this report would not have come to fruition. The assistance of the NRC and FEMA staffs is appreciated, particularly that of the NRC's Emergency Preparedness Branch Office of Inspection and Enforcement, who gave the final draft a thorough reading ano made many helpful comments. The emergency planning authorities of numerous states, including Maryland, New Jersey, Delaware, Connecticut, Wisconsin, Nebraska, New York, Massachusetts, Illinois, Pennsylvania and Washington reviewed early drafts and, in a i
number of instances, submitted fresh ideas regarding the arrangement, tone, and applicability of
{
the final report. Also, some industry personnel not formally part of the Task Force contributed will-ingly of their time and should be thanked for their help. A working group of the AIF Subcommittee on Emergency Preparedness and Siting chaired by Johnny Elliott of Duke Power Company followed the study and acted as a peer review panel for drafts, giving guidance on many technical points.
Thanks are due Charles Daverio of Long Island Lighting Company for offering many constructive
- comments based on practical knowledge gained during the creation of LlLCO's Emergency Plan.-
Tom Tipton and John Siegel of the AIF staff contributed their knowledge of current regulatog -
issues..R. B. Maxwell and his staff at the Tennessee Valley Authority made technical suggestions about the health physics issues and dose reduction factors covered in the report. Finally, Carl Mar-zola of Stone and Webster did a superb job of helping edit the final report and offered his expertise in meteorology ;t certain critical points.
l The Task Force hopes that this report will encourage e// offsite emergency planners to work.
together with utility stafft. for the common good of protecting the health and safety of the public.
Melinda S. Renner Manager, Special NESP Projects National Environmental Studies Project
4.2 Review of Disaster R'esearch 4.2.1 Individual Decision Processes and Actions Research on protective response to emergency warnings has covered.a' range of hazards including floods (Perry, Lindell and Greene,1981),
hurricanes (Baker.11979) and tornadoes (Sims and Bauman,1972), as well as hazardous materials accidents (Burton, et al.1981; Perry, Greene and Mushkatel,1983) and a nuclear. power plant accident (Houts,. Lindell, Hu, Cleary, Tokuhata and Flynn, in press).
There is a common theme across this literature that can be referred to as a " protective action decision model" (Lindell,1984) in which the affected individual seeks a balance -
between two goals: the maintenance of normal activities and the protection.
of persons and property.
In order to maximize protection while minimizin{f disruption, the individual assesses the characteristics.of the hazard and of alternative responses and then uses these assessments.in choosing.a means of coping with the hazard.
According to this model (see Figure 4.1), the relevant characteristics of the hazard are assessed in terms of perceived certainty, severity and innediacy of the impact of the hazard (box 5). When all of these factors are present, evacuation tends to be the resultant decision. When certainty is low, particularly when there appears to be.no immediate threat to persons or property, ti.me and effort will be allocated to monitoring the situation, but evacuacicn is not likely. When the warning recipient considers the threat to have minimal consequences, especially when he does not consider himself to be in the impact area, there is also little likelihood of evacuation.
Possible responses to the hazard are assessed in terms of the perceived efficacy of a protective action in neutralizing.the threat to persons or property (box 6).
The action considered most likely to succeed 4-6
e l
n a
la o
n r
s o
i r yi n o
t v
et idi o
a p r h
ed ta t p a e
e o m
b er k r et p e o d
o s
f e
r r e donP oSI n
ne pe e
s ta e l r 7
n o
s t
f i
n t
c
- o. o a
ot e a c p y ei t
t n
i t
y c c ys e t
mnt i
t a 'c r m
eI i i l
a r e
a e d r e
e d t e
l e e d r ci t
r l eeifvir p s
r r v vt a e e vt t f
i c a m o ea zCS ccEB C
i i
a e ae cr r ah e e cr t ee e
eh f r a o eh r
P co P cp 5
6 e
s l
f e
at e
l f
n c
u n i
r t
e o
e ou c v b
o a
i e
e t
s d c d
c f
f l
n od od e a a r i
e na t
oz e o nt n ne s a l
i r t o c o ev mt rh ea i e i
i t j t
e t
m p o pm c
pt c
u a o p
e r
ir i
r r ot r o r
c cc o f
ob a n s r s e r i
eo er p r a h
I P
C D
D t
4 1
3 I
FI I
l y
l
[
g I
l
- b
1 q
is the action most likely to be implemented. Regrettably, the action i
)
considered most appropriate by experts may not be seen by a threatened individual as the most effective.
