ML19296C037
| ML19296C037 | |
| Person / Time | |
|---|---|
| Site: | Black Fox |
| Issue date: | 04/01/1977 |
| From: | Dalton A Citizens Association for Sound Energy, DALTON, A.T. |
| To: | Atomic Safety and Licensing Board Panel |
| Shared Package | |
| ML19296C015 | List: |
| References | |
| RTR-WASH-1400 NUDOCS 8002250096 | |
| Download: ML19296C037 (24) | |
Text
~
~
~,.yJc. % ths
.... -. w v.:=
i
_ - $ q-
?
. 2::s EXHIBIT C
-;.{y 2
i_'f:.
UNITED STATES OF AMERICA NUCLEAR REGULATORY COMMISSION BEFORE THE' A' COMIC SAFETY AND LICENSING BOARD IN THE MATTER OF
)
)
PUBLIC SERVICE COMPANY OF
)
DOCKET NOS. STN 50-556 OKLAHOMA, ASSOCIATED ELEC- )
STN 50-557 TRIC COOPERATIVE,'INC. AND )
WESTERN FARMERS ELECTRIC
)
COOPERATIVE, INC.
)
)
(Black Fox Stations, Units )
1 and 2)
)
INTERVENORS', C.A.S.E.
ILENE YOUNGHEIN, LAWRENCE BURPSLL AND CHARLES GLYMORE, SUPPLEMENTAL SUBMISSION RE:
CLASS NINE ACCIDENTS Attached hereto is the named Intervenors supplemental sub-mission regarding Class Nine Accidents.
Dated this 1st day of April, 197
~
CDDi' ANDREW T. DAL'f0N, JR.
Attorney for Intervenors 2536 East 51 Street
' Tulsa, OK 74105 "g
8002250 g
EXIIIDIT A.
s I
In view of the above-described uncertainty of the likeli-hood of pressure vessel rupture it is essential that the Black Fox plant include the design of systems to prevent the Class 9 accidents which could occur from the following vessel related failures:
1.
Vessel rupture by rapid expansion of circumferential or axial cracks into through-wall cracks.
2.
Rupture of the vessel due to the thermal shock caused by interjecting cold emergency cooling water into a hot ves-sel in an attempt to restore water level.
3.
The vessel failure due to aging and/or neutron-induced metal embrittlement.
4.
Vessel rupture from development and propagation of ves-sel flaws undetected due to a build up of radioactivity preventing proper inspection and testing.
-=
~
II s
The reactor pressure vessel is one of the most critical elements of a reactor system with respect to single-failure accidents having massive potential consequences and little possibility of mitigation.
Catastrophic rupture of the vessel is surely a possible event.
It can have a number of consequences, which involve a considerable likelihood of a very massive release of radio-activity prcr.pcly and directly to the atmosphere.
1 RSS states for PWR Vessel Ruptures:
"Potentially large ruptures in the vessel were considered that could prevent effective cooling of the core by ECCS.
Since certain of these ruptures appeared to be capable of causing missiles (such as the reactor vessel head) with s
sufficient momentum to rupture the containment, this area was examined with some care.....There is some small probability that a large vessel missile could in fact impact directly on the con-tainment and penetrate through the wall. (Sic).
This type of rupture could involve a core melt-down in a non-intact containment."
Similarly massive vessel ruptures in a BWR below the reactor core shroud could result in rupture of-the primary containment and possibly the secondary containment.
Such failures might also lead to excessive blowdown forces on reactor intervals which could disrupt the core and result in failure of the emergency cooling function.
RSS also states that:
"Very large vessel ruptures could result in high energy missiles, such as the reactor vessel head,
~;
jy or could cause motion of the reactor vessel to An intermediate breach of the primary occur.
containment structure and loss of leakage integrity could result in either case since the pressure sup-
~
d pression Lyctem could not accommodate the rate of e
onergy released.
