ML20083N682
| ML20083N682 | |
| Person / Time | |
|---|---|
| Issue date: | 03/31/1982 |
| From: | NRC |
| To: | |
| Shared Package | |
| ML20083N681 | List: |
| References | |
| FOIA-82-551 NUREG-0773, NUREG-0773-DRFT, NUREG-773, NUREG-773-DRFT, NUDOCS 8302020483 | |
| Download: ML20083N682 (107) | |
Text
_ _.
..;p t
'l
,M NUREG-0773 i
DRAFT REACTOR ACCIDENT SOURCE TERMS:
DESIGN AND SITING PERSPECTIVES March 1982 l
I I
l l
8302020483 821202 PDR FOIA ELLIOTT82-551 PDR i___._-
l 4
Table of Contents Page Introduction 1
Historic Perspective of Regulatory Practice and Source Term....
3 Environmental Impact Assessment 6
The Reactor Safety Study...................
7 Accident Descriptions.......................
8 Uncertainties and Improvements in Reactor Safety Study 25 Rebaselining of Reactor Safety Study Results 28 Probabili: tic Risk Assessments (Post WASH-1400)..........
48 Indian Point Risk Comparison Study 53 Consequences of Major Accident Release Sequences 57 Ri s k C u rv e s............................
62 Risk Insights Into Reactor Design.................
90 Siting Source Terms........................
96 References 104 I
i i
t
~.
L
-~
INTRODUCTION In establishing the basis upon which criteria can be formulated to license the construction and operation of nuclear power plants, it is necessary to obtain a perspective of the spectrum of probabilities, magnitudes and characteristics of release of radioactive materials associated with potential reactor accidents and the associated consequences.
Such a perspective should provide a consistent foundation for the establishment of meaningful regulations upon which assurance of no undue risk to the public health and safety can be made.
This report presents the current available information on the potential reactor accidents that have been analyzed for various reactor designs, and develops a set of potential radioactive releases (source terms) which will serve to represent the spectrum of accidents. This set of accident source terms could then be used to form the basis for the development of consistent regulations for siting, emergency planning, minimum engineering safety features and degraded core cooling. The source term estimates represent our current knowledge and understanding of a very complex problem as of about early 1981. These es*.imates should be taken as " state of art," however, it must be fully understood that there are large uncertainties associated with these estimates.
The uncertainties associated with the source term are discussed in NUREG-0772,
" Technical Bases for Estimating Fission Product Behavior During LWR Accidents."1 Recommendations that are made in that report have not been incorpoated in the source term estimates presented.
Mcwever, until revised physical process a'nd source release codes, i.e., MARCH and CORRAL, have been reissued, better quantitative estimates on the accident sequence source terms cannot be made.
v
. -,,... - -. ~,,
-,._m.
-,-m..__7-
__,_.,m,
-.,,.,m_,
m.,,
.,..._m.
It is expected that there will be significant source term reductions at least for those sequences that have longer times associated with their development; such as for small LOCAs, or where the particulate material must travel tirough long paths in which water would be present; such as in many transient. events.
Estimates of the impacts on the NRC regulatory process of the changes in
~
source term assumptions are presented in NUREG-0771, " Regulatory Impact of Nuclear Reactor Accident. Source Term Assumptions.u2 In addition, research programs are in progress to gain better insight into the physical processes involving the transport and release of the radioactive material from the reactor to the environment.
Much of the information presented 1.s this report reflects a collation of results available from various LWR studies using ;he probabilistic risk assess-ment (PRA) techniques which were iritially developed for use in the Reactor and subsequently u'pdated.
Risk variation as a function of 3
Safety Study various LWR designs and design features are displayed and these results are j
subject to future changes and refinements as on going PRA based studies from which they derive are completed (e.g., the Reactor Safety Study Methc' ology Applications Progrsm, and the Integrated Reliability Evaluation Program).
It is, howev^r, our cu.rrent judgment that future changes and refinements by these l
l PRA studies are unlikely to significantly alter these overall risk perspectives that are useful and relevant to siting and design, analyses, and to agency 1 risions and policies.
(
From the collated PRA results, this report endeavors to synthesize a set of source terms that would cover a spectrum of severe core damage accidents occurring across a population of existing LWR designs that use widely different designs and safety features. This set of siting source terms represents our 2
--._. --,..,------,--~,,,
f 4
current best judgments about a set of source terms that woul'd cover most of i
the current L'WR designs. These source terms could be used, esser.tially inde-pendent of particular plant designs, for additional' comparative siting analyses.
It should be stated, however, that insights available from various PRA studies and study groups (n.b., NUREG-0715, " Task Force Report on Interim Operation of Indian Point"4) suggest that risk variations associated with.WR designs can be large (e.g., order of magnitude or more) and that the risks.associatea with more populated sites can be offset by favorable design features.
It is expected that planned ruleaakings will recognize these LWR design variations and a?ti-mately assure that accident risks will fall considerably within those that would be projected by use of the set of sca a terms set forth in this report.
Historic Perspective of Regulatory Practice and Source Terms Since the earliest days of reactor power plant development, attempts have been made to define the probabilities, source terms, and consequences associated with u ;.ential reactor accidents.
In 1957, Brookhaven Nationel Laboratory published WASH-740, " Theoretical Possibilities and Consequences of Major Accidents in large Nuclear Power flants."5 That report presented three reactor accident scenarios as being " typical" cases. The first scenario, "the contained case," is one in which no activity is released but a gamma shine do,e from the i
contained source is calculated.
In the second scenario, "the volatile release case," signf'icant. fractions of the volatile fission products (e.g., noble gases, halogens) are released.
The third scenario, "the 50 percent release case," is an accident in which 50 percent of all fission products in the reactor are released from the containment to th atmosphere.
No explicit probabilities were assigned to the'se scenarios except that a probability range
~
3
of between one in 100,000 to one in a billion was discussed.
Calculated conscquences ranged from none to 3,400 fatalities, 43,000 injuries, and 2.3 billion dollars in property damage.
With WASH-740 in mind, in 1961 regulations for site selection were develope'd as part of 10 CFR Part 100, " Reactor Site Criteria."8 In conjunction with Part 100 the concept of a maximum credible accident was developed as the mechanism to evaluate the acceptability of the potential site. The Maximum Credible hccident concept was devised to evaluate siting limits and containment design requirements.
In 1962, the maximum credible accident was codified in TID-14844, " Calculation of Distance Factors for Power and Test Reactor Sites."7 The TID source term, as it is known, p;stulated a loss-of-coolant accident (LOCA) upon complete rut re of a major coolant pipe, followed by a meltdown of the fuel and partial release of fission product inventory to the atmosphere of the reactor building (containment). One hundred percent of the noble gases; fifty percent of the radioiodines; and one percent of the other parti-culate materials (solics) in the fission product invento-" was assumed t-be released from the fuel. The containment was assumed to leak at one tenth of one percent (0.001) per day, indicating that the containment structure was av med to be fully effective in meeting its design leak rate. However, any further deterioration in leakage, or the possible dispersion of the solid fission products was neglected.
Since Part 100 allowed engineered safeguards to offset less favorable site characterictics, these release assumptions also came to be used in design of safety systems engineered to mitigate the release of fission products to the environs.
As this practice evolved, the assumed iodine releases into the l
4
t a
containment atmosphere were recognized as being conservative but this was assumed to compensate for the uncertainty involved in p;ojecting the course of an accident and the release and transport of fission products to the environment.'
A decade later (early 1970s), a design bases was established for control of hydrogen evolved by metal-water reaction based on an ass'umption of localized overheating of the core, but not melting.
This basis was modified (reduced) some years after the issuance of Appendix K to Part 50 (see Par. 50.55 and R.G.1.7), but still assumed only that localized overheating of the core could result.
The increased thermal margins provided oy the regulations led to definition
~
and refinement of design bases for many systems, parti.*ularly for auxiliary systems, which by assuming no core damage either minimized the safety signi-ficance of certain systems, e.g., the auxiliary feedwater systems or failed to fully recognize conditions for which they should be designed and qualified.
This resulted in a rather stylized nonmechanistic analysis of the release of fission product iodine and this approach was later incorporated into Safety Guides 3 and 4 (subsequently renamed as Regulatory Guide 1.3 anu 1.4).
These regulatory guides have incorporated a reduction of the iodine source term by a factor of two conservatively acknowledge natural deposition pressures for elemental iodine on primary system and ccitainment interior surfaces.
In addition, these regulatory guides codified staff ssumptions particulate forms of radionuclides as discussed above (this also included organic iodines) then in vogue.
The various behaviors and formations of the iodines (e.g.,
particulates.or organic and elemental forms).
5 1
Environmental Impact Assessment.
In 1971 the Atomic Energy Commission issued for comment, a set of assumptions for a " realistic" asses.sment of the environ-mental impact o' accidents at nuclear power reactors, pursuant to the requirements of the National Enviror.untal Policy Act of 1969. These assump-tions, specified in a proposed Annex to Appendix D of 10 CFR Part 50 includ'ed a system to be used classifying accidents according to a graded scale of severity and probability of occurrence. These were based largely on staff
~
judgments at this time and did not have benefit of more systematic and rigorous results from use PRA techniques that were to come nearly a. half decade later.
Nine classes of accidents were defined in Appendix 0, ranging from trivial'to very serious.
It directed that "for each class, except classes 1 and 9, the environmental consequences shall be evaluated as indicated." Class 1 events were not to be considered because of their trivial consequences.
Class 9 events were recognized to have the potential for severe consequences, but their probability of occurrence was considered to be so low, that their risk for the environment was low enough not to be considered in environmental l
impact assessment.
Although this classificatio:. scheme was intended for l
environmental statements only, the term " class 9" accident has been widely used, and is often, associated with core melt accidents having severe consequences.
It should be noted, however, that t N proposed Annex defined this category of accidents as those sequences of " postulated successive failures more severe than those postulated for the design basis for protective systems and engineered safety features." In other words, multiple failure scenarios wh'm could result in severe core damage accidents were beyond the single failure critera set forth in the 10 CFR 50 General Design Criteria and need not be assessed for either the environmental or safety review purposes.
l 6
e In contrast to the intentionally more conservative treatment of accidents in the NRC's Safety Evaluation, (e.g., maximizing releases for the accidents cer.sidered) the environmental impact assessments were intended to be realistic (i.e., best estimate).
A comparison of similar accidents, therefore, provides someindicationofthestaff'sperceptionofthedegreeofconservatisminckuded in its Safety Evaluations. This comparison, however, would be based only on one part of the risk equation, that is the consequence part; the other important part of the risk equation, i.e., probabilities of release, would be missing and could distort such a comparison.
- The Reactor Safety Study.
In 1975, the Reactor Safety Study (WASH-1400)s was published.
The study revealed that the offsite risks from most reactor acci-dents was small, and that accidents more severe than those design basis accidents then in use by the NRC's licensing staff tor siting and plant design purposes would dominate the risk. The largest risk was associated was only a core melt, but deterioration in the capability of the containment to limit the release of radioactive materials to the environment.
This study was the first systematic attempt to search out a large spectrum of accidents and use quanti-tative assessment techniques to estimate the probabilities, source terms, and public consequences in an integral way.
Event tree and fault tree reliability assessment techniques were used to evolve the accident sequences and successive failures to result in core damage and to assess the probabilities associated with such sequences.
In addition, models of the physical processes associated with particular accident sequences over the spectrum identified were developed to assess the magnitudes and timings associated with the release, transport, and deposition of the rcdioactive materials from the core, through the primary system and containment and into the environment.
Consequence models were also 7
O O
developed to define the dispersion and impacts of radioactive materials in the environment and to assess the distr'ibution of risk (probabilities and consequences) associated with such accidents.
Accident Descriptions Two specific reactor designs were analyzed in the WASH-1400:
a 3-loop Westinghouse Pressurized Water Reactor (PWR) with a dry containment (Surry);
and General Electric Boiling Water Reactor (BWR) with a Mark I, vapor sup-pression containment (Peach Bottom).
A spectrum of accident release characteristics (source terms) and associated probabilities was generated far each reactor.
Nine PWR release categ3 ries and five BWR release categories that were developed '.o cover these core damage accidents identified by the study are presented in Tables 1 and 2, respectively. The information in these tables summarizes the inputs required to characterize a release of radioactive material to the environment.1 Time of release represents the time from the initiation of the accident (reactor shutdown) to release to the environment.
Duration of release refers to the time during which the significant majority of the radioactive materials would be released to the environment.
Warning time is defined as that time before the release to the environment that the plant ooeraoors would recognize that the plant was in serious trouble a'd n
notify the offsite authorities to initiate an emergency response. The release height and energy associated with the release are parameters in characterizing, the manner in which the radioactive materials would be discharged into the atmosphere for subsequent dispersion.
1 iThese inputs are also described in Appendix V of WASH-1400.
8
e
.h f
i TABLE 1 REACTOR SAFETY STUOY PWR ACCIDENT RELEASE CATEGORIES (SURRY) b
' nfction of Core Inventory Released Accident Prob.
Time' Duration Warning" Enerp i
I Xe-Kr l'
Cs-Rb Te-4.
Ba-Sr Ru' ta Sequence Per (hr)
(hr)
(hr) 10 Reactor Stu/hr Year PWR 1 9 x 10 7 2.5 0.5 1.0 20 & 520 0.9 0.7 0.4 0.4 0.05 0.4 3x10 8 PWR 2 8 x 10 8 2.5 0.5 1.0 170 0.9 0.7 0.5 0.3 0.06 0.02 4x10 8 PWR-3 4 x 10.s 5.0
- 1. 5 2.0 5
0.8 0.2 0.2 P.3 0.02 0.01 3x10 8 PWR-4 4 x 10 7 2.0 3.0 2.0 1
0.6 0.09 0.04 - 0.01 5x10 8 3x10.a 4x10
- PWR-5 7, 10 7 2.0 4.0 1.0 0.1 0.3 0.03 9x10 8 5x10.a 1x10 8 6x10 4 7x10.s PWR-6 6 x 10 8 12.0 10.0 1.0 N/A 0.3 8x10 4 8x10
- 1x10 8 9x10 5 7x10 5 1x10 8 f
PWR-7 4 x 10 5 10.0 10.0 1.0 N/A 6x10 8 2x10 5 1x10.s 2x10 5 1x10.s 1x10.s 2x10 7 PWR-8 4 x 10 5 0.5 0.5 N/A N/A 2x10 3 1x10
- 5x10 4 1x10 8 1x10.s 0
0 PWR-9 4 x 10 4 0.5 0.5 N/A N/A 3x10.s 1x10 7 6x10-4x10 e 1x10 28 0
0
- Time interval between start of hypothetical accident (shutdown) and release of radioactive material to the atmosphere.
lotal time during Which the major portion of the radioactive esterial is released to the atmosphere.
b Time interval t<.seen recognition of lopending release (decision to initiate pub 11 protective esasures) and the C
release of ra'aicactive material to the atmosphere.
organic lodine is combined with elemental lodines in the calculations, Any error is negilg Mle since the release fraction is d
relatively small for all large ralease categories.
