1. Planning for the
protection of the public in the event of a radiological release that
extends beyond the boundary of TMI was highly complex. It involved the
utility and government agencies at the local, state, and federal levels.
That complexity posed problems in the case of the accident at Three Mile Island;
some of the written plans that existed had not been coordinated and
contained different systems for classifying accidents and different
guidelines for notifying government officials.
2. In approving
sites for reactors, the NRC has required licensees to plan for off-site
consequences of radioactive releases only within the "low population
zone" (LPZ), an area containing "residents, the total number and density
of which are such that there is a reasonable probability that
appropriate protective measures could be taken in their behalf in the
event of a serious accident." As calculated for the design-basis
accident for TMI-2, this zone was a 2-mile radius.
3. Emergency
planning had a low priority in the NRC and the AEC before it. There is
evidence that the reasons for this included their confidence in designed
reactor safeguards and their desire to avoid raising public concern
about the safety of nuclear power.
4. The NRC has not
made the existence of a state emergency or evacuation plan a condition
for plant licensing. A state may voluntarily submit a response plan to
NRC for concurrence, and if the plan meets NRC guidelines -- which do
not have the force of law -- the state receives a formal letter of
concurrence. At the time of the accident, Pennsylvania
did not have an NRC concurred-in plan. The NRC concurrence program has
been called ineffective by federal and state emergency preparedness
officials.
5. The utility has
the responsibility to prevent or to mitigate off-site radiation releases
and to notify the government agencies designated in its emergency plan
in the event that an emergency is declared. Federal, state, and local
agencies are responsible for off-site response to radiation releases. At
the time of the TMI accident, the local and county governments had the
primary action role once notified of the emergency. None of the local
communities within the 5-mile radius of the plant had emergency plans,
and the existing county plans did not include detailed evacuation bf.
plans.
6. At all levels of
government, planning for the off-site consequences of radiological
emergencies at nuclear power plants has been characterized by a lack of
coordination and urgency. For example, a federal response plan in
preparation since 1974 by federal emergency preparedness agencies was
unfinished at the time of the accident because of an interagency
jurisdictional dispute and lack of communication. Pennsylvania did not
begin to develop a radiological emergency plan until 1975, even though
nuclear power plants had been operating within its borders for at least
a year prior to that time. People who attempted to generate interest in
radiological emergency planning at the local level near TMI found local
officials apathetic. Part of the reason for this was the attitude
fostered by the NRC regulatory approach, and by Met Ed at the local
level, that radiological accidents having off-site consequences beyond
the LPZ were so unlikely as not to be of serious concern.
7. Interaction
among NRC, Met Ed, and state and local emergency organizations in the
development, review, and drill of emergency plans was insufficient to
ensure an adequate level of preparedness for a serious radiological
incident at TMI.
8. Although NRC
personnel were on-site within hours of the declaration of a site
emergency and were in constant contact with the utility, the NRC was not
able to determine and to understand the true seriousness and nature of
the accident for about 2 days, when the fact of extensive core damage
and the existence of the hydrogen bubble were generally recognized
within NRC.
9. During the first
2-½ days of the accident, communications between the NRC Incident
Response Center in Bethesda, Maryland, where the senior management was
located, and the site were such that senior management officials found
it extremely difficult to obtain up-to-date information. Communications
were so poor on Friday morning that the senior management could not and
did not develop a clear understanding of conditions at the site. As a
result, an evacuation was recommended to the state by the NRC senior
staff on the basis of fragmentary and partially erroneous information.
Communications did not improve until Harold Denton, designated the sole
source of information, arrived on the site and communicated with NRC
headquarters, the Governor's office, and the White House by White House
communications line.
10. The reality of
possible evacuation was quite different from the theoretical planning
requirements imposed by the NRC and Pennsylvania before
the accident. The 5-mile emergency plans were developed according to a
Pennsylvania requirement for emergency planning within a 5-mile radius
of nuclear power plants. The Pennsylvania requirement was stricter than
that prescribed by NRC,
which only required TMI to have a plan for a radius to 2
miles. (See finding D.2.) It is known that the consequences of a
postulated major release to the atmosphere from a reactor accident could
lead to significant doses of radiation being received many tens of miles
from the site of the accident. At TMI-2, although the radiation releases
were significantly lower than the design-basis accident, evacuation was
being considered for distances much greater than 2 miles. During the TMI
accident, NRC believed that the consequences of the accident might
extend far beyond the 2- or 5-mile radius. As a result, evacuation plans
were hurriedly developed for distances of 10 and 20 miles from the
plant.