Spontaneous evacuation insnediately prior to tornado impact is probably due to the recognition that sheltering in place during the tornado does not guarantee survival. This observation i
also holds true for many victims of fires in high rise buildings who have attempted unsuccessfully to evacuate when sheltering in their rooms would likely have saved their if ves.
Barriers to implementation (box 6) considered by warning recipients include lack of knowledge of a safe place to go, a safe route of travel, or even, in some cases, a means by which they can transpo'rt themselves to safety.
In other instances, the separation of family members is considered to be a barrier to evacuation.
lintil family members can be reunited--or at least until they can establish contact and agree upon a place to meet--evacuation is unlikely to occur.
i Categories of cost (box 6) considered in deciding whether or not to evacuate include out of pocket expenses, such as travel costs and temporary living expenses, lost wages, and the possibil.ity of property losses due to the action of the hazard agent or to looters.
In many cases, the desire to protect property is a major concern that leads individuals to delay their departure until the last possible moment.
The perceiv' d characteristics of the hazard impact and of the e
available protective actions are both affected by four sets of factors.
First, the perceived characteristics of the hazard impact are determined by the individual's prior beliefs about the hazard (box 1).
These, in turn, are modified by the source and content of information about the specific event itself (boxes 2 and 3).
Potential sources of infonnation 4-8
include friends, relatives and neighbors (unofficial sources), police, fire and other emergency personnel (official sources) and the newsmedia.
The most important characteristic of a source is its credibility, which is a function of expertise and trustworthiness. While expertise is characterized by knowledge of correct information about the hazard and an ability to process that information to project future events, trustworthiness refers to a willingness and ability to comunicate that i
information without bias (e.g. Hass,1981).
Perceptions of bias may result from one or the other of two beliefs about the communicator: (a) that he is deliberately withholding information that, if not withheld from the public, would damage his ability to achieve his own goals (unwilling to communicate without bias), or (b) that he is communicating all that he inows, but is not aware of all of the relevant infomation (unable to comunicate without bias). Finally, the content of the messai;e refers to the description of the actual or projected events. When the message I
content is specific, it reduces the recipient's uncertainty by conveying precise information about the severity and inmediacy of impact.
It reduces perceived barriers to implementation by describing a feasible protective action (box 4) and a means by which the action can be implemented (e.g., a destination and a route and mode of travel).
4.2.2.
The Social Context of Individual Decisions Individual protective actions involve receiving and assessing a warning of some kind, preparing to undertake the selected action (preparing to evacuate or shelter), and taking the action. While the protective action model above addresses the decision processes of the individual, both the individual assessments of the warning information and
- 4. 9 li b
l the taking of protective actions have social contexts.
The social context includes the primary groups to which individuals belong and the formal organizational structure responsible for the health and safety of the population. Primary groups consist of those sets of closest relationships which individuals maintain, referring in particular to the innediate t
family unit and to other kin, but also including friends and neighbors.
It is well recognized that community protective organizations are mobilized in the face of community emergencies to fulfill functions of i
J i
warning the population, directing protective actions when necessary, and l
l providing care for temporarily displaced populations.
It is less well l
l recognized, perhaps, that the persons for whom these things are being done j
are not simply responding to the information and guidance being provided by the protective organizations. Social science research on behavior during and after disasters has shown that persons most typically make 1
decisions and take actions as families, not as individuals (Barton,1969; i
Mileti, Drabek and Haas,1975; Drabek and Stephenson,1971; Perry, Lindell i
and Greene,1981; Quarante111,1980).
In the post-disaster setting, relatives, and to some extent, friends will be the major--and pn.ferred--source of temporary shelter, subsistence needs, and various I
other kinds of support for those dislocated by the event (Barton,1969; Mileti, Drabek and Haas,1975; Drabek and Key,1983; Bolin,1982).
1 Protective actions most typically are taken only after the actions of others have been taken into account, and protective actions most typically 1
are undertaken as a group.
With respect to the warning phase, other than environmental cues to the onset of a hazard impact (e.g., funnel clouds in the case of tornadoes), information about the hazard comes primarily from other i
4-10 f
J
i y
persons.