' Includes Ru, Rh, Co, Mo, Tc.
I Includes Y, La, Zr, Nb, Ce, Pr Nd, Np, Pu, Am, Ca.
f
l TABLE 2 REACTOR SAFE 1Y STUDY BWR RELEASE CATEGORIES (PEAcil BOTTOM) l I
b Irob.
Durat on War g*
En Fraction of Core Inventory Released f,, f I
p, Xe-Kr 1
Cs-Rb Te-Sb Ba-Sr Ru' La Reactor Stu/hr Year BWR 1 1 x 10.s 2.0 0.5
- 1. 5 130
- 1. 0 0.40 0.40 0.70 0.05 0.5 5x10 8 BWR 2 6 x 10.s 30.0 3.0 2.0 30 1.0 0.90 0.50 0.30 0.10 0.03 4x10.s BWR-3 2 x 10.s 30.0 3.0 2.0 20 1.0 0.10 0.10 0.30 0.01 0.0.2 4x10 3 BWR-4 2 x 10 8 5.0 2.0 2.0 N/A 0.6 8x10
- 5x13.a 4x10 8 6x10
- 6x10
6x10 88 4x10 8 8x10 88 8x10 8*
0 0
" Time interval between start of hypothetical accident (shutdown) and release of radioactive material to the atmosphere.
bTotal time during which the major portion of the radioactive material is released to the atmosphere.
' Time interval between recognition of impending release (decision to initiate pubitc protective measures) and the release of radioactive material to the atmosphere, arganic lodine is combined with elemental lodines in the calculations. Any error is negligible since the release fraction is d
relatively small for all large release categories.
' Includes Ru, Rh, Co, Mo,'ic.
I 1ncludes Y, La, Zr, Nb, Ce, Pr, Nd, Np, Pu, Am Ca.
W t
0
A short description of each of the release categories (taken from Appendix VI of WASH-1400) is repreated below to help characterize the phy:,ical processes associated with the postulated containment failure mechanisms.
More detailed analyses and description of these accidents and associated processes are given in appendices V, VII and VIII of WASH-1400.
PWR 1 This release category can be characterized by a core meltdown followed by a steam explosion resulting from contact of molten fuel with the residual water in the reactor vessel. The containment spray and heat removal systems are also assumed to have failed and, therefore, the containment could be at a pressure considerably above ambient at the time of the steam explosion.
It is assumed that the steam explosion would rupture the upper portion of the reactor vessel and breach the containment barrier, with the result that a substantial amount of radioactivity might be released from the containment in a puff over a period of about 10 minutes.
Due to the sweeping action of gases generated subsequently by reuctor vessel melt-through and during containment-vessel melt-through, the release of radioactive materials from the leached containment would continue but at a relatively lower rate.
The total release was estimated to contain approximately 70 percent of the iodines and 40 percent of the alkali metals present in the core at the time of release.1 Because the contain-ment w'ald contain hot pressurizc ' gases at the time of failure, a relatively high release rate of sensible energy from the containment could be associated 1The release fractions of all the chemical species are listsi in Table 1.
The release fractions of iodine and alkali metals are indicated here to illustrate the variations in release with release category.
11
^
~
with this category. This category also includes certain potential accident sequences that would involve the occurrence of core melting and a steam f
explosion after containment rupture due to overpressure.
In these sequences, l
the rate of energy release would be somewhat lower, although still relatively high.
PWR 2 This category is extended to cover the failure of core-cooling systems and core melting concurrent with the failure of containment spray and heat-removal systems.
Failure of the containment barrier would largely occur through overpressure such as and might occur and to less of all ESFS by loss of electric power systems,1 causing a substantial fraction of the containment atmosphere to be released in a puff over a period of about 30 minutes.
Due to the ' sweep-ing action of gases generated subsequently by reactor and containment vessel melt-through, the release of radioactive material would continue at a relatively lower rate thereafter. The total release would contain approximately 70 per-cent of the iodines and 50 percent of the alkali metals present in the core at the time of release. As in PWR release Category 1, the high temperature and pressure within containment at the time of containment failure would result in u
a relatively high release rate of sensible energy from the containment.
l IPWR 2 category was also intended to cover pressure systems (e.g., low pressure ECCS) located outside containment.
In such sequences the ccre is assumed to melt with the releases essentially bypassing the containment and containment mitigating systems.
12
PWR 3 This category involves an overpressure failure of the containment due to failure of containment heat removal which in turn interacts with and fails core coding systems.
Containment failure would occur prior to the commence-ment of core melting. Core melting would then cause radioactive materials to be released through a ruptured containment barrier.
It is estimated that approximately 20 percent of the iodines and 20 percent of the alkali metals present in the core at the time of release would be released to the atmosphere.
Most of the release could occur over a period of about 1-1/2 hours. The release of radioactive materials from containment would be caused by the subsequent meltdown processes and by the sweeping action of gases generated by tha reactiori of the molten fuel with concrete.
Since these gases would be initially heated by contact with the melt, the rate of sensible energy release to the atmosphere would be moderately hty5.
PWR 4 This category involves failure of the core-cooling system and the containment spray injection system after a loss-of-coolant accident, together with a con-current failure of the containment system to properly isolate.
This would result in an estimated release of almost 9 percent of the iodines and 4 per-cent of the alkali metals present in the core at the time of release.
Most of the release would occur continuously ever a period of 2 to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />.
Because of the containment recirculation spray and heat-removal systems are assumed to operate to remove heat from the containment atmosphere during core melting, a relatively low rate of release of sensible energy would be associated with this category.
13
I PWR S This category involves failure of the core cooling systems and containment isolation and it is similar to PWR release category 4, except that the contain-ment spray injection system would operate to further reduce the quantity of*-
airborne radioactive material and to initially suppress containment temper-ature and pressure. The containant barrier would have a large leakage rate due to a concurrent failure of the containment system to isolate, and most of the radioactive material would be released continuously over a period properly l
of several hours.
Approximately 3 percent of the iodines and 0.9 percent of the alkali metals present in the core is estimated be released in this cate-gory of accidents.
Because of the operation of the containment heat-removal systems, the energy release rate would be low.
PWR 6 This category involves a core meltdown due to failure in the core cooling systems after a LOCA or transient initiating event. The containment sprays were assumed in effective in mitigating the radioactive material released into the containment, but the containment barrier would retain its integrity until the molten core proceeded to melt-through the concrete containment base mat.
1 In this category the continment pressure would remain relatively high but below the estimated failure pressure. The radioactive materials would thus be suddenly released into the ground, with some leakage to the atmosphere occur-ring upward through the ground with most of the atmospheric release being noble gases not mitigated by the soils.
Direct leakage to the atmosphere would also occur at a los rate prior to containment-vessel melt-through.
It 14 r
~
was also assumed that this direct leakage occurred at s1*/ day. Most of the release would occur continuously over a period of about 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />.
The release would include approximtely 0.08 percent of the iodines and alkali metals pre-sent in the core at the time of release.
Because leakage from containment to the atmosphere would be low and gases escaping through the ground would be '
cooled by contact with the soil, the energy release rate would be very low.
Because containment sprays would operate and core melting would not occur, the energy release rate would also be low.
PWR 7 This category is similar to PWR release category 6, except that containment sprays would operate to reduce the containment temperature and pressure as well as the amount of airborne radioactivity.
The release would involve 0.002% of the iodines and 0.001% of the alkali metals present in the core ut the time of release. Most of the release would occur over a period of 10 hours1.157407e-4 days <br />0.00278 hours <br />1.653439e-5 weeks <br />3.805e-6 months <br />. As in PWR release category 6, the energy release rate would be very low.
PWR 8 This category approximates a PWR design basis accident (large pipe break),
except that the containment would fail to isolate properly on demand.
The other engineered safeguards are assumed to. function properly. The core would not melt. The release would involve approximately 0.01% of the iodines and 0.05% of the alkali metals. Most of the release would occur in the 0.5-hour period during which containment pressure would be above ambient.
Because 15
cantainment sprays would operate and core melting would not occur, the energy release rate would also be low.
PWR 9 This category approximates a PWR design basis accident (large pipe break), but with about a 10 fold deterioration in the Containment design leakage rate, in which only the activity initially contained within the gap between the fuel pellet and cladding would be released into the containment.
The core would not melt.
It is assumed that the minimum required engineered safeguards would function satisfactorily to remove heat from the core and containment.
Because of the subatmospheric features of the containment the release would occur over about a 0.5-hour period during which the containment pressure would be above ambient. Thereafter, the leakage would be very small. Approximately 0.00001 percent of the iodines and 0.00006 percent of the alkali metals would be released. As in PWR release category 8, the energy release rate would also be very low.
BWR 1 This release category is representative of a core meltdown followed by a steam explosion in the reactor vessel. The latter would cause the releases of a substantial quantity of radioactive material to the atmosphere. The total release would contain approximately 40 percent of the iodines and alkali metals present in the core at the time of containment failure.
Most of the releases would occur over a 1/2 hour period.
Because of the energy generated l
in the steam explosion, this category would be characterized by a relatively 16
high rate of energy release to the atmosphere. The category also includes certain sequences that involve overpressure failure of the containment prior to the occurrence of core melting and a steam explosion.
In these sequences, the rate of energy release would be somewhat similar than for those discussed above, although it would still be relatively high.
BWR 2 This release category is representative of a core meltdown resulting from a transient event in which decay-heat-removal systems are assumed to fail.
Containment overpressure failure would result, and core melting would follow.
Most of the release would occur over a period of about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />. The contain-ment failure location would be such that radioactivity would ersentially be released directly to the atmosphere without significant retent% n of fission products within the negative pressure confinement building which surrounds most of the BWR primary containment.
This category involves a relatively high
~
rate of energy release due to the sweeping action of the gases generated during core meltdown pressures. Approximately 90 percent of the iodines and 50 percent of the alkali metals present in the core are estimated to be released to the atmosphere.
l BWR 3 l
This release category represents a core meltdown sequences caused by transient j
events accompanied by a failure to scram or a failure to remove decay heat.
Containment failure cold occur either before core melt or as a result of gases and steam pressures generated during the core meltdown ana reactor-vessel i
17
'~~
O melt-through.
Some fission product retention would occur either in the suppression pool or the reactor building prior to release to the atmosphere.
Most of the release was assumed to occur over a period of about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> and is estimated to involve rele;.se 10 percent of the alkali metals to the atmosphere.
2 For those sequences (e.g., loss of decay heat removal) in which the contain ment would fail due to overpressure before core melt, the rate of energy release to the atmosphere during the subsequent core melt, would be somewhat smaller, although still moderately high.
BWR 4 This release category is representative of a core meltdown with large enough containment leakage to the reactor building (e.g., due to failure of contain-ment to isolate) to prevent containment failure by overpressure. The quantity of radioactivity released to the atmosphere would be significantly reduced by transport paths in the reactor building and by potential mitigation by the secondary containment ventilation and filter systems.
Condensation in the containment, in the reactor building, and the action of the standby gas treat-ment system on the releases would also lead to a low rate of energy release.
The radioactive materials are assumed to be released from the reactor building or the stack at an elevated level. Most of the release would occur over a 2-hour period and would involve approximately 0.08 percent of the iodines and 0.5 percent of the alkali metals.
m
BWR S This category approximates a BWR design basis accident (large pipe break) in which only thc activity initially contained within the gap between the fuel pellet and cladding would be released into containment and partly retained in
,the pressure suppression pool. The core would not melt, and containment leakage would be small.
It is assumed that the minimum required engineered safeguards would function satisfactorily. The released activity from con-tainment to the reactor building would be filtered and would pass to the atmosphere through the elevated stack.
It would occur over a period of about 5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> while the containment is pressurized above ambiant and would involve approximately 6 x 10 8 percent of the iodines and 4 x 10 7 percent of the alkali metals.
Since core melt would not occur and containment heat removal systems would operate, the release to the atmosphere would involve a negligibly small amount of thermal energy.
As discussed above, Tables 1 and 2 were generated by grouping the accident sequences into release categories based upon the similarity in system failures and in magnitudes and characteristics of release and the physical phenomena involved. Tables 2 and 4 give the dominant release sequence groupings and key I
to the symbols for the PWR categories. Tables 5 and 6 give the same informa-tion for the BWR categories. Two important conservative assumptions were made in generating the release categories containing the many individual accident sequences.
First, a probability " smoothing technique" was used across release categories. The probability that was assigned to each category was not only the combination of all the sequences that would be calculated to have similar release characteristics but was computed by adding 10 percent of the probability 19 t
t TABLE 3 - PWR DOMINANT ACCIDENT SEQUENCES VS. RELEASE CATEGORIES PWR DOMINANT ACCIDENT SEQUENCES vs. RELEASE CATEGORIES
- ~ ~ " " -
.mes c.,
um c
-a
-c a
i e
i e
i
'~
d
" H.a.u
- d.us aus.
==-e
==-,
i
-e
^~ -u
- w
-e am um
.a a
a -e
.m na -u u-= uim tu.-"
an sue au
- a.
- d d
-a ar-.
d
"~.u
- a
- .'m.u
-e use.
m ice sue
.us.o ad.u.-.
u4
- i. -'
a.-*
.~
u -'
s.m
~
a e
uu aceu ac,..
'i~,-.
=c:.u ac,
=c.i. =>~
=1;;,0.,
=>,,,0 ac.a 'c..
- ce-.
'ce ace-. ace-.
\\
s~
- a s,~.
s,
.a e,"un ia ma
=>~ -ie
- ,"as
_m.
=
ae 4
s-a.-
e.a.-
6 4.--
30~
.ae as-
)
e 4
e, uu 8
I 3
-le 3
IU -G
'2
+32 4
2 4
2 4
8 g
S CDar A ed S te,,
8,94 4 g
4 g
4 S e4,g SM g
taae asas taae
&a&G 3sse esae g
g 4
s,ce-. e s,=4.ie s,r-8.,
8:"nue..
um.
um um 8
8 ** *ge 3
4 3
Jane gene 5 4 OC4 2
g tsaa.
s Saae nue '
mac' mm*
um' ma" ua*
l ame
a, e uuu 94
,9 K*T 9-.
33
.gg taas,
naae 3a&G Inas asaem m ar-6
,gg survens e e 1s44
'C4
-u L.ae
- II laae""
taae 3m44*
1aa.
' nat.
Sane e Peu.a.444u has n
N
' esas 4
ersvuus asua N3mm embygg. e
-e ot ae en&G*I esM
,,Ag
,gg T pre.e.61asi a M,gg M* y, Messe,g N's,,
Me'eg, M,,
Ema Paae saae Sans amas e.as gen BWF fuse
- 4 m,,
M.40 g, W =e pruar
- 9 3a&G has 3
3 mas W=e
-e a.as 4 48 d
was
- a.ae mas *
.an'*
d d
ma.