11. During the most critical phase of the accident, the NRC
was working under extreme pressure in an atmosphere of uncertainty. The
NRC staff was confronted with problems it had never analyzed before and
for which it had no immediate solutions. One result of these conditions
was the calculational errors concerning the hydrogen bubble, which
caused the NRC to misunderstand the true conditions in the reactor for
nearly 3 days.
12. On Friday and Saturday, certain NRC officials incorrectly
concluded that a hydrogen bubble in the reactor vessel would soon
contain enough oxygen to burn or explode. Ignoring correct information
supplied by a B&W employee and certain members of its own staff, NRC
relied instead upon incorrect information supplied by other members of
its staff and by others that sufficient oxygen was being formed from
water radiolysis to reach a concentration sufficient for a burn or
explosion. Based on this information, the NRC commissioners began
formulating new recommendations for evacuation. On Sunday, NRC staffers
obtained information from several national laboratories and from General
Electric and Westinghouse that sufficient oxygen could not form. The
Sunday information ended the concern about oxygen formation and
evacuation. This incident suggests that NRC lacks sufficient knowledge
and expertise in water radiolysis.
13. The role of the NRC commissioners and their
decision-making process during the accident were ill-defined. Although
the commissioners on Friday assumed responsibility for making
recommendations to the Governor concerning protective action, there was
no apparent procedure by which issues and staff recommendations were
explored and resolved. The commissioners were preoccupied with matters
such as the details of evacuation planning and the drafting of a press
release.
14. Existing emergency plans were not designed to meet the
demands of a protracted crisis. The plans had no mechanisms for
establishing reliable communications among the on-site and the several
off-site organizations responsible for various aspects of the emergency
response.
15. There were no hospitals within 5 miles of TMI, but there
were several hospitals within the expanded, proposed evacuation zones.
The NRC estimated that it would be able to give officials a few hours
"lead time" for evacuation. But hospital administrators estimated they
would need substantially more time to evacuate patients.
16. During the TMI accident, the actual radiation levels
outside the plant were low, but there was uncertainty about the
possibility of serious releases on short notice. Federal and state
officials disagreed about the nature of the information on which to base
evacuation decisions and other protective actions during the emergency.
Some officials based their decisions on actual radiation exposure
levels, while others based their decisions on concerns about potential
releases of large amounts of radioactivity. For example, the
Pennsylvania Bureau of Radiation Protection told the Governor on Friday
that radiation levels indicated that no protective action of any kind
was required; on that same morning, NRC Chairman Hendrie recommended
that pregnant women and young children be advised to leave the area near
the plant because of his concern about possible releases of
radioactivity.
17. At approximately
12:30 p.m., March 30,
Governor Thornburgh advised pregnant women and preschool aged children
to leave the area within a 5-mile radius of TMI until further notice. A
substantial number of other persons, including health professionals,
voluntarily left the area around the plant during the weekend of March
30 through April 1. The advisory to pregnant women and preschool
children was lifted on April 9.
18. Throughout the accident, the Pennsylvania Emergency
Management Agency (PEMA) received reports concerning conditions at the
site from the Bureau of Radiation Protection. During the first 2 days of
the accident, however, the director of PEMA also received background
information on the status of the plant from the Governor's office by
attending meetings and press conferences and relayed that information to
county organizations, which, in turn, informed the local civil defense
directors. Starting Saturday, the PEMA director was no longer included
in these meetings with the result that PEMA and county and local civil
defense organizations had to rely primarily on the news media for
information about conditions at the site. They found this an
unsatisfactory source of information and believed that this arrangement
compromised their effectiveness in responding to the accident.
19. The TMI emergency plan did not require the utility to
notify state or local health authorities in the event of a radiological
accident. (See also finding C.7.)
20. For over 25 years, the use of blocking agents such as
potassium iodide to prevent the accumulation of radioiodine in the
thyroid gland has been known. The effectiveness of potassium iodide
administration for thyroid gland protection in the event of releases of
radioiodine was recognized by the National Council on Radiation
Protection and Measurement in 1977. The Food and Drug Administration
authorized use of potassium iodide as a thyroid-blocking agent for the
general public in December 1978. However, at the time of the TMI
accident, potassium iodide for this use was not commercially available
in the United
States in quantities sufficient for the population within a 20-mile
radius of TMI. At the time of the accident. Met Ed had no supply of
potassium iodide on-site. A crash effort by the federal government and
private industry resulted in delivery of substantial supplies of
potassium iodide to
Pennsylvania
within 2 days of the decision to obtain such supplies.
