ML20211C841

Forwards Draft Summary of Results Presented in Seabrook PRA, for Review & Comment

ML20211C841
Person / Time
Site:	Seabrook
Issue date:	10/02/1984
From:	Ariuska Garcia LAWRENCE LIVERMORE NATIONAL LABORATORY
To:	Davis S NRC
Shared Package
ML20209C800	List: ML20209C798 ML20209C897 ML20209D266 ML20209D278 ML20209D286 ML20209D327 ML20209D479 ML20209D512 ML20209D547 ML20209D870 ML20209E015 ML20209E274 ML20211C165 ML20211C259 ML20211C363 ML20211C389 ML20211C430 ML20211C484 ML20211C558 ML20211C632 ML20211C641 ML20211C679 ML20211C769 ML20211C792 ML20211C841 ML20211C904 ML20211C920 ML20211C942 ML20211D042 ML20211D068 ML20211D076 ML20211D092 ML20211D109 ML20211D129 ML20211D140 ML20211D169 ML20211D178 ML20211D191 ML20211D198 ML20211D208 ML20211D226
References
FOIA-87-6 NUDOCS 8702200235
Download: ML20211C841 (48)

Text

{{#Wiki_filter:f g Lawrence Livermore National Late n. October 2, 1984 SS-099/DDD Ms. Sarah M. Davis U.S. Nuclear Regulatory Comission 7920 Norfolk Avenue Bethesda, MD 20814

Reference:

Sumary of Results Presented in the Seabrook PRA A rough draft of a "Sumary of the Results Presented in the Seabrook PRA" is enclosed for your review and comment. (Note that Section 3.2 is incomplete, as described below.) The content of this sumary, which may become a chapter, or part of a chapter in the final report of our review of the Seabrook PRA, is intended to provide a concise description of the results presented in the PRA, without comment or elaboration; 1.e., from their perspective--not from ours. This was accomplished by intentionally drawing heavily from the text of the PRA. Many quotes and near quotes from the PRA are included in the rough draft (all eventually to be credited). As consequence, the material presented in this rough draft may appear to be too concise and lacking development. Your comments on this aspect of the content would be welcomed. In addition, after beginning Section 3.2 of the enclosed sumn'ary, it occurred to me that Section 2.3 of the Seabrook PRA, which is titled " Major Contributors to Risk" would probably better present this portion of the Sumary of Results, so I did not complete this section. It may also be worthwhile to consider the possibility of using Seabrook PRA Sections 2.3 and 2.4 as replacements for nty entire rough draft. These sections, with minor modifications and additions to improve completeness (e.g., adding a table to define the 13 release categories) would provide a reasonably comprehensive sumary description from their perspective. If neither the approach taken in the rough draft or in Seabrook PRA Sections 2.3 and 2.4 provide the type of content and/or style of summary desired, a new start may be necessary. Please call me if you have any questions or suggestions on this material. Sincerely, f024-fr7-006 AA.cM Abel A. Garcia Ob(/ Principal Investigator G702200235 gyog3, PDR FOIA SHOLLYB7-6 PDR .,. : n... - .,, u s

~ n DRAFT FOR COMET Sumary of Results Presented in the Seabrook PRA 1.0 Ouestions Answered by the PRA The three questions listed below provide a structure for the analytical work of the Seabrook PRA (SPRA) and a framework for organizing the numerical results. o What is the likelihood of core melt? What is the likelihood of release of radioactive materials as a o function of release magnitude? o What is the likelihood of damage to public health and property as a function of the level of damage? The answers to these questions developed in the SPRA are briefly described in Section 3.0 herein. 2.0 What was Considered in the Analysis 2.1 Initiating Events The SPRA included consideration and quantification of the 58 initiating events listed in Table 1. The table includes the code designator used for each initiator 5 in the study. 2.2 Plant Damage States f The SPRA considered 39 plant damage states in the risk model, i There PDSs are listed and defined in Table 23. Their relationships to one another are illustrated in Table 26. l b h t't! I 1 u n.- w ! nnr r rr tw.., R i

I 2.3 Release Categories The SPRA used 13 release categories to represent the spectrum ofrelease states. These 13 release categories can be furthercategorizea into three groups: one group for categories in which the containment structure remains intact and isolated, a second for release categories that involve a gradual, long term degradation of containment integrity, and a third which involves early containment Definitions of the release categories in each of these three failure or bypass. groups are presented in Table 3. Each release category is represented by a three-The code designator consists of an S, which denotes applicabilityto part code. Seabrook Station, a number to indicate the absenceof containment failure mode or .) to indicate the absence of containment filtration by the containment building spray system, and a V to indicate the presence of an additional vaporization component of thesource term for scenariosin which the molten core debrisexperiencessustained elevated temperatures. Certain symbol combinations corresponding to additional release categories which can be hypothesized are precluded by Seabrook Station design features. For example, there is no 55 because the containment building sprays must function with adequate heat removal to keep the containment structure intact. 3.0 Results 3.1 Classes of Results Presented l The SPRA presents results for the 7 classes of damage indices i listed below, with consideration of uncertainty for each. o Core melt frequency Early fatalities; those occurring wi. thin a short time af ter o exposure. Injuries (radiation illness not leading to fatality :. o {