- a.an v:
- uu n ne cri _
er an. -w empasse em amanse enemmer m
Seas '
taaS*
e.as esas eene eene tist M taae" emne
.eae d
d enas'*
emas
- d enas*
smas"'
ame tene Ename emas"*
e.as '
e an E,=nem enas"*
e.as
- e.as n.as"*
ae amae
,,,,-e
,,,,-e
,,,,.s 4
d
-e
.-se
- r..f one w.sendet. s, esse eau e my w. *a.
- e..a. Ast i.u e.
a e
e en.nasta m pas.
'-fe.unem.
M seam p.ama e.
P=as t '.y age e.&3 e
20 h
TABLE 4 - KEY TO PWR ACCIDENT SEQUENCE SYMBOLS Intermediate to large LOCA.
A Failure of electric power to ESFs.
B Failure to recover either onsite or offsite electrical power within B'
about 1 to 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> following an initiating transient which is a loss of offsite AC power.
Failure of the containment spray injection system.
C Failure of the emergency core cooling injection system.
D Failure of the containment spray recirculation system.
F Failure of the containment heat removal system.
G Failure of the emergency core cooling recirculation system.
H Failure of the reactor protection system.
K Failure of the secondary system steam relief valves and the auxiliary L
fet2 Water system.
Failure of the secondary system steam relief valves and the power M
conversion system.
Q Failure c' the primary system safety relief valves to reclose after opening.
Massive rupture of the reactor vessel.
R 5
A small LOCA with an equivalent diameter of about 2 to 6 inches.
1 A small LOCA with an equivalent diameter of about 1/2 to 2 inches.
S 2
Transicat event.
T LPIS check valve failure.
V Containment rupture due to a reactor vessel steam explosion.
a S
Containment failure resulting from inad.:quate isolation of containment openings and penetrations.
1 Containment failure due to hydrogen burning.
Containment failure due to overpressure.
6 c
Containment vessel me'..-through.
21
>-x Oa wH 5
w b
w oc 2
g 5
m a
-1" g
7 t'3 T
~
8 3
3 3j w
4 y
a 3
H b
I ]',
r 8
f f
f t
i l.
2.s 12
?
v27: rt :2 t
12 :2 72 72 12
?s s
s s
s s
a s s a :.s s er as e: as us 51
,d,
=
z
'5 s... 5 :
i u
a r: : : :
s a
a a
s w>
n
.[ I,!
w a
lis uu lb {
5 NF n
- e v
- e, e e v
e e t
- ; e l
I! ;-
Y a
d u
a-t-
t r 3.,. 3.. 3 :r
- a :
v v 2 2 5 a e e e.-
3.e 3
[3 r3 3
.:3 4li.
3 E
3 3
3
- .: e.313 v3 63 3 63 ta
-g:
r3 r r: 13 g
e p*
N u
3 4
w u.J e
- tis, u.
s.-,
a o
~
e'
- 3*
() laa
?.
T.
T. t. ?.
. T.
T.. t.. ?..t.
. T.
t. T.. T.,7
- f.,t.
II, 5: n-
~
.uj W y) m 2 2 7
o T.
u s.3 :s.2 rg.ra g3 3
p, - -
3.3 3
3 3
3 3 _,:
. j 'c3 73.ru r: 63 r
i.
s.s,3,3 r3
- s 3.,. - s ~.s.-
e,.:
u u a a
z
- 1:e i
d =.
m l
g g "Lg a
S S S S CY N
S S S
=
w x=
ieA v3 {.
(..,..3 I3 a3 73 IX I213 v,.,
's
.i. g 3, 9, 3 '3 3
3 1
3
..c3 8*
v3 v3 '3 v3 1-3 3
2 s
3 g
gg s
=f.'
g
=e
- ljE
'l E
a I
a I
r o
t t
I
- Ij g'!
j S
E 1
2 I
u 3
5 3
l 2 3 3
3 "I
3 1
1 E:5 d
1I ~:
o
=
0 I.
=
=
d k
I j
S I n, l
8 i
i
=
z I
A g
2 E,.
e lh I
.d e
s g
a x
~3:
to 0
tn w
e I
I
TABLE 6 - KEY TO BWR ACCIDENT SEQUENCE SYMBOLS Rupture of reactor coolant boundary with an equivalent diameter of A
greater than six inches.
Failure of electric power to ESFs.
B Failure of the reactor protection system.
C Failure of vapor suppression.
D Failure of emergency core cooling injection.
E Failure of emergency core cooling fun' tionability.
F c
Failure containment irolation to limit leakage to less than G
100 volume percent per day.
Failure of core spray recirculation system.
H Failure of low pressure recirculation system.
I Failure of high pressure service water system.
J Failure of safety / relief valves to open.
M Failure of safety / relief valves to reclose after opening.
P Q
Failure of normal feedwater system to provide core make-up water.
A small pipe break with an equivalent diameter of about 2 inches S
y to 6 inches.
Small pipe break with an equivalent diameter of about 1/2 inches S
2 to 2 inches.
Transient event.
T Failure of 'PCI or RCIC to provde core make-up water.
U Failure of low pressure ECCS to provide core make-up water.
V Failure to remove residual core heat.
W a
Containment failure due to steam explosion in vessel.
p Containment failure due to steam explosion in containment.
y Containment failure due to overpressure - release through reactor
'Juilding.
y Containment failure due to overpressure - release direct to atmosphere.
1 Containment isolation failure in drywell.
6 s
Cc7tainment isolation failure in wetwell.
(
Containment leakage greater than 2400 volume percent per day.
q Reactor building isolation failure.
Standby gas tra. ment system failure.
0 l
i 23
~
'~
from adjacent categories to account for uncertainties in the releas.e magnitudes i
(in other words, the uncertainty that a particular sequence would fall into only one of the assigned release categories). Second, the release fraction for each isotope group that was assigned to a category (see Tables 1 and 2),
generally was taken as the highest calculated fraction from all of those accident sequences (Tables 3 and 5) assigned into the release category.
Therefore, the release category was only a synthesized way of covering a large number of the accide.nt sequences assigned to the release rategory and could represent an overestimate on release for many of the assigned sequences.
Alternative work revealed however, that the smoothing and synthesis techniques were not particular significant to the overall societal risks projected by WASH-1400.
The smoothing and synthesis techniques did however 2rtifically inflate the importance of particular release categories (such as Category 1 involving vessel steam explosions) and did not enable a good recognition about tnose particular sequences and systems documenting the release categories and the overall projected risks.
These latter problems with use of the WASH-1400 smoothing and synthesis techniques are most significant primarily when the objective is to specify design risk reduction values associated with various improvement in LWR design. This latter objective was not however, the objec-tive of WASH-1400, as will be shown later in this report, the update PRA work subsequent to WASH-1400 has given a greater recognition to the risk importance of particular accident sequences and essentially has eliminated the smoothing and synthesis techniques of WASH-1400. This update PRA work should provide a reasonable foundation for those whose objective it is to improve LWR safety.
24 e
y.
Uncertainties and Improvements in Reactor Safety Study In 1977, the Nuclear Regulatory Commission sponsored the Risk Assessment Review Groups to independently review the accomplishaents and limitations of the Reactor Safety Study.
The general conclusion that was reached was that the uncertainties associated with the absolute values used in WASH-1400 were large and caution should be applied whenever using such abselute values and the PRA techniques.
Significant research efforts have been ongoing to improve and formalize the risk assessment techniques developed in WASH-1400.
All aspects of the work have undergone extensive u,...j to determine sensitivities of various a. sump-tions, and to revise approaches that were critized.
Improvements in data bases and in the experience in using the techniques have given greater levels of understanding to their application and their weaknesses. Of prime interest is the improvements which have occurred in the modeling of the physical processes of the severe core damage accidents including the releases from containment.
Significant progress has been made to improve the analytical methods of meltdown analysis and containment response relative to the RSS approach.
The " Meltdown Accident Response Characteristics,"s MARCH code, model has been developed to provide analyses of various thermal-hydraulic processes during reactor meltdown accidents and in particular to give a pr.ority focus on those small LOCA and transient restricted events identified in WASH-1400 are very important to the overall risk. Additionally, the " Containment of Radionuclides Released After LOCA,"10 CORRAL code, has been modified and generalized for use in studies, including a much larger population of LWR designs.
l 25
The MARCH (Meltdown Accident Response Characteristics) code has been under development at Battelle Columbus Laboratories with an objective of providing a fast-running analytical capability for investigating the thermal-hydraulic aspects of meltdown accident physical processes. The principal physical phenomene and accident processes which are analyzed in the cede include:
The time scale of the accident, particularly tte times for the start and completion of core melting; The time required for the molten core to fail the reector vessel Lottoa head; Possible energetic interactions when the core debris fall to the floor of the reactor cavity, including the likelihood of containment failure due to such interactions; Long-term pressure-time history within the reactor containment, including the likelihood and time of containment failure due to overpressure; The probability and consequences of hydrogen burning or detonation within the containment building; and The interaction of the core debris with the concrete fcundation.
MARCH was publicly released in October 1980.
Subsequently, the code has seen use in a number of regulatory issues and has been the subject of several substantive reviews.11'12 The general results of these reviews indicate that 26
4
=
e the code has a number of deficiencies and that great care should be exhibited in its application. This has been noted to be especially true for instances when the code is being used in situations other than that originally intended (that is, for probabilistic risk assessment).
Improvements to the code are now being initiated to attempt to correct the more important deficiencies; completion of this effort (with public release) is expected in the latter half of 1982.
Improvements to the CORRAL code following the release of the Reactor Safety.
Study have been performed in two phases.
In the first phase, the code was Laneralized to be applicable to a broader spectrum of containment designs (beyond the Surry and Peach Bottom designs). This chage, as well as certain others, were incorporated in the CORRAL-2 code, which was publicly released in 1977.
Since 1978, the second phase of CORRAL improvements has been underway with an objective of making additional substantive changes to both the code structure and its models (to reflect improved data in some areas). This work is now l
nearing completion.. A peer review has been performed on the code and its l
l documentation; the major comments of this review will be accounted for in the code prior to public release. This release is now planned for the latter half of 1982.
Because of the extensive changes made to the code, it has been l
decided to rename the code to MATADGR (" Method for Analysis of Transport and Deposition of Radionuclides).
In the discussion of PRA results which follow and are used in the synther,is of the source terms, versions of the above codes actually used were:
a pre public release version of MARCH (circa 1978); and CORRAL-2.
27
Rebaselining of Reactor Safety Study Results It was noted above that, following the release of the Reactor Safety Study (RSS), work was initiated to develop improved codes for the prediction of the consequences of core meltdown accidents (i.e., MARCH, CORRAL-2, and CRAC-2).'
In 7.978, when the CORRAL-2 code had besa completed and an initial version of the MARCH code developed, work was undertaken to reevaluate ("rebaseline") the results of a number of RSS accident sequences. Through this effort the effects of code improvements could be measured for both specific accident sequence phenomenological predictions and overall risk predictions.
A relatively wide spectrum of sequences was chosen for investigation, including risk-dominant sequences, sequences contributing significantly to the predicted core melt probability, and sequences having particular phenomena of interest.
In most cases, the reanalyses resulted in relatively minor changes to release pre-dictions.
In one case, however, the change in predicted release was sufficient to shift the released category designations of an accident sequence. The BWR accident sequence TQUV thus had its y' and y failure modes shifted from Release Categories 2 and 3 to categories 3 and 4, respectively.
In addition, this work permitted the updating of the RSS asults with respect O-to certain new data related to steam explosions. Based on this new informa-tion,13 the likelihood of c'ontainment failure caused by an in-vessel steam explosion was reduced from 1 X 10 2 to 1 X 10 4 for accident sequences in wnich the reactor coolant system pressure was relatively high at the time of the core collapse into the water in the lower reactor vessel head.
- Further, the assumption that some accident sequences had a steam explosion-caused containment failure mode (the a failure mode) in PWR Release Category 3 was 28
altered.
In the rebaselined results, all a containment failure modes were placed in Release Category 1.
These changes thus have the effect of altering the relative probabilities of PWR Release Categories 1 and 3 and BWR Release Category 1.
The results of the reanalysis of the predicted radioactive releases for individual accident sequences are shown in Tables 7 and 8 for PWRs and Tables 9 and 10 for BWRs. These results are also shown graphically for certain sequences in Figures 1 to 3 for PWRs and Figures 4 to 6 for BWRs. The effects of proba-bility modifications due to different assumptions related to steam explosions (and the shifts for the TQUV Sequence) are shown in Tables 11 and 12 and Figures 7 and 8.
It can be seen from these data that the predicted release magnitudes for individual radionuclide groups changes somewhat for many accident sequences.
However, the information shown in Figures 7 and 8 indicates that the overall RSS results were not hignly sensitive to the modeling changes.
In response to a request by the Commission to address the question of continued operation of the Indian Point Units 2 and 3, a task force was formed to com-pare the risks from Indian Point with other reactors.4 As part of this effort I
the release categories of the Reactor Safety Study (see Appendix B of NUREG-0715 Rebaselining of +.he RSS Results) were reevaluated using current techniques.
Tables 13 and 14 present the rebaselined accident release categories for the PWR (Surry) and BWR (Peach Bottom), respectively.
The symbols for the release sequences are identical to those given in Tables 4 and 6.
l l
29
Table 7 RSS baseline calculation - summary of MARCH results ECCS CSS Core Core Melt Vessel Cont.
Fail.
Concrete Start Stop Start Stop Uncovery Start End Failure Time Pres.