Also of considerable significance--and often overlooked--is the fact that attempts are routinely made to confirm existing information and
{
l to seek additional information from sources other than the one from which
)
a warning was first received.
If a message was originally received from an informal source such as a friend, relative, or neighbor, an attempt is often made to confirm it through contact with public safety officials or through the mass media. Conversely, a warning first received from an official source is frequently discussed with informal sources in order to I
assess its relevance.
Furthemore, persons will contact relatives and i
friends to ascertain that they have received the warning, thereby extending and reinforcing the warning process.
Under normal circumstances
'j persons exchange important information over the phone or in person with others significant in their lives.
The major difference during an emergency is that large numbers of persons have a similar piece of i
information at the same time, thereby contributing to jams of communication channels.
Preparations for the selected protective action, and the eventual action, also are characterized by social interaction rather than by individual self-interest.
It is the family that typically is the unit of response to a threat. With few exceptions, it has been found that families separated at the time of initial warning seek to unite and then take action as a unit, or at least to communicate before leaving and agree to meet at another location. At Mississauga, for example, some parents -
who had gone out on Sunday morning, found that barricades set up in their absence significantly delayed their attempt to return home to evacuate their children. Many had arguments with police or circumvented the barricades (Burton, et al.,1981, pp. 6-14).
4-11
l The time of day and day of week can affect the extent to which i
families are likely to be all together and thus the amount of time they are likely to spend in coordinating prior to taking action.
Until the details of who is where and who will go where have been worked out, u
families are reluctant to take an action such as evacuation.
The two best ti documented counterexamples of the rule of families evacuating as a unit were observed in Anchorage following the 1964 earthquake and at Three Mile Island in 1979.
In both cases, it is possible that family members disagreed about the degree of threat, with the members judging the threat to be less severe staying behind. Altet natively, it might have been that 1
fears about personal safety were counterbalanced by concerns about I
property (e.g., beliefs about $he threat of looters) or costs of evacuation, such as icst wages.
Once the protective action has been adopted, primary groups continue to structure behavior.
For example, groups of friends or two or more families may get togetnbr to wait out a hurricane, or to engage in protective actions like sandbagging in 4' flood.
When evacuation is the selected protective action and family trembers have been gathered, the most typical destination is the home bf other relatives or friends outside the danger area.
It must be noted that not all persons in a comunity will be integrated into families or friendship networks. Furthermore, because of the numbus of persons likely to be affected by a comunity emergency, and because some of the problems faced are of a technical nature not suited to 1
the capabilities of families, protective functions are shared between primary groups such as families and fonnal community organizations with special capabilities and responsibilities.
In particular these l
4-12 l
1 L- -
.1:
l i.
1' l~
organizations will be needed to handle'what can be called " response--
1 generated demands" (Quarantelli,1982). As families 'and indiv; duals respond to information about the threat, appropriate organizational l
support will be needed to facilitate communication, continue assessment of-i the emergency, mobilize necessary human and material resources, and coordinate and control the actions of the population as.it responds.
1 1
1 4.2.3 Requirements for Organizational: Support l
Under certain conditions, protective action decisions are made and-implemented by the affected individuals without any intervention by public.
I authorities. Perry, et al. (1981), for example, studied a site struck.by.
j a flash flood in which levee failure was detected by nearby residents and l
passed on directly.to friends, relatives and neighbors in the risk' area.
.-l l
Siit:e the threat area was small-and its boundaries easily defined by local residents, evacuation transportation was not a' problem.. Also, because the -
risk area was small, most evacuees were able to stay with friends, relatives or neighbors until the emergency was' over..
In other cases', there may be a strong need for local. authorities to j
l l'
provide support for the warning, evacuation and temporary housing of local residents. Needs in each of these areas can differ significantly across segments of the population. The general public--those with no impediments; to warning receipt, no need for evacuation transportation and no requirements for care--needs little more than a specific message defining-the threat and the appropriate protective action.
The transit dependent--those without private automobiles--may need buses or other mass transit in addition to a specific warning. Special populations, such as invalids, handicapped, school children, prisoners and others, may have l
l 4-13 n
\\
very restrictive requirements associated with their movement.
The characteristics of the hazard impact--especially the amount of forewarning, the speed of onset and the scope and duration of impact--are alm important factors to consider in anticipating the need for support ftam public safety agencies. These characteristics can impose significant constraints upon the protective action process.