In addition, the containment is small in size and its proximity to the reactor vessel would indicate that severe vessel ruptures might either tear or cause missiles to penetrate the containment shell."2 Two BWR vessels at Oyster Creek and Nine Mile Point had serious cracking of furnace sensitized stainless steel vessel components prior to or shortly after startup.
Repairs were made but reoccurrence remains a fear.
The Hatch vessel became the subject of a lengthy dispute during its manufacture.
It was judged acceptable by boilder ccde minimum standards, but advanced ultra-sonic techniques gave evidence of extensive sub-surface flaws, which in turn gave rise to concern about what had been missed with past and accepted inspection practice.
/
A review of the non-nuclear vessel rupture experience was s
carried out in 1970 by Oak Ridge National Laboratory 3
metallurgist Monroe Wechsler.
His study fails to provide much support for the RSS-selected vessel rupture probability.
3 Some of his conclusions were:
"The safety analyses for nuclear power plants are based on a maximum credible accident that falls short of large-scale vessel rupture.
Based on pressure vessel failure statistics from England and Germany and the increased use of nuclear power reactors in the years ahead, the wisdom of such a safety policy is ques tioned. "
"It is interesting that the British and German studies both imply a probability rate of about 2 x 10-5 per vessel year for catastrophic or large-scale failure in vessels comparable to nuclear pressure vessels."
O "In no reasonable sense of the word is a large-scale vessel rupture ' incredible'."
The British government was considering in 1974 the pur-chase of approximately 18 PWR power plants.
A number of eminent s
metallurgists were troubled by the vessel rupture possibilities.
Chief among these was Sir Alan Cottrell, a fellow of the Royal Society, and Chief Science Advisor to the British government.
4 He has stated that:
"The security of (a light water reactor) vessel must... depend on the maintenance of an immaculate standard of manufacture and quality control.and on regular in-service inspection of the most rigorous and detailed kind.
I hope that the safety of the public in this country (Great Britain) will never be made dependent upon almost superhuman engineering and operational qualities."
The United Kingdom Atomic Energy Authority had instituted a study in 1973 to consider such questions, and the results were published in late 1976 as the Marshall Report.
This report concludes:
" Subject to a number of considerations, we recom-mend it WOULD now be possible for the Nuclear Inspector to be satisfied about the safety of pres-sure vessels.
." (Emphasis added).
5 The qualification included:
"That it is essential to confine the operational transients to unusually narrow limits in order to avoid excessive crack growth by metal fatigue."
~~
"That emergency core cooling water should be injected at an unusually high temperature in order to minimize the risk of fracture by ther-mal shock."
'.c "The report makes a further 25 recommendations for improved confidene:3-including the need for a considerable amount of scientific effort aimed mainly at guarantee.ing a long, safe life for the pressure vessel."
N Such improvements, if satisfactorily concluded, may reduce s.
vessel rupture probability, but the fact remains that these precautions have not and are not now being taken by U.S.
operational plants.
e
- s.
4 REFERENCES 1.
RSS, Appendix I, p.
47.
2.
RSS, Appendix I, p.
56.
3.
- Wechsler, M., "The Radiation-Embrittlement of Pressure-Vessel Steels and the Safety of Nuclear Reactor Pressure Vessels," Oak Ridge National Laboratory, May 1970.
4.
- Cottrell, A., " Fracture of Steel Pressure Vessels", sub-mitted to the Select Ccmmittee on Science and Technology -
January 22, 1974.
5.
The Enercy Daily, Vol. 4, No. 83, October 8, 1976, p. L.
4
.. e--
III MEN INSIGHTS FROM THE REACTOR SAFETY STUDY a
BLACK FOX NUCLEAR PLANT s
A.
SUMMARY
The latest technology in the assessment of risk to public health and safety as a result of a reactor accident is provided by the Reactor Safety Study (RSS), WASH-1400.
The scope of the RSS was narrow, only covering the nuclear plant operation por-tion of the fuel cycle.
Also, the RSS Report did not include the risk from sabotage and equipment aging.
The RSS Team con-cluded that the chance of a core melt was much higher than previously believed, now approaching one chance per 4000 per reactor-year as an upper limit.