Helt Start Sequence (Min)
(Min)
(Min)
(Min)
(Min)
(Min)
(psia)
(min) 17 30 51 91 0
15 91 V
1827 1870 1929 2044 1787 100 2054 S C-6 0
1787 2
209 226 261 289 261 61 418 TMLB'-a 209 226-261 271 279 99 409-TMLB'-6 TMLB' c E09 226 261 271 1490 100 409 2
35 34 49 67 78 0
15 218 S D-p g;
35 67 34 49 67 109 67 18 218 S D-a 2
AHF a 1
65 1
50 102 125 145 204 145 47 365 AHF y 1
65 1
50 102 125 145 207 214 90 365 i
'AHF-6 1
65 1
50 102 125 145 207 215 77 365 AHF-c 1
65 1
50 102 125 145 207 1450 100 365 4
e W
O 4
P e
Table 8 RSS baseline calculation CORRAL results Cumulative Release Factors Sequence Xe-Kr I-Br Cs-Rb Te Ba-Sr Ru La V
- 1. 0
.642
.820
.410
.097
.044
.006 S C-6 1.0 0.52
.332
.188
.036
.018
.003 2
TMLB'-a 1.0
.484
.527
.361
.063
.402
.002 TMLB' y 1.0
.308
.388
.145
.044
.018
.002 TMLB'-c
- 0. 7 4(4)
.001
.001 1(4) 7(5) 1(5)
S 0-p 1.0
.002 009
.007
.001 6(4) 9(5) 2 S0a 1.0
.369
.439
.313
.057
.323
.002 2
AHF-a
- 1. 0
.369
.439
.313
.057
.323
.002 AHF y 1.0
.148
.236
.126
.029
.013
.002 AHF-6
. 1. 0
.140
.224
.129
.028
.013
.002 AHF-c 0.7 2(4) 7(4) 8(4) 8(5) 6(5) 1(5)
The notation 5(5) is an abbreviation of 5 X 10 5 31
TABLE 9 RSS BASELINE CALCULATIONS -
SUMMARY
OF MARCH RESULTS FOR BWR ECCS Core Core Melt Vessel Cont. Failure Concrete Start Stop Uncovery Start End Failure Time Pressure Melt Start Sequence (min)
Time (min)
(min)
(min)
(min)
(psia)
(min) 1 9
50 76 48 175 76 AE (dry) 1 19 51 77 49 175 77 AE (wet) 31 56 1 01 117 117 175 118 SE 2
e TC I
77 77 100 133 145 77 175 146 124 159 191 203 251 175 203 TQUV M
3185 3264 3390 3405 2820 175 3405 TW O
e 9
e
~
TABLE 10 RSS BASELINE CALCULATIONS - CORRAL RESULTS FOR BWR Cumulative Release Fractions Sequence Xe-Kr I-Br Cs-Rb Te Ba-Sr Ru La AE-a' N 1.0 0.3 0.6 0.4 0.07 0.3 0.004 AE-a 1.0 0.2 0.4 0.4 0.05 0.3 0.003 AE-Y 1.0 0.02 0.05 0.06 0.006 0.005 8(4)
AE-Y' 1.0 0.09 0.3 0.4 0.03 0.03 0.005 S E-Y 1.0 0.003 0.02 0.08 0.001 0.005 0.001-2 S E-Y' l.0 0.03 0.1 0.4 0.006 0.03 0.005 2
TC-Y 1.0 0.09 0.2 0.1 0.03 0.01 0.002 TC-Y' 1.0 0.5 0.6 0.4 0.07 0.04 0.006 TQUV-Y 1.0 0.001 0.02 0.07 9(4) 0.004 9(4)
TQUV-Y' 1.0 0.01 0.09 0.4 0.005 0.02 0.005 (1) With no pool scrubbing of melt release
-5 (2) The notation 5(5) is an abbreviation for 5 x 10 i
33
RSS TMLB's Rebaseline TMLB'a 0
I Cs Te Ba Ru La 10 m
w 10-1 eQ
$n w
W m
M W
W ec GI>
E
=
W l
3 10-2 I
u
=w y
J.0-3 FIGURE 1 34 y-m.
RSS TMLB'6
_ Rebaseline TMLB'6 I
cs To Ba au T.a g
wwwwWWWWI e
pyww w 10-1 CO wy M5.
A W
en R3 W
V
,w w.w..
W>
+J M
5 10-2 5
0 10-3 FIGURE 2 35
RSS S C6 2
Rebaseline 5 C6 2
I Cs Te Ba h
La 100 i
10-1 C
r-y 2
u.
43 E
a:
I.
's 1o-2 5
o 10~3 FIGURE 3 36
l l
0 10 g
10_1 U<
M' W
Co<
W J
W E
W>
"g 10-2 i=
U 10-3 FIGURE 4 37 i
l RSS AEY REBASELINE AEY I
Cs TE 3A Ru b_
10 l
I i
Z
$ 10-1 o4 12:
6 I
W l
f./)
I l
W JW Qll 1
ammmmme W>
- s 10-2 Eo 10-3 1
I FIGURE 5 38 j
l
--.-..--,-..--,n.-
RSS TW Y REBASELINE IW Y I
Cs TE BA Ru LA 10 2
O
~
10 "
u M
w LM to<
WJ uJ M
uJ>
4-8
_ - same
.m E
10 "
=
U 10-3 FIGURE 6 39
.e
-~
t Table 11 RSS baseline calculation dominant accident sequences Release category 1
2 3
4 5
6 7
S C-6 ACD-p 2
2 2
2 3x10 s 4x10.s 2x10 s 1x10 11 2x10.s 6x10 7 9x10 8 S H-S S CD-c S H-c S CD-p SHa TM LB '-6 S F-6 2
2 2
2 1
2 3x10 8 3x10 8 1x10 7 1x10 11 1x10 8 2x10.s 6x10 s AD-a S G-6 S 06-S 0-S B-c TML-c 2
2 1
2 2x10 s 9x10 s 1x10 12 6x10 9 8x10 8 6x10 8 S H-p 5 8-c S D-c S C-a S F-6 t
1 t
2 2
2x10 s 3x10.s 5x10 9 2x10 8 3x10 8 AH-a S G-6 AD-p S H-c 2
t 1x10.s 3x10 8 4x10 8 3x10 8 i
5 F-a AF-6 AH-p TKQ-c 2
1x10 8 1x10.s 3x10 9 3x10 s l
AD-c 2x10 s AH-c 1x10 8 TKMQ-c 1x10 s l
1.1x10 7 7.0x10 8 2.3x10 s 2.1x10 11 4.9x10 s 6.3x10 7 3.4x10 s 1
8.3x10 7 7.2x10.s 3.0x10 8 3.1x10 7 4.8x10 7 4.0x10 s 3.4x10 5 s
40 l
=
Table 12 RSS BWR key accident sequences Relaase catagory 1
2 3
4 Ae-a TW y' TW y TQUV y 2(9) 3(6) 1(5) 4(7)
S E y' TC y AEG-6 S E-a 1
1 2(9) 4(8) 1(5) 7(10 TW-a AE y' AE y S GHI-6 2
2(9) 3(8) 1(7) 6(10 S J y' SEy S GJ-6 S J-a 1
2 2
1 1(9) 2(8) 1(7) 6(10)
S I y' R
S EG-6 SIa 2
2 2
1(9) 2(8) 1(7) 3(10)
S HI y' S HI y TC-a 2
2 1(9) 2(8) 9(8)
SIy 2
9(8)
SJy 2
8(8)
TQUV y' 8(8) 9.9(9) 3.1(6) 2.1(S) 4.0(7) 4 5.3(7) 5.2(6) 2.1(5) 2.5(6) 41
. -.. -,. _ -.. _, -. _ _ _ _,, _ _. _ _ ~. _. _ _ _ _., - - _ _ _ _. -. _. _ _ _ _ _ - - _. _ _ _. _ _ _ _
T
///////////b 7
//////'//l 6
7/b 5
RW P
D E
N Y
I L
/ k' E
G S
E A
T 7
BE 4
A R
C E
R E
U S
G I
AE F
L E
R 7//// //k 3
/
V//I///,//////
2 f, 6 / 6 1
6 1
,!eeBi i
I l
O N
N
\\
4 YR 3
O G
E 8
TA E
C R
E UG S
I A
F ELE R
R W
B D
2 E
N I
L E
S A
B S
E S
R R
s 1
~
0 60 70 1
- 1
- 1 5
gn. $EQ82 0
l
.I!
I 2,.i iI l
- l l!!
'l
!l!I:i
1 l
j ll l
1 i
e.
TABLE 13 REBASELINEO RSS PWR ACCIDENT RELEASE CATEGORIES (SURRY) b Fraction of Core Inventory Released Accident Prob.
Time" Ouration Warning' Energy Xe-Kr I
Cs-Rb Te-Sb Sa-Sr Ru' La#
d 10 l
Sequence Per (hr)
(hr)
(hr)
Stu/hr Reactor s
Year
- 1. Event V 2 x 10 8 1.0 1.0 0.5 0.5 1.0 0.64 0.82 0.41 0.1 0.04
.006
- 2. IMLB' 3 x 10.e
- 2. 5 0h 1.0 170 1.0
.31
.39 15
.04
.02 2x10.a i
- 3. PWR-3a 3 x 10.e 5.0 1.5
- 2. 0 6
0.8 0.2 0.2 0.3 0.02 0.03 3x10.s Melt 8 4 x 10 6 10.0 10.0
- 1. 0 N/A 6x10 8 2x10 6 1x10 6 2x10 6 1x10 8 1x10.e 2x10 7
- 4. PWR-7
' Time interval between start of hypothetical accident (shutdown) and release of radioactt've material to the atmospfwre.
iotal time during which the major portion of the radioactive material is released to the atmosphere.
D
' Time interval between recognition of impending release (decision to initiate pubilc protective measures) and the release of radioactive material to the atmosphere.
0rganic lodine is combloed with elemental lodines in the calculations. Any error is negligible since the release fraction is d
relatively small for all large release categories.
' Includes du, Rh, Co, Mo, Tc, Ilncludes Y La, Zr, Nb, Co Pr. Nd, Np, Pu, Am. Ca.
' Updated calculations with MARCH and CORRAL Indicated tat these release cat %gorles were not importantly changed and could be retained for use in PRA. These particular categories have little
)
impact on the overall predicted risks.
C j
4 t -
TABLE 14 REBASILINED RSS BWR ACCIDENT RELEASE CATEGORIES (PEACH BOTTOM) b Fraction of Core Inventory Released Accident Prob.
Time" Duration Warning' Enerp I
Xe-Kr I
Cs-Ab Te-Sb Ba-Sr Ru' La Sequence Per (br)
(hr)
(hr) 10 d
Reactor 8tu/hr Year
- 1. AEa 5 x 10
- 0.8 0.5 0.5 120 1.0
.29
.57
.42
.071
.29
.004
- 2. AEa, 10
- 0.8 0.5 0.5 120
- 1. 0
.024
.04
.3
.0025
.28
.002 8
2 x 10 8 1.5
- 2. 0
- 1. 0 I
1.0
.45
.67
.64
.073
.52
.0083
- 3. TCy 8
3 x 10 8 50.0 2.0 40.0 I
1.0
.098
.27
.41
.025
.028
.005
- 4. lWy 8
3 x 10 7 2.0 0.5 1.0 15 1.0
.095
.30
.36
.034
.027
.005
- 5. IQUVy
- 6. TCy 8 x 10 8 1.5 2.0
- 1. 0 0
1.0
.07
.14
.12
.015
.01
.002
- 7. TQUVy 10.s 3.5 0.5 1.0 0
1.0
.02
.055
.11
.006
.007
.0013
- 8. TWy 1 x 10
- 50.0 2.0 40.0 0
1.0
.003
.011
.083
.011
.007
.001
- ilme interval between start of hypothetical accident (shutdown) and release of radioactive material to the atmosphere.
biotal time during which the major portion of the radioactive material is released to the atmosphere.
' Time interval between recognition of lapending retsase (decision to initiate pubile grotective measures) and the release of rasloective material to the eteosphere.
d0rganic lodine is combined with elemental lodines in the calculations. Any error is negligible since the release fraction is relatively small for all large release categories.
- Includes Ru, Rh, Co, Mo, Tc.
I Includes Y. La, Zr, Nb, Ce, Pr, Md. Np, Pu, As, Ca.
I I
l r
I a,._.
9
The accident sequences identified in the rebaselining effort which are expected to dominate risk of the RSS-BWR design are briefly described below.
TCy' and TCy These sequences involve a transient event requiring shutdown of the reactor while at full power, followed by a failure to make the reactor subcritical l
(i.e., terminate power generation by the core). The containment is assumed to be isolated by these events; then, one or the other of the following chain of l
events is assumed to happen:
(a) High pressure coolant injection system would succeed for some time in providing makeup water to the core in sufficient quantity to cope with the rate of coolant loss through relief and safety valvos to the suppression pool of the containment.
During this time, the core power level varies, but causes substantial energy to be directed into the suppression pool; this energy is in excess of what the containment and containment heat removal systems are designed to cope with.. Ultimately, in about 1-1/3 hours, the containment is
(
estimated to fail by overpressure and it is assumed that this rather l
severe structural failure of the containment would disable the high pressure coolant makeup system. Over a period of roughly 1-1/2 hour after breach of containment, it is assumed the core would melt.
This has been estimated to be one of the more dominant sequences in terms of accident risks to the public.
46
(b) A variant to the above sequence is one where the high pressure coolant injection system fails somewhat earlier and prior to con-o tainment overpressure failure.
In this case, the earlier melt could result in a reduced magnitude of release because some of the fission products discharged to the suppression pool, via the safety and j
i relief valves, could be more effectively retained if the pool remains subcooled.
The overall accident consequences would be somewhat reduced in this earlier melt sequence but ultimately, the process accompanying melt (e.g., noncondensibles, steam, and steam pressure pulses during reactor vessel melt-through) could 7.ause overpressure failure (y or y') of the containment.
TWy' and TWy 1
The TW sequence involves a transient where the reactor has been shut down and containment has been isolated from its normal heat sink (i.e., the power conversion system).
In this sequence, the failure to transfer decay heat from the core and containment to an ultimate sink could ultimately cause overpres-sure failure of containment.
Overpressure failure of containment would take many, many hours, allowing for repair or other emergency actions to be accom-plished; but, should this sequence occur, it is assumed that the rather severe structural failure of containment would disable the systems (e.g., HPI, RCIC) providing coolant makeup to the reactor core.
(In the RSS design, the service water system which conveys heat from the containment via RHR system to the ultimate sink was found to have a single point vulnerability and represent the dominant failure contribution in the TW sequence.) After breach of contain-ment, the core is assumed to melt with the release being discharged either 47
_,,---n a ---
into the reactor building (y) or directly to the atmosphere (y') without significant placement or retention by the surfaces and compartments in the reactor building.
[TQUVy', AEy', S Ey', S Ey'] and [TQUVy, AEy, S Ey, S Ey]
t 2
1 2
Each of the accident sequences shown grouped into the two bracketed categories above are estimated to have quite similar conscquence outcomes and these would be somewhat smaller than the TCy', y and TWy' sequences described above.
In essence, these sequences, which are characterized as in the RSS, involve failure to deliver makeup coolant to the core after a LOCA or a shutdown transient event requiring such coolant makeup.
The core is assumed to melt down and the melt processes ultimately cause overpressure failucs of contain-ment (either y' or y).
The overall risk from these sequences is dominated by the higher frequency initiating events (i.e., the small LOCA (S ) and shutdown 2
transients (T).
Probabilistic Risk Assessments (Post WASH-1400)
Since the completion of the Reactor Safety Study (RSS), additional probabilistic risk assessments have been performed under the direction of the Nuclear Regulatory Commission's Office of Research.
It was recognized that the two designs analyzed in the RSS did not necessarily represent the various reactor power plant designs in operation and construction.