- 4. 2. 3.1 Notification A requirement for an organized system of hazard detection and notification dissemination emerges when a hazard does not provide environmental cues that the general public can use as a means of forewarning of hazt.rd onset.
The hazard detection / notification j
dissemination system assumes a particularly significant role when the speed of onset of the hazard exceeds the speed at which the affected population can respond.
In such cases, the burden of hazard detection l
l must be assumed by specialized agencies who can transmit infomation to local notification dissemination networks. For natural hazards, this function is perfomed by the National Hurricane Center, the National 1
Severe Storms Center and other su:h agencies.
In the case of technological hazards, this function is performed by the plant personnel at fixed facilities, or employees of the carrier in a transportation
)
i inci dent.
In bither case, information about the hazard is passed on to State and/or local public safety officials who then evaluate the threat and disseminate a warning to the population of the risk area. When a hazard is detected in advance, a given notification mechanism or combination of mechanisms can disseminate a specific warning to the entire population of the risk area in a timely manner.
The most rapid means of
)
I l
4-14 I
__ ___ ____ - 2
9 8.1. 2 Perceived Characteristjes of Alternative Protective Actions At present, the data en phblic perceptions of the characteristics of alternative protective actions are very sparse.
There are no data on perceptions of the efficacy of sheltering. With respect to perceptions of evacuation, there is one result reportec in the Seasonwein (1983) survey that deals with perceptions of the barriers to-implementation. Almost two-thirds of the respondents in this survey rated as a strong argument the assertion that an evacuation of the area surrounding the plant (a densely populated area) "would be impossible in light of the numbers of people to be moved, the existing road network and the lack of. protective equipment." Concerns about barriers to implementation of evacuation need not necessarily stimulate compliance with recommendations to shelter.
However, Houts, et al. (in press) found thai. oncerns about a later ' forced evacuation were expressed by some of those who were voluntary evacuees during the TMI accident.
Houts and his colleagues inferred from the TMI.
data that perceived barriers to a mass evacuation (that ultimately was never ordered) helped to stimulate the early, voluntary evacuations.
8.1.3 Intentions to Comply with Official Plans In contrast to the paucity of data on perceptions of the characteristics of radiation hazard and of the alternative protective actions, data op intentions to comply with recommended protective actions are not difficult to find.
Data on behavioral intentions can be somewhat difficult to interpret.
The difficulty in interpretation arises from questions about the degree to which a survey interview evokes the behavioral cues that would actually guide an individual's response in an actual emergency.
Unless great care is taken in conducting the interview, 8-4
behavioral cues that would be.important in an actual emergency might.be omitted.
In a brief telephone survey, for example, respondents' attention would be directed toward only the most salient aspects of the situation rather than to all of the relevant aspects of the~ situation. Since all three of the available studies (de Pujo,1982; ~Seasonwein,1983;. Zeigler and Johnson,1984) addressing intentions to comply with recommended protective actions were based upon telephone interviews, these surveys and their results should be interpreted with caution.
The results of three studies will be presented in some detail' below; the implications for a nuclear power plant emergency that can reasonably be drawn from them will be presented in Section 8.1.4.
8.1.3.1 Intended Response to Prompt Notification d
In the,de Pujo (1982) study, respondents were asked what they would do if the sirens were sounded, approximately half (45%) of the respondents said that they would seek. further information or instructions, which is, i
of course, the officially recommended response.
Of the remainder,10%
would prepare to evacuate and 19% would begin to evacuate.
Many of those who would_ evacuate spontaneously indicated that they would monitor the situation on their car radios as they evacuated.
When respondents were asked in another question if they would generally follow official instructions, approximately half (47%) said they would do so.
Most of the rest would make their own plans (23%) or comply with official plans if these matched their personal plans (26%).
8-5 s
_a___-__-_.
l I
i i
8.1. 3. 2 Compliance with a Recommendation to Shelter In a study by Zeigler and Johnson (1984) respondents were asked how they would respond to an advisory for all people within 5 miles of the plant to stay indoors. A majority of the respondents within the affected area (52%) said that they would shelter, while a significant minority (40%) said that they would evacuate.
In addition, a number of residents outside the affected area said that they would shelter (49% of those 6-10 miles from the plant) or evacuate (40% of those 6-10 miles from the plant).
A somewhat similar response was obtained in the Seasonwein (1983) l When asked what they would do if instructed to shelter, the survey.
majority (66%) said they would comply with official instructions.