RSS goes on to state that addi-tional factors which give rise to large radioactive releases e
s include containment breech, wind, ueather, and population den-sity.
Finally, the RSS concluded that a major reactor accident would result.in a 3,300 prompt deaths, 45,000 early illnesses, 2
2 over 8400 km of contaminated land, and over 560 km requiring the relocation of people. ~Because of the new assessment of the higher probability of a core melt accident, it is now appropriate that the plant design and emergency planning criteria should be based on a core melt accident with containment rupture.
It is important to also note that the RSS Executive Summary does not show the most important consequences of a reactor acci-dent; that is, the long-term health impacts'due to the increased incidence of cancer in the downwind area.
For example, further y
analysis of the RSS shows that less than 1 prompt fatality be-comes 6700 1ctent deaths, and 10 prompt fatalities becomes 1-
12,000 latent fatalities.
The result is that the RSS Execu-
[
tive Summary is highly misleading.
The Black Fox emergency plan does not now adequately address the consequences of massive releases of strontium 90 into the groundwater and subsequently into the nearby river.
Emergency planning for Black Fox should be carried out in all directions to a distance of 25 miles.
In addition, the effectiveness of emergency planning can only be assured by training the public through frequent " mock" drills and mass edu-cation.
A rigorous assessment of risk due to sabotage should be un-dertaken and safeguards strengthened where appropriate.
In view of the difficulty eliminating the possibility of sabotage, strict measures should be adopted regarding permissible popula-tion densities in the nuclear plant accident affected area.
Finally, I am particularly concerned that the nuclear safety issues, such as those we have discusseu, are being dealt with as separate and individual problems.
It is absolutely vital from my perspective that they be dealt with in the aggregate.
The present regulatory bodies do not appear to have the inclina-tion to assume this responsibility.
Consequently, each individual problem is declared " solved", or "under study", or "not important" and the system of nuclear power generation is degraded from its publicized theoretical base in bits and pieces.
You have the opportunity to do this essential, but neglected job.
B.
Most RIe' cent Reacto[ 5afety Study IG
'\\
The latest technology in the assessment of the risk from nulcear power plant accidents is provided by a report released
in October 1975 by the U.S. Nuclear Regulatory Commission s
entitled The Reactor Safety Study.
It represents the results of a three-year long study whose principal obj ective was "to try to reach some meaningful ~ conclusions about the risk of nuclear accidents using current technology."
The study's direc-tor was Professor Norman Rasmussen.
The scope of the RSS was quite narrow, only quantifying the risk from the power reactor itself.
Other potentially hazardous portions of the nuclear fuel cycle, i.e.,
uranium mining, uranium enrichment, fuel fabrication, spent fuel transportaion, fuel reprocessing, and radioactive waste dis-posal, are not addressed in the RSS analysis.
These are impor-tant omissions when one is comparing energy alternatives.
u-In adcition to the narrow scope, there are other limits to the study's results.
Whole categories of accident-initiating events a.
such as earthquakes and fires were dismissed without adequate analysis.
b.
Sabotage was explicitly not considered.
Tae in~ crease of failure rates as plants age was c.
not taken into consideration.
It was for such reasons that the American Physical Society's Reactor Study team concluded after reviewing the draft RSS that
" based on our experience with problems of this nature involving very low probabilities, we do noc now have confidence in the g
C~
presently calculated absolute values of the probabilities."
In addition, because of the manner of averaging over many s
sites, the RSS results cannot be considered more than a rough approximation of the risk to any population group near a specific nuclear power plant.
C.
Older Ideas Overturned - Probability of Core Melt Very Likelv The RSS has swept aside a number of older and apparently confirmed ideas about the nature and risk of big nuclear acci-dents.
One example of concepts overturned by the RSS concerns the probability of core melting.
Up co the release of the draft version of the RSS the industry and regulatory view was that a core melting accident would have very serious consequences owing.to the near certain likelihood of massive radioactive release.
The low risk to the public was attributed to the extremely low probability of a melting accident.