Therefore, a research program involving the use of PRA was begun in order to explore the effect of variations in design on accident reactors. This program, titled " Reactor l
Safety Study Methodology Application Program," RSSMAP, was designed to analyze a group of plants using the insights from the RSS on the dominant accident l
48 l
sequences and by comparing systems against those system fault and event tri.es developed in the RSS.
Four reactors were analyzed:
Sequoyah, a Westinghouse S
4-loop PWR with a steel shel1~- ice condenser containment; system ; Oconee, a Babcock and Wilcox 2-loop PWR with concrete dry containment 10; Calvert Cliffs, a Combustion Engineering 2-loop PWR with concrete dry containment; and Grarid Gulf, a General Electric BWR with Mark III containment systems, the latter plant is in addition to RRSMAP, another research program, the Interim Reli-ability Evaluation Program (IREP), was begun in late 1979 in order to develop a somewhat standardized approach toward evaluating the risk of power plants as part of the regulatory safety review processes.
The first plant analyzed by the IREP was Crystal River -3, a Babcock and Wilcox 2-loop PWR design with a
" dry" concrete containment.
Four additional plants, Millstone #1 (BWR),
Arkansas Nuclear One,- Unit #1, (PWR), Browns Ferry and Calvert Cliffs (for benchmark purposes) are currently being analyzed in the IREP.
At this time only preliminary quantitative estimates on Crystal River -3 are available from the IREPts and as mentioned earlier these results are subject to future ctlanges and refinement.
(a) Rebaselined Surry vs. Oconee 3 and Crystal River-3 From the RSSMAP results on Oconee 3 and the preliminary IREP results on Crystal River 3, it is estimated that these two units should have release characteris-tics comparable to those of the rebaselined Surry Unit in WASH-1400. The principal differences relative to the rebaselined Surry unit would be associ-ated with changes in probability of the various accident sequences because of the differences in plant system design and system operation. Qualitatively, the types of accident sequences identified by the RSSMAP24 17 and IREP18 49
studies for the B&W plants were not dissimilar to those identified for Surry in WASH-1400; only the quantitative (probabilistic) estimates differed.
Like Surry, the small LOCA and transient relating events also dominated the spectrum of core melt releases.
b)
Rebaselined Surry vs. Sequoyah 1 The RSS-Map for Sequoyah 1 indicated tha'. the presence of an ice condenser containment system led to somewhat different risk dominant accident sequences.
Also the release characteristics for many of the core melt accidents were somewhat different than for the large dry containment - in part, due to the presence of buffered ice as a passive retention mechanism.
For example, the scenario TMLB' was found to be an important risk contributor for the Surry design.
This sequence was of considerably lesser importance in the Sequoyah 1 design because of the additional retention of radioactive materials by the passive ice in the ice condenser. The contribution of the interfacing LOCA scenario, Event V was assumed to be some less for the Sequoyah plant, than in the Surry design due largely to yearly testing planned for the check valve 1
barriers between the low pressure injection /RHR system and the primary coolant system and (2) the retention by ice of a fraction of the Event V release after the molten core has breathed the reactor vessel.
l A rather unique potential fault was identified for the Sequoyah plant that was associated with the compartmented nature of the containment design and the 1
1 Note the RSSMAP results 'ar the Sequoyah Event V14 did not reflect this testing program.
50
need for this drain water between the upper and lower compartments, fault related to a having blocked coolant drains between the upper compartment to the lower compartment following a LOCA when the ECC and containment spray and heat removal systems were aligned into a recirculation mode of operation.
The dominant fault leading to drain blockage was assessed to be human errors in failing to remain flanges that would be used to block the drains during refueling. This resulted in a common failure mode for all of the post-havoc coolant recirculating systems.
Relative to the Surry design this sequence (designed as AHR, S, HF, SzHF)1 for Sequoyah were of higher probability and they were found to importantly dominate the overall accident risk in Sequoyah.
For those HF sequences the ice suctioning the ice compartment would be depleted prior to failure of the recirculation systems (HF) thereby resulting in no fission product mitigating benefits by ice to about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br /> the containment was estimated to fail in the overpressure mode leading to some of the largest release magnitudes identified for this plant.
The MARCH and CORRAL results set forth in reference 14 were revicwed and further characterized into various release groupings for use as input to the CRAC consequence modeling.
These inputs are given in Table 15.
More complete I
explanations on the releases from individual sequences and the associated l
l probabilities can be found in ref. 14. Table 16 shows those particular Sequoyah sequences that determined the releases magnitudes and release probabilities in l
Table 15. The accident sequence designations used in the Sequoyah PRA study are much the same as those for Surry.
Table 17 provides a brief description of these Sequoyah sequences.
See Table 14.
51
J TABLE 15 SEQUOYAH ICE CONDENSER RELEASE CATEGORIES b
Fraction of Core Inventory Released Accident Prob.
Tlee*
Duration Warning' Enerp Me-Kr I
Cs-Rb Te-5b Ba-Sr Ru*
La' d
Sequence Per (hr)
(hr)
(hr) 10 Reactor Stu/hr Year IC 1 1 x 10 7 1.0 0.5 0.5 6
1.0
.27
.37
.42
.045
.43
.603 IC 2 9 x 10 7 0.6 1.3 0.5 5
1.4
.48
.42
.09 052
.016
.002 IC 3 5 x 10 8 3.0 0.5 1.5 40 1.0
.13
.57
.49
.068
.042
.007 IC 4 6 x 10.s 4.0 1.0 2.0 40 1.0
.0%
.26
.34
.033
.024
.004 E
IC 5 10 5 2.G
- 1. 0 1.0 40 1.0
.007
.04
.150
.003
.009
.002 IC 6 2 x 10 5 1.5 1.0 0.5 40 1.0 5x10 4 6x10 4 3x10.a 4x10.s 2x10
- 3x10 5 IC 7 4 x 10.s 1.5 1.0 0.5 40 1.0 7x10 7 3x10 7 3x10 7 3x10.e 2x10 8 4x10 *
" Time interval between start of hypothetical accident (shutdown) and release of radioactive material to the atmosphere.
]
Diotal time during which the major portion of the radioactive material is released to the atmosphere.
Time interval between recognition of impending release (decision to initiate pubile protective measures) and the C
release of radioactive material to the atmosphere.
darganic lodine is combined with elemental lodines in the calculations. Any error is negilgible since the release fraction is relatively small for all large release categories.
- Includes Ru, Rh, Co, Ma, Tc.
IIncludes Y. La, Zr, Nb, Ce, Fr. Nd, Np, Pu, As, Ca.
4 G
Because it has an ice condenser containment, the Sequoyah unit had different releases. These categories for Sequoyah are given in Table 15, and the accident sequences that have been grouped into the categories are given in Table 16.
Complete explanations of the release categories and accident sequences can be found in the specific plant reliability reports.14 17 Information on the other plants not reported here are still being developed and should be available by the middle of 1982.
In addition to the studies sponsored by the NRC, other power plants have been analyzed using probabilistic risk assessment techniques. The German Reactor Safety Study analyzed the Biblis-B PWR.
Diablo Canyon was analyzed to assess the seismic vulnerabilities.
Recently, Indian Point, Zion, and Limerick were ordered by the NRC to perform risk assessments because of their high population densities.
Indian Point Risk Comparison Study As part of the effort to assess the risk at Indian Point Units 2 and 3,(4) the staff performed a limited systems risk assessment of the units.
The characteristics of the accident sequences, including system unavailability and containment failure mode, were reviewed to determine which of the fission produ:t release source terms developed as part of the RSS or RSS rebaselining effort were appropriate.
Several fission product release categories were then established as being characteristic of accident' sequences involving the various containment failure modes.
Table 18 provides 2: listing of Accident sequences 53
TABLE 16 SEQUOYAH ICE CONDENSER DOMINANT ACCIDENT SEQUENCES VERSUS RELEASE CATEGORY Ice Condenser Principal Sequences Probability of Release Category in Release Category Release Categorv IC1 S Ha 1 x 10 7*
1 A Da S HFa 2
IC2 Event V 9 x 107 S HFy 5 x 10 s IC3 2
IC4 TMLy 6 x 10 s S HFy6 1
ICS S H6 1 x 10 s 2
TMLB'y 2
S H60 2
IC6 S 06 2x10 5 2
S H6 1
IC7 S 06 4 x 10 s 1
AD6 IProbability values represent the summation of individual sequences in a release category where more than one sequence has been shown.
2The containment failure mode designations used (i.e., a, p, y, 6) for Sequoyah are identical to those used for Surry in WASH 1400 except for 8, which desig-nates a possible failure mode caused by core debris fragmentation upon contact with water in the reactor cavity.
54 eC1
TABLE 17 ICE CONDENSER PWR-KEY ACCIDENT SEQUENCES Sequences Description AD,5 0, S D A LOCA initiated by a large (>6 in.), medium (2-6 in.),
1 2
or small (<2 in.) break, respectively, in the primary coolant system accompanied by failure of the emergency core cooling injection.
AH, 5 H, S H A LOCA initiated by a large (>6 in.); medium (2-6 in.),
1 2
or small (<2 in.) break, respectively, in the primary coolant system accompanied by the failure of emergency core cooling recirculation.
AHF, S HF, S HF A LOCA initiated by a large (>6 in.), medium (2-6 in.),
1 2
or small (<2 in.) break, respectively, in the primary coolant system accompanied by the simultaneous failures of emergency core cooling recirculation and containment spray recirculation (includes the principal containment heat removal path).
TMLB' Failure of the power conversion and auxiliary feedwater systems given the initiating transient event of loss of offsite ac power with a failure to recover either onsite or offsite ac power within about 3 hours3.472222e-5 days <br />8.333333e-4 hours <br />4.960317e-6 weeks <br />1.1415e-6 months <br />.
TML Fail'ure ( the power conversion and auxiliary feedwater systems given an initiating transient even that interrupts main feedwater delivery.
V Failure of LPIS check valves leading to a LOCA, together with failure of the LPIS and resulting in core melt and a discharge of radioactivity essentially bypassing containment.
I l
l l
1 55
Table 18 Accider.t sequences and containment failure modes 01 02 03 G4 AHGa ADFGa AGa S HGa S DFGa S Ga AHa t
t t
S Ha 1
ADa S Da t
$3x109 N3x109 N5x109 N6x10s 6
6 6
6 1
2 3
4 TM BLM'B"6 AHGS AGS ADFG6 2
S HG8 S GS S DFG6 t
t 1
S HG6 S GO S DFG6 2
2 2
TM LM'B"6 AZ6 t
S Z6 1
S Z6 2
TM LXB'6 1
TM LM'XB'6 t
$2x10s g4.5x107 N1.5x10s gjg.s
&1
&2 Imod AHG&1 AH&2 AHI TM LX&
t
- 083[d od SH$ mod S HG&t S H&2
~
1 od t
t 1
l D W'B' mod S H& mod 5 H&2 5 HG&t 2
1 2
2 TM LM'X&
ADFG&1 AD&2 AD&(mod
'XBi t
d od SD S DFG&1 S 0&2 1
1 t
t TMBm'&m(mod SD&(mod S 0&2 S 0FG&t 2
d 2
2 od 2
TM BLM'B S LX AZ&t S LX&2 2
mod 2
2 TM LXB'&t VE2 V&
TMKX'&
ADNf(mod ADF&(md t
TM LX&2 TM LM'XB'&1 1
1 od d
TM LM'X&2 I DFG S DF TM BLt.'B'&t t
t od 2
1 S 0Fqod TM B W'&2 f.:DFG&
l 2
2 AHG&(mod TM KX' 1
(
ADF&2 S HG 1
d S HG& mod S DF&2 2
t AZ&(mod S DF&2 2
od SZ l
t S Z& mod 2
$5x107
$4x108
$3.2x10s V(bod)
V (&)
V (6)
V (6) 3 4
i 2
V(DCG)
V(DG)
V(D)2 V(D) mod
$1x109 N5x107 N5x107
$5x107 Folded into Other & Release Categories for CRAC Runs 56
1 l
grouped by cmitainment failure mode and release category with the corresponding i
RSS or rebaselined accident sequences identified in Table 19. The fission 4
product release characteristics of each containment failure mode category are provided in Table 20. Table 21 briefly describes the major accident sequences.
Consequences of Major Accident Release Sequences The preceding accident saquence release data provide the basis from which a set of representative accident source terms can be developed. The first step necessary to develop a rational set of accident source terms is to present measures of the importance of the various accident sequences so that compa.isons can be made. Accident consequence analysis provides the method to generate those measures that are needed to determine the importance of the source term.
The CRAC-21s (Calculation of Reactor Accident Consequence) code was used to analyze the importance of the accident sequen:es. This code was originally developed for the Reactor Safety Study and is discussed in detail in Appendix VI 3
of the study and in NUREG-0340, " Overview of Reactor Safety Study Consequence Model."20 To provide a meaningful comparison of the source terms, axpected individual risk versus distance of early fatality and latent cancer fatality were used as 1
the risk measures.
Expected individual risk versus distance is defined as the expected probability (mean) of being exposed to doses which will cause early fatalities (less than one year after exposure) or latent cancer fatalities at a specific distance.
The expected probabilities are derived by calculating the dose and associated area covered within various distances for a variety of l
weather conditions; thus generating a distribution of dose / area relationships.
57
i Table 19 RSS/rebaselined accident sequences corresponding to Indian Point fission product release categories Indian Point RSS/rebaseline equivalent product release category fission product release sequences at AHFa 82 ACDGIa 5 Ca ws 2
a4 AH, Ak 6
TMLB'6 2
62 AHFy
)
6 AHFy
) Rebaselined 3
6s SC
)
2 6
ACDGIS 4
(t TMLG'& - Rebaselined
&2 AD&
l
& mod AD& - terminated at end of vapor release i
V AB(
t V (&)
ACDGI&
2 v + &2 s
Y4*k mod 58
~
_=
Table 20 Fission product release characteristics a
b c
Containment Time Duration Warning Energy Fraction of core inventory released failure (hr)
(hr)
(hr) 108 d
e f
category Btu /hr Xe-Kr I
Cs-Rb Te-Sb Ba-Sr Ru La
- 2. 0 0.5
- 1. 0 100
- 1. 0 0.7 0.5 0.1 0.06 0.02 2x10 3 ha 5.0 0.5 3.0 100 1.0 6x10 8 7x10 5 6x10 5 9x10 8 5x10.s 8x10 7 2
Va 4.0 16.0 3.0 0
0.3 2x10 3 1x10 5 2x10 5 1x10 8 1x10 s 2x10 ?