Most of 4
the remainder (25%) said they would leave the area. A comparable response was found in the survey conducted by de Pujo (1982).
In response to a question similar to the one asked by Seasonwein, 57% said they would shelter if that were the official recommendation and 31% said they would not.
I 8.1.3.3 Como11ance with Evacuation Procedures In the Seasonwein (1983) survey, respondents were asked whether they would take predesignated routes in an evacuation. Only 31% said that they knew that there was a predesignated evacuation route and only 41% intended to take it.
Most of the rest (56%) reported that they would leave the best way they knew how. As was noted in Section 5.1.2.2, 86% of the respondents said that they did not know if they would comply with plans for direct evacuation of school children.
l l
l 8-6
4 8.1.4 Sumary of Expected Response by the Public As was noted in the previous section, the behavioral intentions studies conducted by de Pujo (1982), Seasonwein (1983) and Zeigler and Johnson (1984) should be interpreted cautiously.
This is not to say that the possibility of spontaneous evacuations that is raised by these studies should be rejected.
In fact, there are two examples other than TMI in I
which spontaneous evacuations have been documented.
It should be noted that neither of these examples was associated with a nuclear power facility.
The first of these was reported by Lindell, et al. (1982) in their study of evacuation times following the Mt. St. Helens eruption.
Over 10% of the residents of the Woodland sample evacuated during the May 18, 1980 eruption; no official evacuation warnings were given for this P
For the other case of spontaneous evacuations, the response was area.
i
./
j more extensive.
These took place in the neighborhoods between 4 and 8 kilometers north of the Mississauga train derailment.
In that area where, again, no official evacuation recommendation was given, approximately 60%
of the residents evacuated.
Although a few of them said that they thought they had been given an order to evacuate, most left because of their own perception of the chlorine danger.
The Mt. St. Helens and Mississauga studies indicate that asymmetries in the affected area (due to wind direction) are particularly likely to produce overresponse by the public.
If, for example, protective action is recomended within a 2 mile radius and 5 miles downwind, an overresponse is most likely in the 2-5 mile upwind sectors. The magnitude of the overresponse in other areas of the EPZ (e.g., 5-10 miles upwind and downwind) as well as outside the EPZ is difficult to determine, but is 1
probably overstated by the behavioral intentions surveys conducted to 8-7 l
i
date.
In none of the three surveys cited above did the investigators include any information about the efficacy of sheltering--information which would undoubtedly accompany any recommendations made by offsite authori ties.
In addition, they did not clarify or cite any of the factors that have been found to inhibit evacuation (e.g., concerns about' lost pay, looters,etc.).
Three general implications follow from the data reviewed in the previous sections. First, the available data suggest' that one can anticipate reasonable levels 'of compliance with officially recommended protective actions.
A majority of the respondents in all three behavioral I
intentions surveys report that they would shelter if advised to do so.
It should be remembered that the level of compliance may be quite issue r
specific. The intended level of compliance was lower when questions were asked about following posted evacuation routes or about picking up
)
children at schools rather than at reception centers, i
The second point is that a significant minority of the residents of I
the affected aree (and in some cases outside the affected area) may not comply with one or more official recommendations.
The failures to comply would tend to be more in the direction of overresponse than underresponse.
This means that voluntary evacuations would begin earlier and have a broader scope than the officially recomended evacuations.
Residents would compare themselves to those for whom protective action is recomended.
The closer they are in terms of geographic distance and the more similar they are with respect to geographic proximity or demographic characteristics (e.g., pregnant women and preschool children) to the target group that has received a specific protective action recommendation, the more likely they are to also take that protective action (Houts, et al., in press).
8-8
Third, none of the data indicate that panic is a condition that is -
likely to occur if the offsite response to an accident is managed properly.
The available data do indicate that a reactor accident would'be perceived to put residents in severe danger and it is perceived that this danger could be rapid.
This is similar, but not identical to the first of-the four conditions required for panic.
The critical difference here is between the perception of " rapid"_ and "immediate" danger.
To the degree that the situation is portrayed as serious, not desperate, and requiring 1
timely, but not instantaneous response, the likelihood'of panic will be j
low. A similar situation holds #or the second and third conditions, limited escape routes that are about to close.