This proba-bility has variously been estimated to be between 10-6 (one in
~
-8 (one in 100 million) or so per reactor year.
a million) to 10 The RSS has now concluded that these estimates are quite wrong.
The RSS Study Team concluded that there was a chance of 1 in 20,000 for each reactor year of operation of having a melting accident.
They put uncertainty bounds on this; ard
_i according to the RSS, there was an upper limit of one chance in 4,000 per reactor year of having a melting accident.
This is i
actually quite an appreciable probability.
Moreover, the true I
fA
number may even be larger.
The RSS numbers, if accepted, imply that for the Black Fox reactor with a service life of 40 years, there is a chance approaching one percent that it will melt in its lifetime.
It might be higher than that, perhaps by a factor' of two or three more, perhaps by a factor of ten.
I do not know for sure.
We see now that the principal basis on which the smallness of the risk of catastrophic accidene has been officially founded for many years is, acco: ding to the RSS, wrong.
The burden of establishing the safety of the reactor program must now be borne by some other considaration.
This consideratica, according to the RSS, is that a meltdown is, on the whole, innocuous.
RSS goes on to state that additional factors which give risk to large radioactive release (containment breech, vind, population and weather) are among those required to yield catastrophic results.
D.
Accidents.Consecuences and Probabilites The RSS concluded that the principal risk is from accidents in uhich the core of the reactor melts.
The report arrives at the genera' conclusion that the probability of an accident which leads to a maj or dispersal of radioactivity, which in turn exposes thousands of citicens to an increased likelihood of cancer, is 1 chance in 1 billion per year of reactor operation.
This estimate of likelihood is approximately composed as follows:
~
a.
Chance of a meltdown of the 1 chance in 100,000 reactor core to occur per year of reactor c;*
operation
4 b.
Chance of meltdown of the core 1 chance in 10 leading to a large release of radioactivity to the environ-ment Chance of large release of I chance in 10 c.
radioactivity occurring during a time of unfavorable weather d.
Chance of high population 1 chance in 100 density in the direction of the unfavorable winds The worst case accident of a 1000 megawatt reactor, according to the final RSS report, would result in 3,300 early fatalities, 45,000 cases of early illnesses, and 32 billion marks in property damage.
Long-tern health effects from the same accident are estimated as 45,000 latent cancer fatalities, 240,000 thyroid nodules, and approximately 5,000 genetic effects in the first generation after the accident.
The uncertainty range is a factor of 3 to 6.
The preceding is the average effect of the worst Case accident.
E.
RSS Presentation Masks Long-Term Risks -
The RSS report itself, in all its 2200 pages of detail, is virtually impenetrable to all but the professional reader.
The 12 page Executive Summary was, there fo re, prepared for the vast majority of readers who were interested in the conclusions of the report, but didn't have the time or the expertise to read
~
even a significant fraction of the report itself.
Unfortunately, the RSS comparison is deceptive.
The RSS Executive Summary does not show thejmost important consequences
,2
of a reactor accident; i.e.,
long-term health consequences which include many fatal injuries due to an incrase incidence of cancer in the downwind area.
For example, consider the one in 1,000,000 per plant per year probability accident the RSS group showed as resulting in less than 1 prompt fatality.
From Appendix VI of the RSS, and with a little bit of arithmetic, you will find that this acci-dent results also in 6700 latent deaths.
Without explanation about delayed health effects, the risk comparison curves in the Executive Summary which show caly early. fatalities are totally misleading.
Another example is a class of reactor accident that leads to 10 early fatalities (reactor year probability of 3 in 10 million) which would yield the following additional health effects.
10 early fatalities; 1000 early illness (respiratory impairment only),
75,000 cases of hypothyroidism; 75,000 thyroid nodules,- about 2300 fatal thyroid malignancies-9000 latent cancer fatalities; and about 6750 genetic abnor=alities total among live births.
Of these genetic defects 10 to 20*' may be sufficiently serious to produce life shortening, i.e.,
to be in a sense fatal.