V 2.0 16.0
- 1. 0 0
5x10 4 7x10 8 1x10 5 5x10 8 9x10 7 5x10 7 2x10 7 4
a3 2.5 2.5 1.0 90
- 1. 0
.37
.44
.31
.057
.323 2x10 3 a2 1.0 0.5 0.5 90 1.0 0.2 0.1 0.2 0.01 0.3 8x10 4 23.0 0.5 20.0 6
1.0
.3
.2
.1
.03
.1 1x10 3 as Ir4 1.0 0.5 0.5 6
1.0 3x10 2 6x10 8 2x10 2 8x10 4 4x10 2 4x10 5 6
2.5 0.5 1.0 70
- 1. 0
.31
.39
.15
.044
.018 2x10 3 1
6 3.5 0.5 2.0 100 1.0
.15
.24
.13
.03
.013 2x10 a 2
6 23.0 2.0 20.0 0
1.0
.052
.33
.2
.04
.02 3x10 3 3
6 20.0 0.5 18.0 100 1.0 6x10 3 7x10 5 6x10 5 9x10 s 5x10 8 8x10 8 4
(3 2.0 15.0 1.0 0
.7 4x10 4 10 3 10 3 10 4 7x10 5 1x10 5 (2
2.0 15.0 1.0 0
.3 5x10 4 8x10 4 2x10 5 1x10 8 1x10 8 2x10 7 (mod 2.0 4.0 1.0 0
5x10 4 5x10 8 1x10 5 1x10 5 1x10 s 1x10 8 2x10 7 aTime interval between start of hypothetica; accident (shutdown) and release of radioactive material to the e
atmosphere bTotal time during which the major portion of the radioactive material is released to the atmosphere.
Time interval between recognition of impending release (decision to initiate public protective measures) c and the release of radioactive material to the atmosphere.
d 0rganic iodine is combined with elemental iodines in the calculations. Any error is negligible since its release fraction is relatively small for all large release categories.
" Includes Ru, Rh, Co, Mo, Tc.
IIncludes Y, La, Zr, Nb, Ce, Pr, Nd, Np, Pu, Am, Ca.
Table 21 Description of major accident sequences Accident sequence Description V
Failure of check valves which isolated LPIS from RCS and causing loss of LPIS resulting in core melt.
AD, S D A LOCA in the RCS accompanied by a failure of 2
ECCS in the injection mode.
AH, S H A LOCA in the RCS accompanied by failure of ECCS 2
in the recirculation mode.
ADF, AhF A LOCA in the RCS accompanied by ECCS failure 5 DF, SpHF and containment spray recirculation failure.
2 AG, S G A LOCA in the RCS with failure of containment 2
cooli1g causing loss of conta' unent integrity by overpressure failure.
Subsequent interaction causes ECCS loss and core melt.
AZ, S Z A LOCA in the RCS accompanied by component 2
cooling water system failure which causes loss of ECCS and containment ESFs.
TM LX Interruption of main feedwater accompanied by 1
failure of AFWS and " feed and bleed" operation.
TM KX Interruption of main feedwater and failure of 1
reactor trip without successful initiation of
" feed and bleed" mode of operation.
TM LB' a TM BLM'B' Loss of offsite power causing interruption of 2
2 main feedwater with failure of AFWS and failure of onsite AC power sources.
60
The dose is calculated assuming the entire cloud exposure (inhalation and cloud gamma shine pathways), plus a one day ground exposure.
Shielding is associated with normal activities.
Given a calculated dose, the probability of being fatally injured or contracting a fatal cancer is computed.
For l
example, if the computed dose is 510 rem to the bone marrow, there is a 50 percent chance of being killed from such an exposure; for those individuals who survived, there would be on the order of an additional 10 percent chance of contracting a fatal cancer from the 510 rem exposure.
Individual risk is simply those health effect probabilities multiplied by the chance that an individual would be exposed given a release. This exposure probability is the ratio of contaminated area within a given distance to the total area within the distance.
(This assumes a uniform wind direction proba-bility.) CalculationsaremadeforarepresentativeAdkofmeteorological conditions. The expected individual risk is calculated by averaging the individual risk probabilities for all of the meteorological sequences at each distance.
Expected individual risk versus distance provides an easily viewed perspective of tne relative importance of the accident sequences. The early and cancer fatality curves represent two distinct types of radiation effects.
Early i
fatalities provide perspectives on source terms in which large fractions of the core inventory is potentially released into the environment.
Radiation doses exceeding 300 rem to the bone marrow are required to cause such an effect. Therefore the early fatality measure is a threshold phenomena and cannot be caused by relatively small releases. The source term issues which have been raised,1 could potentially reduce the chance of accident sequences 61
' rom generating the doses necessary for early fatalities.
Thus the uncertainties associated with early fatalities could have a very large impact en the results of this analysis.
When latent cancer fatality risk is considered as the risk measure, the effects are much more a simple measure of dose received with no defined thresholds.
Thus the curves do not terminate at some distance, but continue slowly dropping off.
Risk Curves To gain a perspective on the importance of the accident source terms, Figures 9 through 20 are presented.
These figures assume a unit probability for the release. Thus, they are conditional curves which display the relationship of the source term to the individual risk measures without regard for their probability of occurrence.
They are simply the risk faced by an individual at a distance assuming the accident has occurred. The first six figures present expected early fatality risk versus distance for the six previously discussed designs. The last six figures present the expected latent cancer fatality risk versus distance perspectives.
A second set of curves, Figures 21 through 35, are also given. These figures present the identical information as the first set of curves, except that the probabilities of the accident sequences, as measured in the risk assessments have been included. Thus, these curves represent our current evaluation of l
the absolute individual risk posed by the accident sequences.
Figures 21 through 28 give the expected individual risk of early fatalities versus 62
PROBABILITY OF ACUTE FATALITY T'O AN INDIVIOUAL VERSUS DISTANCE GIVEN* YARIOUS CORE MELT RELEASES 10-2 _
=
CAT.fi TWe, TCa RSS - BWR DESIGN g
(PEACHBOTTOM)
NN 4
10-3 g
\\
CAT.f3 TCy, TWy N
~~
10 CAT.f2 TWy' TQUYy' g
~
3 c_
10-5 10-6 4
0 0.5 1.0 1.5 2.0 2.5 10 DISTANCE (MILES)
ASSUMPTIONS:O
- RELEASE PROBABILITY (f.e. SEQUENCE PROBABILITY) ASS TO BE UNITY PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARCH-CO CODES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL o ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 9 63
PROBABILITY OF ACUTE FATALITY TO AN INDIVIDUAL VERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES 10-2 '
3 5
-TCy' TQUVy J EY RSS - BWR DESIGN Z
\\[<
A ls ET REBASELINE
\\
k
- $ Ey[r 2
~
^r 5 '-
2 i
~ E
\\
B 5
/
w m
g.
g:
3 AE2, 5)Em (DF=100)-
TQUYy' 3
iAEv' Z
~
1 S Ey' j
5 Ey' g
2 g
10-5 i \\
1 i
{
10-6
~
~
10'7 0 0.5 1.0 1.5 2.0 2.5 DISTANCE (MILES)
Tw -
r s
ASSUMPTIONS: 0
- RELEASE PROBABILUY (i.e. ' SEQUENCE PROBABILITY) ASSU TO BE UNITY PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARCH-CORRAL CODES O EXPECTED CONSEQUENCES FRCli 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL 0 ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 10 64
i PROBASILITY OF ACUTE FATALITY TO AN INDIVIDUAL YERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES I
10' 3
- CAT.f1 1MLB's, 5,Ca RSS - PWR DESIGN 3
}
(SURRY)
CAT.f2 TMLa'a,y, EVENT Y
]
p
=
10-3 s
i CAT.f3 S C4 l,
2 1
\\
10~4 E
CAT.f4(SMOOTHING)
Z 3
lis
\\
\\
g
\\
\\
10-5
=
10-6
=
=_
l
(
L l
0 0.5 1.0 1.5 2.0 2.5 10 DISTANCE (MILES)
~
ASSUMPTIONS: d
- RELEASE PROBABILITY (i.e. SEQUENC TO BE UNITY PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BC CODES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUE 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSUR SHIELDING BASED ON NORMAL ACTIVITY FIGURE 11 65
--,-m.
,--.--,.,,,n
-,..n
PROBABILITY OF ACUTE FATALITY TO AN INDIVIDUAL VERSUS DISTANCE GIVEN* VARICUS CORE MELT RELEASES 10~I I
~
(SURRY) 10-2 3
3 UNY 3 C4 2
3
[
d a
TMLS'4,7 8
E
=
10
=_
10-5
=
~
~
~
4 2.0 2.5 10 0
0.5 1.0 1.5 DISTANCE (MILES)
ASSUMPTIONS: O* RELEASE PROBABILITY (i.e. SEQUENCE PROBABILITY) AS TO BE UNITY PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARCH-CO CODES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 12 66
PROBASILITY OF ACUTE FATALITY TO AN INDIVIDUAL VERSUS DISTANCE GIVEN* VARICUS CORE MELT RELEASES 10-2 _
g
~
ICE CONDENSER Z
DESIGN 10-3
[5HFy s
2 s
fx 5,
x 5,w s
s
~
k 4
s E
/
?
QF,Y g
2 3
mg WMV 10-5 1
I
~
=
10-6
~_
1.5 2.0 2.5 10-7 0 0.5 1.0 DISTANCE (MILES)
ASSUMPTIONS: O
- RELEASE PROBABILITY (i.e. SEQUENCE P TO BE UNIH PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL M CODES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENC 3200 MWT POWER LEVEL o ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVI U FIGURE 13 67
~
~
PROBABILITY OF ACUTE FATALIT TO AN INDIVIDUAL VERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES 10'2 g
l
=
\\
[3 INDIAN POINT DESIGN 10'
==
=
~
V mod j
a2
- 2 n
10-4
=_
C g
3
~
3
\\-
3 4
.g a
., - a3 l
10-5
\\
i 10-6 i
l
_~
A 4
0 0.5 1.0 1.5 2.0 2.5 10 DISTANCE (MILES)
ASSUMPTIONS: 0
- RELEASE PROBABILITY (i.e. SEQUENCE PROSABILITY) A TO BE UNITY.PER REACTOR YEAR o RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARCH-CORRA y
i l
N CODES o EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL OE IRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE k
dN SHIELDING BASED ON NORMAL ACTIVITY h\\
i FIGURE 14 68
PR08 ABILITY OF LATENT CANCER FATALITY TO AN INDIVIDUAL VERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES 10" g
~
R$$ - BWR DESIGN
'it*
(PEACH 80TTOM)
I' b
.I\\
10-4 4Ca-3
- p.,N g CAT.f1 TWs,
.N
! 'N
\\
i i.
c CAT.f2 M '. Tpy?
[
3 li.
7 s
g c
d s
E s
.[
3 CAT.f3 TCy, twt i
g g,
s
[-
~~
10-0 CAT.!4 (SMOOTHING)
_~
~
10"I
.i
=
l
=
~
CAT.f5 NO MELT
\\
c l
0 5
10 15 20 25 10-8 t
DISTANCE (MILES)
ASSUMPTIONS: O
- RELEASE PROBABILITY (i.e. SEQUENCE PR08 ABIL TO BE UNIT ( PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARCH-CODES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES W 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 15
~
69 1
'ii
?
..s PROBABILITY OF LATENT CANCER FATALITY TO AN INDIVIDUAL l
VERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES j
.- --i in-2 _
I
~
~
REBASILINE c-r.,
10~3
-.=
V
'~~
l~. \\\\\\
s
~
j
- j. fTcy',,
\\
s 10.A -
=
lT AEa,' $ Em:
v.
TQUVy' '
~
3 W
AEv'.
i.,
s 3
1, i
a g
I
\\
I ~5
\\
~
f j
h \\\\
\\
N
~
K N
\\
10-6 N
5
=
l AEs, S Em (DF=100)
TQUVy j
y W
EY 2
10~70 S,
10 15 20 25 DISTANCE (MILES)
ASSUMPTIONS:o
- RELEASE PROBABILITY (i.e. SEQUENCE PROBABILITY) A TO BE UNITY PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARCH-CODES o EXPECTED CONSEQUENCES FRCH 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL 0 ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING JASED ON NORMAL ACTIVITY FIGURE 16 70 e
w-
~
PROSABILITY OF LATENT CANCER FATALITY TO AN INDIVIDUAL VERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES.
10-2 E
-- [
T y.
~
i.
10-3 h'.,
CAT.fi (COLD)-
'. ',[-.'
g.
g\\
N
. x 4
10
- CAT.f 3 4
g
.il CAT.f2 s
g E
6 i
e x
10-5 9ys -
e
,g i
LCAT.fi(HOT)
CAT.f4 s
g
~
^
'- CAT.f 6 N 10 3 '\\
3
_\\
- CAT.f 8
{
\\
(NONELT)
CAT.f5 s
_(\\
CAT.f7 10~70 5
10 15 20 25 DISTANCE (MILES)
ASSUMPTIONS: O
- RELEASE PROBABILITY (i.e. SEQUENCE PROBABIL TO BE UNITY PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARC CODES O EXPECTED CONSEQUENCES FRQ4 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND. EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 17 71
,-.n.,
~
PROBABILITY OF LATENT CANCER FATALITY TO AN INDIVIDUAL
.YERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES 10'2 g
=
~
~
(SURRY) p s
i
.i s
10-3 c-.
i _.
ri s
N i
s
i l-
\\
EVENT Y
\\
i i
\\
l t
N 10 4
=
i 6.,i
- 5 C4
~.
- 2 Z
C g
g c2
. ' ' i
-TMLB' 4.y E
g 10_
g N
10'O
=
(
1 10'7 0 5
10 15 20 25 DISTANCE (MILES)
ASSUMPTIONS: 0
- RELEASE PROBABILITY (i.e. SEQUENCE PROBABIL TO BE UNIU PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MARC CODES O EXPECTED CONSEQUENCES FROM*91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 18 72
~
PROBASILITY OF L4 TENT CANCER. FATALITY TO AN INDIVIDUAL VERSUS DISTANCE GIVEN* VARIOUS CORE MELT RELEASES 10~2 ~-
l
==
=
INDIAN POINT
~
~ ~ -
- DESIGN i
i i
' ~ ~ ~ ~
10~3 i
i g
Z 1
\\
\\
~
,t '
\\
.I l
i
.i
. l !..
\\
h l
10'4
~
j R
- i. ~ '
8
~!
3 g
3 V md
~
f.
i
~
.. - i
~
y-
/y- =2 88 s
_',a i
~
10,,
7_.4
' M 4\\
10-6
-N'x
=
g N
(
m Vmd
- 1
\\
\\-
x t,1 2
\\
.. e mod j
Y md mod 3
[
\\4
- 2 e
10 0
S
.10 IS 20 25 DISTANCE (MILES)
ASSUMPTIONS: O
- RELEASE PROBABILITY (i.e. SEQUENCE PROBAB TO BE UNITY PER REACTOR YEAR O RELEASE MAGNITUDES ESTIMATED FROM USE OF BCL MAR CODES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES 3200 MWT POWER LEVEL 0 ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 19 73
,,.._7 y
y
RIst OF EARLY FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10-3 TCy, TWy RSS - BWR DESIGN g
_=
CAT.f3 2
N (PEACH.BOTTCN)
[
s N
\\
10-4
(
s
~
p CAT.f2 TWy'. TQUYy'
~
CAT.f1 TWa, TCs i
s 10-5
~
~
g s
g ce ca.
