15

O

"RES would prefer an integrated approach looking at the
source terms and risk from all of the six
SARRP/NUREG-1150 reference plans to determine what the
generic NRC positions should be and what, if any,
plant-specific alternatives might exist. To proceed
first on a plant-specific basis could preclude a more
rational, policy-level look at the problems, which could
result in inadvertent, unfortunate decisions."

The memorandum concluded by saying that "We understand that you have sent a letter to BG&E to the effect that it is premature to consider reducing the EPZ, and we strongly support that action."

A January 6, 1986 memorandum to Denton from Thomas Murley, NRC's Regional Administrator for Region I, also offered pertinent comments on the BG&E proposal. In that memorandum, Murley wrote:

Two

"...The consideration of this exemption
request should be approached cautiously.
important questions must be answered before
this request, and others of this type which
may follow, can be considered solely for their
logistical and technical merits. The first
question is... Who determines the size of the
EPZ?... One may argue that the state and local
governments should play a major role in
determining who is to be protected and
subsequently in setting the size of the EPZ.
The State of California, for example, requires
planning for an Expanded Planning Zone
(20-mile radius)..........

"The most conservative approach would be to
consider and resolve the generic questions
prior to considering exemptions on a

plant-specific basis. This is our [the Region
I staff's] recommendation."

Murley's comments on the appropriate role of state and local governments in determining the size of the EPZ sound somewhat incongruous in light of the staff's present involvement in an attempt to craft a technical argument designed precisely to circumvent state and local government concerns about emergency planning by summarily excluding those governments from the EPZ.

The NRC staff position in January 1986 clearly opposed dealing with such requests on a plant-by-plant basis. We see no evidence that the situation 11 months later is substantially different.

PAST NRC STATEMENTS REGARDING REACTOR SAFETY, CONTAINMENT
FAILURE, AND ACCIDENT PROBABILITIES

16

CURRENT NRC WORK ON RISK REBASELINING AND SEVERE ACCIDENTS

After the TMI accident the NRC began a comprehensive review of the risks of a severe accident. According to discussions between NRC staff and Subcommittee staff, this review has taken more than 5 years and has cost more than $250 million. The objective of the review is to attempt to establish with greater precision appropriate source term assumptions for a severe accident and to refine risk assessments so as to improve upon the work of the 1975 Reactor Safety Study (WASH-1400).

According to staff, the draft document resulting from this review, NUREG-1150, is currently scheduled to be released for review and comment early in 1987. Following a 120-180 day review period, staff expects to take an additional 120-180 days to digest comments and make changes to the draft document. While the staff has informed us that the objective includes various rulemaking proceedings, they do not expect those proposed rules to be issued until the final NUREG-1150 is released late in 1987 or early in 1988.

Many of the issues under review in the NUREG-1150 process are directly relevant to the issues under discussion with regard to Seabrook. It is extraordinarily difficult to understand why the NRC staff seems intent upon fine-tuning a study for one plant when a sweeping generic review is drawing to a close. The NRC bears an extremely heavy burden to justify its departure from the rationale offered in the Minogue memorandum (above). The NRC has provided no reason why it could not await the promulgation of rules pursuant to the publication of NUREG-1150 before taking any action on the size of the Seabrook EPZ that depends upon containment and risk analyses for the Seabrook plant. To proceed otherwise directly contradicts positions taken by staff only months earlier.

LARGE UNCERTAINTIES IN SAFETY-RELATED ISSUES

In an April 30, 1986 memorandum from Victor Stello, Jr., Executive Director for Operations to Commission Secretary Samuel Chilk, Stello identified a long list of safety-related issues that

17

require further research. Many of them bear directly on issues related to containment performance under severe accident conditions, and therefore on off-site consequences resulting from a severe accident. Although the purpose of Stello's memorandum was to argue for additions to the NRC budget necessary to pursue research programs in these areas, the issues themselves paint an extraordinary picture of the gravity of outstanding safety-related issues and the extreme lack of knowledge on the part of the NRC about severe accidents. Given the number of items enumerated and the breadth of their implications, it is difficult to understand why the Commission staff is intent on expending its time and public funds on the Seabrook technical studies in light of these many extant technical uncertainties. Indeed, the breadth of the safety issues appears so great that we recommend the Subcommittee consider an independent investigation into their implications for operating reactors.