The risk comparisons in my view ought to be based on the 12.000 latent fatalities expected rather than on the 10 early fatalities.
2 In ad'dition, there would be over 3000 (8400 km ) square miles of land contaminated with radioactivity above acceptable levels, and over 200 (560 ka ) square miles where people would have to c"
be relocated.
Obviously, the long-term consequences trom a reactor acci-dent are very much greater than the projected short-term conse-quences.
The result is that the RSS risk comparison presenta-tion is highly misleading.
When one adds the long-term fata-lities to the RSS risk assessment curve for man-caused accidents, one sees that the risks from nuclear accidents--even assuming the RSS models are correct--are comparable to those of other man-caused disasters.
It is important to reemphasize I do not believe the RSS probabilities and consequences to be correct.
There is good cause to believe that RSS probabilities are low and that conse-(including both prompt and delayed effects) are under-quences stated.
s F.
Strontium-90 Releases and Water There are a number of differences in the manner that the consequences of a nuclear accident have been presented to the public.
An example is the release of strontium-90 to water.
The consequences of the possible massive contamination of water by strontium-90 on health, water supplies, costs, or interdiction schemes are not adequately addressed in the RSS or by the Nuclear Regulatory Commission.
This is, of course, a significantomission.
~
Since reactors are typically located near large bodies of water and these bodies of water are often important sources of C,
drinking water or food, the consequences of water contamination as a result of a reactor meltdown becomes a vital issue.
The
< typical amount of strontium-90 which is calculated by the RSS to go into their groundwater and subsequently the nearby river e
is now 150,000 curies released over a period of about one year.
The strontium-90 contamination would require over 300 km 3 gg water to dilute the solution to equal the Maximum Permissible Concentrations (MPC).
For co=parison, this is about equal to one
. half the annual average flow of the 51ssissinoi River.
The final RSS still indicates that "the effects of contamination on water supplies has not been considered in detail" (p.76, p.
134) and still assumes without an analysis that streams and rivers would be contaminated for only a "short time" (p. 134).
This assump-tion is very questionable.
G.
Emergency Planninz-Evacuation for Class 9 Accidents The present emergency plans, as determined by the NRC, following a major accidental release of radiation assumes evacua-tion of only the residents living within a few miles of the plant site.
For comparison sake, the introductory mention of evacua-tion in the RSS assumes that evacuation is carried out in all directions to 5 miles from the plant and in a downwind 45 sector to 25 miles.
Specifically, the 30% of the population moves away from the plant at an effective speed of 11.7 km/hr, 40% at 2 km/hr and 30% remain immobilized due to confusion or misdirection (o km Regardless of the initial response to the emergency, all persons are assumed to leave the contaminated area after no more than 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> of exposure to the radioactivity deposited on the ground.
{.,
Several other important observations-from the RSS are that the proba-
4
{
bility of death is 10 times higher for those who are unable to evacuate as compared to those who evacuate at 2 km/hr.
On the other hand, failure to evacuate the contaminated area within 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br /> after cloud passage might increase the nu=ber of deaths
, and illnesses by a factor of 3 or 4.
The differences between the RSS Evacuation Model, and the rir.nned Black Fox nuclear site evacuation plans ard many.
The most significant difference is the " design basis" for the emergency planning.
The regulations that the NRC impo,ses upon the plant owners requires that the plant be designed to accommodate a " design accident", which generally is a guillotine break of the largest diameter pipe in the primary cooling system.
^
The ECCS, if it does its job, will then prevent a radioactive s
release outside the reactor containment itself, and as a result of that, the accident really does not have maj or off-site effect.s.
In contrase, the RSS concluded that the engineered safety features, the emergency core cooling system and those syste=s imposed to cool and depressurize the reactor building might not function during certain categories of severe accidents and that core melt would result.
In addition, in some of these accidents RSS concludes there would be a containment rupture.
The RSS has identified a number of different scenarios by which this could occur.
Traditionally, a melting accident with containment rupture has been referred to as a Class 9 accident.