10-6
=
=
l
~
10-7 :
I 10-8 0 0.5 1.0 1.5 2.0 2.5 DISTANCE (MILES)
ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSUMED TO BE 10 / RE O S'M0pTHI G%$tijalTED,^EiPtfCd4C10,EN(SF. SUE _NGES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORFJU. ACTIVITY FIGURE 20 74
RISK 0F EARLY FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10~3-TCy
=
R$5-BWR DESIGN
\\
~
REBASELINE N
\\
N.
\\\\
'\\\\
a 10 5
N TCy!
N
- TQUVy s
- AE i
~
T TWy'
'N
- hE 2
N 10-5 I
Ca 2
10-6 g
=_
AEs
~
\\
3 Em
~
1
\\
N
~
10-7
/
=
-=
\\
g EFFECT OF POOL DF=100 N
0
- 0. 5, 1.0 1.5 2.0 2.5 10-8 DISTANCE (MILES) 4 / REACTOR YEAR ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSUMED O f60NfHiAh! IN/ UFCWGE&MMES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUEN 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOS SHIELDING BASED ON NORMAL ACTIVITY FIGURE 21 75
RISK OF EARLY FATALITY TO AN INDIVIDUAL YERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10'3 ::
I
=
~
(SURRY)
~
[
CAT.#2 TML8's.v. EVENT V
- -4 E
5
\\
CAT.f3 5 Ca 2
,\\\\
10-5 I
t
~
~
d
!ll 8
- CAT.fi (COLD) TML8's, $ Ca J
2 10, j
C i
CAT.fi (H0T) TMLB's, $ Ca 2
l
~_
1 x
10 7 w
N s
N CAT.f4 (SM00THINGl-4 0
0.5 1.0 1.5 2.0 2.5 0
DISTANCE (MILES)
ASSUMPTIONS: O OVE ELL CORE MELT PROBABILITY ASS O pdDbtI' ngl,ELIh!STEL Eb(2L-ItNEMT. SEQUENCES I
O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE Q.000 EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY l
FIGURE 22 76 l
RISK 0F EARLY FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUSCES 10~3 =
l
=
(SURRY)
(
f2" 4
=
10
=
2
~
EVENT V
\\
10 5
=
=
3 TML8's,y 2
5
~
10-6 10-7
~_
~
4 10 0.5 1.0 1.5 2.0 2.5 0
DISTANCE (MILES)
ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSIMED TO BE 10 O SHOGTRIM M LI @ DRLIGLTA CCIDEM M EQUENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 23 77
~
l l
RISX OF EARLY FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES
=
l 10~3i-
~
ICE CONDENSER DESIGN
[
\\
s N g /-8 "T 2
N N
4 10 s
-5 HF 7.
2 N
g 3
x B LT i
g N
N 5 F7 l
- 1 N
i 10-5 N
s g
\\
N C
\\
N N
3 s
y E
s EVENT Y
\\
S g
10'0._
--=
N N
rs ad E
t
\\
10-7
\\
E E
\\
\\
10-8 DISTANCE (MILES)
ASSUMPTIONS: O 0VERALL CORE MELT PROBABILITY AS
/ REACTOR YEAR I
O SN0'OTRININ1-!MINATEDfEXPt'ICIT-ACCI9ENLSEQUENGES-O EXPECTED CONSEQUENCES FRm 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O FNTIRE CLOUD EXPOSURE + 24 HOUR GROU.1). EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 24 78
~
,=.
,O RISX OF EARLY FATALITY TO AN INDIVIOUAL VERSUS DISTANCE FOR VARICUS CORE MELT SEQUENCES 10-3
=
INDIAN POINT DESIGN 3
~
4
\\
BEFORE FIX -
~
\\
- AFT R FIX
\\
\\s\\
10
\\
=
\\\\
_~
LAHG4,y
\\
'S)HG4.Y 5 " 8'T g
g 2
\\ 4 10-5 2
3 TMLS's.y g
g s
s N
10-6 y
3
\\ h
\\
s
~
\\ N \\
\\
g
~
' \\' N-
\\\\ \\
~
\\
10 7 s
g sy N
\\\\
\\
s
~
)
S.
1 10-8 0
0.5 1.0-1.5 2.0 2.5
~
DISTANCE (MILES) 4 ASSUMPTIONS: O OVERALL CORE MELT PROBASILITY ASSUMED TO BE 10 / REA O pdOTRIM DIMIMUSILACSEENT--SEQUENCEN.
O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 25 79
l RISK OF EARLY FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10-3
)
g
=_
~
OCONEE S C4 2
EVENT Y
/
4
\\
/
5 s
TMLB's,y 10-5
=
E E
E e
g 8
Z 10-6 10-7
-=
=_
l 10-8 0 0.5 1.0 1.5 2.0 2.5 DISTANCE (MILES)
ASSL2'PTIONS:O OVERAL.L CORE MELT PROBABILITY ASSUM O SM00 THIN'G EkIMINATEDdQPLJCILACCIDENESEQUENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL o ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED OM NORMAL ACTIVITY FIGURE 26 80
. eem
+w r
RISK OF EARLY FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10-3 CRYSTAL RIVER EVENT Y
[
'^A' fY Y 10 N
/
/
=
A
- M.,
[
3 M str
~
,iiMf TMtB a,y Y* YA' 10-5 hb
'W S C4 b
2 g
E j
10-6
=
=
=
1 10-7
=
~
i t
zo-a0 0.5 1.0 1{5 2.0 2.5 DISTANCE (MILES) 4 ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSUMED O SMOOTHING ELIMINATED, EXPLICIT ACCIDENT SEQUENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WIT 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE N
/
SHIELDING BASED ON NORMAL ACTIVITY FIGURE 27 49 8,
RISK OF LATENT CANCER FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10~3 =
l
=
~
(PEACH BOTTIN) a l
10
=
l
]
p CAT.f3 TCy,'TWy
~
CAI.f2TWy',TQUVT' N
I
. 10-5 x
g w
b 5
N 8
cc CAT.,fl TW, TCa I
)
CAT.f4 (SM00 THING) 10-7
?
\\
?
CAT.f5 NO MELT 10-8 0 5
10 15 20 25
)
DISTANCE (MILI3)
ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSUMED TO BE 10 / RE)
O SMOOTHING ELIMINATED, EXPLICIT ACCIDENT SEQUENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE
~ SHIELDING BASED ON NORMAL ACTIVITT FIGURE 28 82
RISK OF LATENT CANCER FATALITY TO AN INDIVIDUAL YERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10~3i RSS-BWR DESIGN RERASELINE
~
\\
\\
a 10 TCy
\\
=
10-5 5
C
?
Twy' g
y l
r-- TCy'
\\
10-6 g 3
3 g
~
l l!,"f-
?A X%.'c ~ ~ ~
'~
t Ey s
N.
S Ey'
~
~
~
10 0
5 10 15 20.
25 4
' DISTANCE (MILES)
ASSUMPTIONS:O OVERALL CORE MELT PROBACILITY ASSUMED TO BE 10 / RE O SMOOTHING ELIMINATED. EXPLICIT ACCIDENT SEQUENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER S' QUENCES WITH E
3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 29 83
-~--T-
-m
~
RISK OF LATENT CANCER FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10~3 -
l g
~
(SURRY)
~
j-CAT f3 '
~
/
I _4 E
E Z
CAT.f1 (COLD)
~
-l
~\\
HI 10-5 4
C A
5 CAT.f4 p
N 10
~
.f6 CAT.f1 (HOT)
CAT.f5 10-7 CAT.i 1
m CAT.f8 N;
~
)
g I
CAT.f9 10-8 20 25 0
S 10 15 TISTANCE(MILES) d ASSUMPTIONS: O,0VERALL CORE MELT PROBABILITY ASSUMED TO BE 10 / REACTOR YEAR O SMOOTHING ELIMINATED, EXPLICIT ACCIDENT SEQUENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE -
SHIELDING BASED ON NORMAL ACTIVITT FIGURE 30 84 v.-.
_ _ ~ _ _ _ _
~
RISX OF LATENT CANCER FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARICUS CORE MELT SEQUENCES 10~3 l
=
'(SURRY) 10 C
Z EVENT Y l
S Cs g
-5
\\
t
\\
E E
-6 7
x
,TMLB'4.T BOX 7 10'7
-=
l
=
10-8 0 5
10 15 20 25 DISTANCE (MILES) 4 ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSUMED TO BE 10 / RE O SMOOTHING ELIMINATED, EXPLICIT. ACCIDENT SEQUENCES O EXPECTED CONSEQUENCES FRCH 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL O ENTIRE' CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY l
FIGURE 31 85
t RISX OF LATENT CMCER FATALITY TO AN INDIVIDUAL VERSUS OISTANCE FOR VARIOUS CDRE MELT SEQUENCES 4
10
=
1
~
ICE CONDENSER
~
D.ESIGN
~
I 10-5
=
u.
Z L
l 1
10-6 l
g n.
..s l
S HF7
-lNLT, S HF7
[
f j
2 y
8
[/
Sp7
~
g 7
/
~
E
\\
!/
xg' l
l'
/
~
2
~
10-
~
S D7 y
l N
10"I 5
10 15 20 25
~
O DISTANCE (MILES)
ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSUMED TO BE 10 / REACTOR O SMOOTHING ELIMINATED, EXPLICIT ACCIDENT SEQU'ENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WITH l
3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVI1Y FIGURE 32 86
RISK OF LATENT CANCER FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10 I
=
~
DCONEE
~_
10
}
{~
1,
}
s
~
EVENT V Y
Y 10~5 -
u, i
p
,r 1-8 3
TMts a,y g
10 5 Cs
/
2
/
-7 b
f BOX 7
[
10-8 0 5
10 15 20 25 DISTANCE (MILES)
- / REACTOR YEAR ASSUMPTIONS:O OVERALL CORE MELT PROBABILITY ASSUMED TO BE 10 O SMOOTHING ELIMINATED. EXPLICIT ACCIDENT SEQUENCES O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES WIT 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 33 87
~
g.,
RISK 0F LATENT CANCER FATALITY TO AN INDIVIDUAL VERSUS DISTANCE FOR VARIOUS CORE MELT SEQUENCES 10-3 l
f CRYSTAL RIVER. -
M~ce-a h
io-4
_=
=
C
~
8 s
EVENT Y I
i 10-5
- =
F i
g 3
g-g-mua.1
\\:
=
l 10_
=
l
~
[
l 10-7 3 Cs 2
BOX 7 10-8 0 5
10 IS 20 2S DISTINCE(MILES)
ASSUMPTIONS: O OVERALL CORE MELT PROBABILITY ASSUMED TD BE 10 / REACTOR YEAR l
o SMOOTHING ELIMINATED, EXPLICIT ACCIDENT SEQUENCES O EXPECTED CONSEQUENCES FROI 91 WEATHER SEQUENCES WITH 3200 MWT POWER LEVEL 0 ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NORMAL ACTIVITY FIGURE 34 88
- - - - - - - - ~ - - - - ' - - - - ' -
Figure 35 89
~
distance for the designs presented; and Figures 29 through 35 give the corresponding expected individual risk of latent cancer fatalities.
These curves contain a significant amount of information concerning our current understanding of the risk dominant sequences and their relationship to the total design.
For example, consider Figures,10 and 22, the risk of fatality curves for BWR rebaselined analysis.
Figure 10 shows on a conditional basis that TCy' and AEa,S Ea sequences have significant potential consequences.
t But, Figure 22 shows that the TCy' sequence is orders of magnitude more likely than the others, and thus deserves relatively more attention.
In this figure you will note that there are a few key sequences which clearly dominate the early fatality risk:
TC; TW; and TQUV.
However, analyses of the Ice Condenser and Indian Point sets of curves indicate that many accident sequences contribute equally to the risk.
For such reactors it would[beJ% difficult to significantly reduce the risk by improving one aspect of the plant.
It should be noted that these curves represent a prospective view of accident risk as seen through the eyes of the fault tree and event tree reliability analysis techniques.
Other factors, such as plant management practices, and operator training, could have an important but presently unquantifiable role in assessing the rea.ctor design risks.
Risk Insights Into Reactor Design In order to get a measure of the total design relationships, the individual risk curves for the sequences can be integrated into a single probabilistically weighted curve.
This is done by summing the accident sequence relative probabilities at each specific distance. The curve then represents the average 90
s risk for a specific design.
Such curves for all of the designs are presented in Figures 36 and 37 for early fatalities and latent cancer fatalities expected individual risk respectively.
The design comparison curves give a perspective on the relationships between the importance of various types of reactor features, and therefore the importances of specific safety systems which control the icey sequences. The overall range given by the set curves represents our best guess of the spread of average individual risk which is posed by nuclear power today.
An additional perspective on the design variations is given in Table 22. The probability of severe core damage per reactor year measures the design's vulnerability to a significant core melt accident.
The probability of at least one fatality given a core melt, represents the relationship between the design and the chance that the design will reduce the impact of the accident to "relatively acceptable levels." Thus, it reflects the design's ability to mitigate the release associated with the severe accident. There is an order of magnitude spread between the design's probability of core damage, and a factor of 30 spread in the design's ability to prevent larger releases. A very rough average core melt probability for the designs which have been studied would predict about one chance in ten thousand (1 x 10 4) per reactor year for an average light water reactor.
The average probability of at least one fatality given core damage for the same average light water reactor would be about 20 percent or one chance in five. These numbers reflect an average and as can be seen in Table 22 there is significant variation about the average. The probabilities of severe core damage listed in Table 22 have very large uncertainties associated with them.
91
EXPECTED RISK OF EARLY FATALITY TO AN INDIVIDUAL VERSUS DISTANCE GIVEN A CORE MELT
- g 10-2 _
I A
8 k ;g$
s.
~
~
2
, gJ
10-3
%*L.
- 5. ICE CONDENSER DESIGN 5
4 3
Yk.-
- 7. OCONEE DESIGN
~._
- 6. INDIAN POINT DESIGN
~
- 8. CRYSTAL RIVER DESIGN
~
b h
\\.
f e
3 x
t C
3 2
x m.w g 3,
.m 6
..El 7-9..
. s; 10~0 I
2 U$
,; (
~
}
3_
5 1-Q.
[,
10-6 0
0.5 1.0 1.5 2.0 2.5 10-7 DISTANCE (MILES)
ASSUMPTIONS: O
- CORE MELT PROBABILITY ASSUMED TO BE 10 O RSS - BWR AND PWR DESIGNS ALL RSS CORE MELT ACCIDENT RELEASE CATEGORIES 1.
ALL RSS ASSUMPTIONS (E.G., SMOOTHING) 2.