Evacuation of individual i
sectors, or even groups of sectors, in sequence would tend to decrease I
publ.ic perceptions of escape routes closing due to overloading.
The strategy of phased evacuation was clearly very effective at Mississauga and would be especially appropriate for other sites with high population densities. The last condition for panic, lack of infonnation, can be l
avoided by providing local-residents with an accurate assessment of the situation that is updated frequently (as of ten as one or more times per ~
hour).
The offsite emergency management effort would almost surely be hampered if rumors or misperceptions stimulated local residents to inappropriate action because of a lack of timely and accurate information about the severity or immediacy of the hazard or availability of evacuation routes.
8.2 Expected Response Time Components for Specific Emergency Scenarios This section will address the expected levels of timeliness and compliance with protective action recommendations for the general public.
8-9
_____._______m__m----
j emergency. When an wergency arises, however, the public needs much of this potentially ignoreo information.
-j A suitable response to this situation requires an understanding of -
two principles. First, planners must recognize that it'is important to persist in disseminating emergency public information.
Although some residents do not read or retain this i formation, there are many that do.
n It is better that some of the public be informed than that none of them be informed. Moreover, those who have read and retained the information may, themselves, be a valuable source of that information to their friends, I
relatives and neighbors in any emergency.
Second, planners must recognize l
1 that emergency information may not be read or retained by some individuals l
because it is perceived to have no relevance to people's normal activities, 9.2 The Sources of Emergency Information The discussion of a source's characteristics presented in Section 4.2.1 described the roles of expertise and trustworthiness in determining the credibility of an information source.
That section noted that, typically, multipl. sources of infonnation are consulted in an emergency.
This has three implications for attempts by any emergency personnel to I
communicate with the public. First, unofficial sources (friends, relatives and neighbors) are likely to be considered to be low in expertise and of mixed trustworthiness.
They would be expected to have a l
limited perspective (knowledge bias) but be willing to consnunicate what information they did possess honestly.
It is likely that different l
members of this group would have different--possibly conflicting--
information and recommendations.
l 9-2 L__________________________________________
i A second implication is that the news media, partia11y'because of the<
role that they see for.themselves as seeking _ out the "whole" story, deliberately seek a diversity of information and opinion from a wide range j
of sources.
Even if all were basing'their assessments on exactly'the same
-information (a rather improbable situation in any emergency), differences 1
of opinion can be expected to einerge.
Public response to sources interviewed directly by the media will vary with each source's position.
h University professors, for example, are often sought out as independent' j
experts because, in addition to their expertise, they are considered to be i
i willing to communicate information without bias, although they may have l
knowledge bias that is caused by overspecialization. Media commentators may be considered credible sources if, although not experts in their own-right, they are considered to be able to elicit several points of view and.
come to a reasoned judgment of the situation.
In addition, an esteemed-media personality could become a link with the outside for a sheltered person and could therefore play a significant role in helping to allay the fears of those being sheltered in any emergency.
A third implication is that official sources--corporate personnel and governmental officials at the federal, state and local-levels--tend, as a group, to have complementary aspects of credibility. Corporate personnel are likely to be considered knowledgeable about the status of their facility, but may not be considered trustworthy because of a perception of l
knowledge bias and, more importantly, a perceived lack of willingness to reveal the true extent of a giver situation. Therefore, as a " composite" source, a group made up of corporate, local, state and federal personnel would have increased credibility if their information notices and protective action recommendations were consistent with each other, i
9-3
7--, -
l
~
.l
' ' ~
ql{
( -- eatrv4'o,, L
.. c UNITED STATES l
. 8
.g NUCLEAR REGULATORY COMMISSION 1
p-
- e WASHINGTON, D C. 20555 og a
w g
,N
..n n liay 26, 1987 MEMORANDUM FOR:
John'Milligan 1
T
.Technassociates
- j FROM
Emile L. Julian
- cting' Chief Docketing and.. Service: Branch l
SUBJECT:
W Af M/n EXHIBITS
-l j
i
' Any documents filed.on the o)en record 'in the suoncHAM pro.
ceeding and made a part of.t1e official hearing record as:an-exhibit is considered exempt from the provisions of the United f
' States Copyright Act, unless it was originally. filed.under seal with the court' expressly because of copyr.ight concerns.
All of the documents sent to TI for processing' fall'.within,the exempt classification.
a 1
l l
1
)
J
.il i
- )
i l
.j
,