The major inconsistercy is that the Class 9 accident is not the design accident.
s The NRC has ignored the RSS conclusion that the probab-ility of a Class 9 accident is not negligible.
The result is inadequate regulations concerning evacuation planning.
A major concern in assessing any Evacuation Model is the consideration of the prccesses by which the reactor operator would be likely to make the decision to initiate evacuation or to notify the authorities in the event of an accident.
In major malfunctions that have occurred at nuclear plants all over the world, the experience has been that the people who are being called upon to make important decisions are doing so under extremely chaotic conditions with information of highly ques-tionable quality on which to base the decisions.
s The Browns Ferry fire on March 22, 1975 in Decatur, Alabama, certainly is one vivid example where the evacuation plan would likely have failed.
In this accident a fire rendered the vital Emergency Core Cooling Systems inoperable.
The control room was filling with smoke, and electrical circuits were becoming inoperative as indicating lights and annunciators signaled anomalously.
Fortunately, the plant operators did not panic and flee the control room.
Had they done so, there could have been a radiation release requiring evacuation of the surrounding countryside.
The effectiveness of emergency plans can be assured only by training the public through frequent " mock" drills, where m
these are possible, and mass education.
Just as obviously, there is a strong suspicion that the nuclear' industry and the 11
4 l
reactor owners do not encourage such widespread public partici-pation for fear of raising questions in the public's mind about the safety of nuclear power plants.
In view of demonstrated weaknesses in existing evacuation plans and the psychological differenc e between hurricane or flood threats (on which the RSS studies are mainly based) and a radioactivity alert in which people including officials would not know whether they are already exposed to radiation, it would seem prudent to re-examine the requirements for evacuation at Black Fox based on the present knowledge and state of technology.
H.
Sab e t_at e_
{.
The low probabilities quoted in RSS and assumed by the regulato v bodies for accidents having maior consecuences were cale lated on the assumption that the reacter accidents considered are bona fide -- that is, that they are consequences
'of machine failure or human error, and not the result of deli-berate acts of destruction or sabotage.
These low risks from a nuclear accident followed, according to Rasmussen, from a low inicial probability of an accident, accompanied by a high likelihood of the presence of mitigating circumstances.
A potential group of iboteurs is subj ect to no
~-
such limitations: if they are properly prepared, their. chance of causing a core melt followed by a large radioactivity re-lease may be quite large; they can furthermore choose a reactor
- c. -
uhich is near a large city, and a day when the wind is blowing toward the city.
In brief, saboteurs can frustrate inten-s s
tionally many otheruise citizatinz circumstances.
Thus, unless it is physically very difficult to sabotage a large power reactor, it follows that the dominant risk to the public frc=
nuclear power reactors may well be precisely from such acts of sabotage.
In fact, the US NRC admitted in March of 1976 that there had been at least 173 actual instances or threats of violence against nuclear facilities across the USA since 1969.
Ninety-nine of these incidents were at licensed nuclear facilities.
which include power plants and research reactors.
Explosives were actually found in two cases,a pipe bomb at a research
~,.
reactor, and an undisclosed amount of dynamite outside e v
Wisconsin-Michigan Power Company plant.
Knowledgeable persons could with relative ease sabotage a large reactor with massive consequences.
There are innumerable ways by which the critically needed cooling uater can be cut off, either by valving, disruption of power supplies or by physical danage to the pressure containing system.
The saboteur could set up conditions that would irreversibly lead to a maj or accident, and still have several hours to escape its affects.
To quote the recently completed Australian nuclear power environmental inquiry (often referred to as the " Ranger Report")
5 v; on the subj ect of sabotage of nuclear facilities -
,u "It needs to be reiterated that no form of attack on a thermal nuclear reactor can result in an atomic
~
explosion, but the possibility rc=ains that through skillful sabo tage, the reactor site could be damaged and the surrounding coc= unity could be seriously harmed by the release of great quantities of radiation."
Such sabotage atte= pts could be directed at the reactor system itself.
Another area of vulnerability is the discharged fuel storage pools.