O RSS REBASELINE AND OTHER DESIGNS
~
1.
SMOOTHING ELIMINATED EXPLICIT ACCIDENT SEQUENCES 2.
NEGLIGIBLE PROBABILITY OF VESSEL STEAM EXPLOSION 3.
O EXPECTED CONSEQUENCES FROM 91 WEATHER SEQUENCES 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND EXPOSURE SHIELDING BASED ON NOR.'iAL ACTIVITY FIGURE 36 92
s EXPECTED RISK OF LATENT CANCER FATALITY TO A VERSUS DISTANCE GIVEN A CORE MELT * -
I
~~
- 1
~
'C..
5.' ICE CONDENSER DESIGN 0-4',4,,t
- 6. INDIAN POINT DESIGN
~
^ -
- 7. OCONEE DESIGN
. CRYSTAL RIVER DESIGN 4'.
7 2 -5 g
V. 'N
~
3
~
y
..y ke 4
E
- A g
~
to-6 N
09 x4f Afgg m.
'J. 2.
~_
i
\\
6 0
10 20 30 40 S0 10-8 DISTANCE (MILES) 4 / REACTOR YEAR ASSUMPTIONS: O
ALL RSS ASSUMPTIONS (E.G., SMOOTHING) 2.
O RSS REBASELINE AND OTHER DESIGNS SMOOTHING ELIMINATED 1.
EXPLICIT ACCIDENT SEQUENCES NEGLIGIBLE PROBABILITY OF VESSEL STEAM E 2.
3.
O EXPECTED CONSEQUENCES FROM 91 WEATHER 3200 MWT POWER LEVEL O ENTIRE CLOUD EXPOSURE + 24 HOUR GROUND.
SHIELDING BASED ON NORMAL ACTIVITY
~
FIGURE 37 93
f e
. o Table 22 Risk insights versus design
- Probability Probability ***
of at least of core damage 1 fatality given-Reactor Name Type per reactor year core damage Peach Bottom BWR Mark I 2 x 10 5 1/3 Surry PWR Dry 6 x 10 5 1/10 Sequoyah PWR Ice Condenser 4 x 10 5 1/25 Indian Point PWR Dry 3 x 10 s 1/100 Crystal River PWR Dry 3 x 10 4
- 1/3 Biblis B (FRG)
PWR Dry 3 x 10 5
- (1/10-1/100)
Calvert Cliffs PWR Dry 2 x 10 '
- (1/4)
Oconee PWR Dry 2 x 10 4
- (1/4)
AThere are large uncertainties associated with the numbers in this table.
- Very rough preliminary estimates.
- Reflects median values point estimates.
94
Unless there are at least order of magnitude differences between the numbers, it is very difficult to attach significance to the estimates.
Representative Source Term The preceding information on the accident sequences and designs, lays the ground work for the development of a representative set of accident releases which can be used in the licensing process. After considering the accident sequence information, it was concluded that no specific accident sequence could be isolated as being representative of the source term. The range of possible selections is very large, starting with a zero point of considering no release therefore requiring no regulations, to developing requirements to cope with the worst physically possible accident regardless of its extremely low likelihood.
It is therefore necessary to develop regulations which will encompass the spectrum of possible accidents and consider the full range of possible consequences.
There are very large uncertainties associated with the development of such a
~
set of source terms, and it must be clearly recognized that these estimates represent our best judgments at the present time. As has been discussed
?.
l previously, a body of information has been presented 1,2 which questions the release of significant fractions of radioactive materials as has been modeled l
in the version of MARCH and CORRAL used in this report.
These data have not been considered in the development of this report and we currently do not feel justified to alter our predictions of the accident source terms other than by recognizing that our estimates could be conservative (over estimates).
95 l
.e In selecting a set of accident releases, which will represent the spectrum of accidents, it is important to clearly state the objective or purpose that such a set of source terms will serve.
It is not clear that one set of source terms will be sufficient for all purposes. The accidents that are used to establish siting or emergency planning regulations would probably not be appropriate as the set used to establish minimum engineered safety features or degraded core cooling requirements.
This is because siting and emergency j
response planning could be regulated on a generic basis, whereas the minimum ~
safety requirements should be much more dependent upon specific plant design consideration.
However, it must be clearly understood that the basis for the selection of a set of source terms should be consistent for any purpose and rest upon a reasoned and documented foundation.
~
Siting Source Term l
The siting source term has been developed to generically encompass several i
i potential reactor accident release characteristics and to cover accident releases ranging from the relatively benign to the " worst case" accident.
In this formulation the range of accident sequence severities is covered by considering release groups which account for a spectrum of possible degraded ESFs.
Selection of source terms involves the development of a set of reactor accident sequence groups and their associated release characteristics. This has been done by comparing the accident sequence type and release characteristics for l
l several currently available PRAs. Table 23 provides such a comparison for a
(
group of accident sequences which could be considered as dominant in terms of I
96 l
^ ' " ~ - ~ ' ' - -
e=wr-
_ - -v
-_m-_
l s
Table 23 Fission product release characteristics for dominant accident sequences obtained in several risk studies Accident sequence Probability Fraction of core inventory released to environment Ref. PRA (RSS terminology)
Per RU Xe-Kr I-Br Cs-Rb Te Ba-Sr Ru La RSS BWR TCy' 2x10 8 1.0
.45
.67
.64
.073
.052
.008 Peach Bottom TCy 8x10 8 1.0
.07
.14
.12
.015
.01
.002 TWy' 10 5 1.0
.1
.3
.4
.03
.03
.005 TWy 2x10 8 1.0
.003
.1
.08
.01
.007
.001 TQUVy
<10 5 1.0
.02
.05
.1
.005
.007
.001 AEa 010 8 1.0
.3
.6
.4
.07
.3
.004 Category 2 010 8 1.0
.9
.5
.3
.1
.03
.004 3x10 8 1.0
.3
.4
.2
.04
.02
~ Event V 4x10 8 1.0
.64
.82
.41
.1
.04
.006 u3 Category 2 010 5 0.9
.7
.5
.3
.06
.02
.004 N
H 10 7 1.0
.27
.37
.42
.05
.4
.003 RSSMAP S1 a HFy 5x10 8 1.0
.13
.57
.49
.07
.04
.007 Sequ ya,h k2 vent V 5x10 8 1.0
.48
.42
.086
.05
.03
.004 RSSMAP Event V 4x10 8 1.0
.48
.79
.44
.09
.045
.006 MQ-ily (THQH) 6x10 8
.8
.2
.2
.3
.02
.03
.003 0conee
}2MLUOy (TMLC) 4x10 8 1.0
.45
.74
.70
.08
.056
.009 2
T B MLU00'6 (TMLB')
10 8 1.0
.54
.74
.64
.08
.054
.009 3 3 T2A 20-6 (TMLB')
5x10 5
.9
.7
.5
.3
.06
.02
.004 T
IREP Crystal R. 3 4
smud
~
v site /public consequences.
It is noted that in general these sequences involve inability to cool the core resulting in core melt and breach of containment without effective operation of containment ESFs. Accident sequences of this type represent the most severe potential consequences.
By comparing the risks assoiated with the various accident sequence fission product releases the TCy
source term was selected as characteristic of this group and has been designated as SST1.
Using insights from previous PRAs on accident sequences and associated fission product releases, a set of characteristic source terms was developed. These are provided in Table 24.
A description of each source term regarding accident sequence and containment failure is provided below.
SST1. This source term is meant to be representative of the most severe fission product releases expected of reactor accidents.
It involves the failure or loss of multiple engineered safety features including emergency core cooling and containment.
Severe breach of containment due to steam explosion, overpressure including effects of hydrogen burning, and bypass of containment through interfacing systems are included in this group.
SST2.
Accident sequences involving a loss of core cooling with containment ESFs available are included in this category. The core is either severely damaged or melts and fission product release is by failure to isolate contain-ment or by delayed hydrogen burn and overpressure.
This source term was obtained from the RSS in Appendix 11, Section B.
It is relatively similar to PWR release category 5 of the RSS in terms of potential consequences.
98 e
^*
Table 24 Siting source term fission product releases Released fractions (to atmosphere)
S.S.T.
Xe, Kr I, Br Cs, Rb Te BaSr Ru La 1
1.0
.45
.67
.64
.07
.05 9x10 3 2
0.9 3x10 3 9x10 3 3x10 2 lx10 3 2x10 3 3x10 4 3
6x10 3 2x10 4 lx10 5 2x10 5 lx10 s 1x10 8 lx10 s 4
3x10 s 1x10 7 6x10 7 1x10 8 10 11 NEG NEG 5
3x10 7 1x10 8 6x10 s 1x10 10 1x10 12 NEG NEG l
m 99
SST3. As in the previous two release categories, failure of core cooling and core melt or severe core damage occurs.
Essentially all fission product release mitigation systems have functioned as designed.
In this case fission product release is by slightly degraded containment leakage (assumed to be 1 percent / day) and basemat melt-through. This release is the same as PWR release category 7 of the RSS.
SST4. This source term is meant to cover accidents in which the core is severely damaged (100 percent clad perforation) but core melt is prevented, as was the case at TMI-2.
It involves degradation of core cooling with contain-ment systems operation as designed.
Fission product release is by containment leakage, again assumed to be slightly degraded to about 1 percent / day.
Release category PWR 9 of the RSS was used for this case.
SSTS. This category encompasses limited to modest core damage accident sequences in which containment systems perform as designed.
It could be considered the equivalent of a Class 8 DBA LOCA.
Containment leakage of fission products is limited to 0.1 percent / day as designed.
As such, the releases associated with this group or one tenth of PWR 9 releases from the RSS.
Mean bone marrow doses vs. distance for the five SSTs are compared in Figure 38.
The doses vary by nearly eight orders of magnitude over the range of releases at all distances.
SST1 doses, and SST2 doses close to the reactor, are poten-tially large enough to cause early fatalities and injuries.
SST4 and SSTS doses do not exceed 1 mrem beyond 1/2 mile which is so low as to be of no concern for early health effects.
100
L l
NYC WEATIIER. NO EVACUATION 10*,w 3
0- SST1 N
sST2
.^
2-10 7 A-SST3 N
i iv 3-10,i
+- SST4 N
2 0
!c
,m sx 10-a-i em 5
M az 10 i o
i m
_4-10 r x<
E N %g 2
10 E s%
10-a-r
.=
w
-t --
10 i
0.5 1
jo DISTANCE (MILES)
Comparison of Mean lionic Marrow Doses for the Five Italease Categor ies.
(Calculations per foreneil using 3412 MWL PWR inventories, NYC weather, s
.. 1
- s Mean thyroid dose vs. distance is shown in Figure 39.
Mean doses from the five categories range over almost eight orders of magnitude for all distances.
SST4 and SSTS doses are below 1 mrem beyond 3/4 mile.
Close to the reactor SST1 and SST2 thyroid doses can be large, with doses from an SST1 large enough to potentially cause thyroid ablation.
Thyroid doses from an SST3 accident exceed the Protective Action Guides * (PAGs)1 of 5 rem thyroid dose close to the reactor.
4 102
---_-~-
s
.s -
.u s m m v m ma 9 F S P P oa m m U3 m m
.-* c U3 U3 U3 G3 m u3 O
6 l
I i
1 c-U.
v>
C 0
4
+
x az
_h
.9 ;
~
+
[
vo cs cu vC "a
4
-S5 7
O 3
o>z E<
r=
B s2 2
z.
-g u{
O M
dn A
5%
ai v
Bi y
!?
_o v
M 3E g
f U
8e o
2 Z
b 'd Es7
>=
b cw=
c h.b z c. a h
m$
u C
- J 0c cw >
oa=
mc Q
"u ": :
Y N.n z
.Eoc CUc Uw O
m W
6
,4 i i pii6 i 6 pini a piiii a piit6 i p istg piu6g risi e p..iig pu6g piang po44 W
M N
o I
i f
I i
3 O ooo o ooo o
o o
o m
m m
+
+
+
+
+
+
+
+
+
t (NHH) 3SOG GIOHAH1 NV3M FIGURE 39 103
t
- ^
References 1.
" Technical Basis for Estimating Fission Product Behavior During LWR Accidents," NUREG-0772, USNRC, June 1981.
2.
" Regulatory Impact of Nuclear Reactor Accident Source Tern Assumptions,"
NUREG-0771, USNRC, June 1981.
3..
" Reactor Safety Study," WASH-1400 (NUREG-75/014), October 1975.
4.
" Task Force Report on Interim Operation of Indian Point," NUREG-0715, USNRC, July 1980.
5.
" Theoretical Possibilities and Consequences of Major Accidents on Large Nuclear Power Plants," WASH-740, March 1957.
6.
" Reactor Site Criteria," Code of Federal Regulations - 10 Energy, Part 100, April 1962.
7.
" Calculation of Distance Factors for Power and Test Reactor Sites,"
TID-14844, March 1962.
8.
" Risk Assessment Review Group to the U.S. Nuclear Regulatory Commission,"
H. W. Lewis, et al., NUREG/CR-0400, September 1978.
9.
" MARCH, Meltdown Accident Response Characteristics, Code Description and Description and User's Manual," NUREG/CR-1711, BMI-2064, October 1980.
104
- ~ - - -
w-eeg
10.
" CORRAL-2 User's Manual," Battelle Columbus Laboratories, January 1977.
11.
" Interim Technical Assessment of the MARCH Code," NUREG/CR-2285, SAND 81-1672, November 1981.
12.
" Preliminary Assessments of Core Melt Accidents at the Zion and Indian Point Nuclear Power Plants and Strategies for Mitigating Their Effects,"
NUREG-0850, November 1981.
13.
" Steam Explosion" reference.
14.
" Reactor Safety Study Methodology Ap'plications Program - Sequoyah #1 PWR Power Plant," NUREG/CR-1659, Vol. 1, SAND 80-1897/3, Draft.
15.
" Reactor Safety Study Methodology Application Program - Oconee #3 PWR Power Plant," NUREG/CR-1659, Vol. 2, SAND 80-1897/2, Draft.
16.
" Reactor Safety Study Methodology Applications Program - Calvert Cliffs #2 PWR Power Plant," NUREG/CR-1659, Vol. 3, 5, SAND 80-1897/3, Draft.
17.
" Reactor Safety Study Methodology Applications Program - Grand Gulf #1 PWR Power Plant," NUREG/CR-1659, Vol. 4, 5, SAND 80-1897/4, Draft.
18.
" Crystal River-3 Safety Study," SAI-002-80-8E, May 1980, Working Draf t for Peer Review.
105
.s-e,,
19.
" Calculation of Reactor Accident Consequences, Version 2 - CRAC2 - Computer Code Input Data Descetption," NUREG/CR-2326, Draft.
20.
" Overview of the Reactor Safety Study Consequence Model," NUREG-0340, June 1977.
9 6
106
)
_