Stello wrote in his introductory narrative:

"In evaluating the various technical issues
that must be resolved to provide a basis for
[regulatory] actions, it is becoming apparent
that some of the uncertainties may be so
large, even with the knowledge gained from the
four years of intense focused research to
date, that NRC may not be able to provide a
satisfactory reduction of these uncertainties
with existing resources.... Some of these
issue may be the chemical and physical form of
radioactive iodine and cesium and their mutual
interaction, the degree of direct containment
heating from the expulsion of molten reactor
core materials, hydrogen generation and loads
and containment performance in resisting these
and other loads placed upon it from a severe
accident." [Emphasis added]

We provide below excerpts from the Stello memorandum.
Safety Issue: Large uncertainties exist in
analyzing in-vessel accident behavior at
core-melt conditions, including clad
oxidations heating, hydrogen generation and
transport. This in turn leads to large
uncertainties in containment heating and
penetration calculations which in turn are
used in health effects consequence
determinations.

Safety Issue:

Large uncertainties exist in the in-vessel fission-product release rates currently used in Source Term analysis with the Source Term Code Package (STCP). These rates are taken from early, old, relatively coarse, out-of-pile data at atmospheric pressure. The relatively volatile fission products not released from the fuel in-vessel and then removed by plate out in the reactor vessel are immediately released ex-vessel by melt-concrete interactions. In cases of early containment failure these fission products can become the largest part of the release to the containment.

Safety Issue: One of the safety issues is
whether or not, in the absence of a hydrogen
control system, given hydrogen-steam mixtures
in large dry containments, would such mixtures
autoignite and burn slowly destroying
containment seals or, alternatively, burn
rapidly and damage containment equipment and
structures. The information would allow NRC
to evaluate the need for changes in equipment
qualification and hydrogen regulations.
Another issue not resolved involves the
quantitative loads to containment walls and
electrical and mechanical safety equipment
from local hydrogen detonations and the
probability of wall and equipment damage."
Safety Issue:

Significant uncertainties exist in the safety analysis codes ability to calculate reactor response to

feedwater-line/steamline breaks. These uncertainties affect NRC's ability to evaluate such effects as: 1) thermal shock to the reactor vessel under pressure in the event of a steamline break; and 2) coolant system overpressurization during feedwater line

breaks.

Safety Issue: To confirm the adequacy of data obtained from plant instrumentation and the availability of electrical equipment under severe accident states which provide the basis for operational and emergency preparedness

18

19

actions; and to provide more accurate data for
deterministic and probabilistic calculations
as they pertain to severe accident states.

Safety Issue: The requirement of separation
of the redundant trains of safe shutdown
equipment prescribed by 10CFR50, App. R cannot
be met in many control rooms. The rule
requires in such cases that an alternative or
dedicated shutdown capability be provided.
Whether in the event of a credible control
room fire the operator will have time enough
and the physical ability to transfer control
of the reactor to the alternative or dedicated
shutdown panel is at issue. There is also the
question of how much of the control room
equipment will survive and whether and when
control room operation can be resumed.

Safety Issue: The currently available
real-time atmospheric dispersion,

precipitation washout and plume rise models
for dose projections in emergency response
situations have not been adequately validated.
Accordingly, there is the potential for an
inappropriate, unreliable or inaccurate model
being used in an emergency response situation
and [sic] which would produce incorrect dose
projections....Risk assessments and other

licensing evaluations in which the models are
used have large uncertainties.

Clearly, there are an extraordinary number of open safety-related issues, some bearing directly on the Seabrook analysis. We fail to understand how the utility and the NRC staff can provide sufficient assurances that the Seabrook containment is virtually "fail-safe" in light of these considerable uncertainties remaining in understanding and modeling basic phenomena.

OTHER COMMENTS ON POSSIBILITIES OF A SEVERE REACTOR ACCIDENT

In light of the efforts by PSNH and NRC staff to demonstrate that the Seabrook containment is virtually "fail-safe" and that a miniscule probability exists of an accident with significant off-site consequences beyond 2 miles, Subcommittee staff felt it useful to review several knowledgeable sources' comments regarding