Again, quoting the Australian report:
"There are opportunities for causing great dacage by sabotage of nuclear installations other than re-actors.
Cooling ponds for spent fuel at reactor sites or reprocessing plants contain highly radioactive material which, if dispersed by a large explosion, would cause serious local hazards."
This risk of sabotage is currently being amplified by the lack of facilities and plans for the disposition of spent fuel.
Almost all reactor sites are currently expanding (by factors of two to four) their on-site spent fuel storage capacity.
Such s
plant modifications will greatly increase the amount of fuel which may be stored on site.
Shortcomings in plant security have been identified, but cocplete protection is impossible._ Again to quote the Ranger Report:
"As with measures for preventing theft of nuclear materials, there have been recent coves to improve the protection of nuclear facilities from sabotage.
Nevertheless, the Ranger Commission formed the opinion that a determined and well-organized attack, especially if assisted by a nuclear expert, could be mounted with a realistic chance of causing very serious damage."
~-
The US NRC conducted an investigation of the status of safe-guards and reached the following conclusions in their report, Strategic Special Nuclear Material, published in February 1977.
es 8
e I
- Safeguards planning must necessarily be based upon hypotheses s
of m tivations and esticates of capabilities.
The pri=ary deter-minants of threat capabilities for theft are the numbers of persons involved, their access to positions of trust with respect to SSNM, the availability of ar=s or special equipment that might facilitate a theft, and their willingness to sacri-fice lives to achieve success.
Historical data on the nu=bers of persons involved in groups cocaitting robberies, assaults, and acts of terrorisa in this and other countries show that s=all groups predccinate.
Groups larger than six persons account for only a feu percent of the cases, although there are isolated instances of very large groups.
Case studies of elaborate crimes or " capers" indicate that group size and composition are dictated by the demands of the job and not by the availability of people.
Such crimes are generally characterized by careful planning and avoidance of violence.
There is some evidence to support the notion of an inverse correlation between group size. and the degree of planned violence.
The possibility of insiders participating in or instigacing the thef t of nuclear materials is considered by many to be more likely than an armed assault by outsiders.
TF2f ts in commerce frequently involve insiders, and internal conspiracies are not uncommon in the hij acking or diversion of commercial freight.
The availability of sophisticated armaments to individuals or criminal groups.is changing rapidly.
Automatic rifles and
plastic explosives are now widely available.
Modern, high-technology weapons, such as wire guided and heat-seeking missiles, recoilless cannons, helicopters, etc., cannot be excluded as possibilities even though they may be unlikely based upon current evidence.
Thus, the evidence of current threat capabilities points toward the possibilities of more than a single insider acting alone or with a small group of persons armed with legally ob-tainable weapons.
With respect to internal threats, high confidence protection against any single insider, regardless of position or clearance, would appear to be a minimum safeguards posture.
The most vexing problem for the industry will be insuring against the insider being a key individual in the security organization.
Prudent safeguards system design would make maximum use of technology and procedures to prevent conspiracy wherever prac-ticable, even where personnel have been cleared.
In any event, clearances should not be deemed adequate insurance against the theft of SSUM by any single insider, regardless of position or trust.
With respect to external threats, high-confidence protection against robbery by a small group of about three persons with in, side. assistance and armed with legally obtainable weapons would appear to be a minimum safeguards posture.
If the indus try safeguards posture must be upgraded agains t
?
the threat of determined violent assaults, then it is evident
<that there is come level of force beyond which high-confidence protection cannot be provided by any practical means within institutional arrangements.
With group sizes substan-current tially in excess of about six persons and with arms much beyond automatic rifles, almost any entity in the nation may be con-sidered susceptible to a. determined violent assault.
Insuring the protection of facilities against determir.ed violent assaults with automatic small arms and explosives for breaching barriers will require substantial changes in the current nuclear facilities.
These would include greatly en-larged and better armed and trained guard forces, additional barriers, alarms and protected guard positions.
Such additions would involve substantial capital and operating costs; and would alter the appearance of the plants toward that of' military reservations.
a w
p.
~
~