9 Thyroid cancer cases (total occurring over a 30-year period); those o resulting mainly from iodine ingestion and curable medically in about 90% of the cases. Latent cancer fatalities; those from cancers other than thyroid o cancers, occurring over a 30-year period. Total population dose, or man-rem (whole body gamma dose). o o Total public property damage and evacuation costs in dollars. The results are also presented in terms of single unit and station (2-unit) risk. 3.1.1 Numerical Results The.esults of the SPRA are presented in a probability of frequency format for each of the seven classes of damage indices. 3.1.1.1 Single Unit Risk The probability of frequency of core damage (melt) is presented in Figure la (probability density) and Figure Ib (cumulative probability). The results for the remaining six damage indices are presented as risk curves including uncertainty in Figure 2a through 2f. 3.1.1.2 Two-Unit Station Risk The evaluation of risk of the Seabrook Station with two operating reactors required that the following factors be taken into account. o The increase in the likelihood of accidents because of the presence of two reactor units and the potential for interactions between the units. The possibility of simultaneous accidents in response to initiating o events that affect both units. I The possibility of common cause failures between units in response o 1 to an initiating event that affects both uaits.

1 The results for core melt frequency are presented in Figures 3a (probability density) and 3b (cumulative probability). The difference inthe mean frequencies of single unit and station (2-unit) core melt events per year is less than a factor of two because of the elimination of double counting of those initiating events that cause a core melt in both units. In other words, Figures 3a and 30 illustrate the probability of frequency of one or more (two) core melts per station year. The contributions to the total event frequency from single and double unit events are illustrated below in terms of mean values. Event Mean Frequency (Events Per Station Year) Core Melt Involving 4.0 E-4 One Unit (2.0 E-4) x 2 Core Melt Involving 0.3 E-4 Both Units Total 4.3 E*4 The effect of two-unit operation on the risk curves is illustrated in Figures 4a and 4D for early fatalities and latent cover fatalities respectively. These figures contain plots of the mean value risk curves for two-unit (station) operations and the corresponding curves for single-unit operations. The two-unit station risk curves includes contributions from single unitaccidents as well as from double unit accidents. Figure 4a is indicative of the damage indices for early I health effects, whereas Figure 4b is indicative of latent effects. 3.2 Principal Contributors to Core Melt and Risk l The matrix formation used in the SPRA allows ready indetificaticr of major contributors to the numerical risk results by a process ofdecompositio" cf tne ris6 r.atrices. Tne procedure for systematically detem. iring the principal r s,

l 9 contributors begins with the risk curves and works progressively backward through the risk model to determine the most important paths insofor as numerical contributions are concerned. 3.2.1 Dominant Release Categories The dominant contributions are identified by determining the relative importances of the release categories at various damage levels, in terms of the relative frequency of exceedance values. Tables 4a and 4b present the relevant matrices for early fatalities and latent cancer fatalities, respectively, and Figures 5a and Sb are plots of the information in the matrices. Summary tabulations of the dominant release category contributions to risk for these risk indices are shown in Tables Sa and 5b, respectively. The results of Figure Sa and Table Sa clearly show that release category S6V dominates the risk of early fatalities across the full range od damage a level presented and S2V makes a small contribution. Release category 51 contributes to an extended tail of the risk curve at frequencies below 10-9 as seen in Table Sa. None of the remaining 10 release categories were found to make a significant contributions to early fatality risk. In Figure Sb and Table 5b, it seems that release categories 56V and 52 V also contribute significantly to the risk of latent cancer fatality, especially in the low frequency range of the risk curve. In the low consequence relatively high frequency range of the risk curve, release categories S3V and S3 make significant contributions. Category S4V makes a small contribution across the full range of the latent cancer fatality risk curve, andcategorySS has only a small contribution to the risk with no potential forlarge numbers of latent cancer fatalities. o sl

The importance of a release category, as measured by its influence on the risk curves, depends not only on its frequency but on its potential forproducing variouslevels of damage. The singular effect ofeachrelease category's potential for producing damage is measured by the elements ofthe S matrix whose results have been plotted in Figure 6a for early fatalities and Figure 6b for latent cancer fatalities. In figure 6a, it is seen that over much of the damage level range, category S1 has the highest potential to produce early health effects given its postulated occurrence, followed by categories S6V and 51, withy category S2V having a much lower potential to produce damage. However, as shown in the previous section, categories S1 and S1 have an extremely low frequency of occurrence relative to categories 56V and 52V; thus, they make extremely small contributions to risk. Hence, upon completing the first step of the risk unraveling process, it has been determined that of the13 release categories utilized in the risk model togroup the billions of accident sequences that were analyzed, only 4 release categories make significant contributions to risk. These are categories S6V, S2V, S3V, and S3. Of these, only categories S6V and 52V make significant contributions to the risk of early fatalities. From the results of the previous section, in particular those in Table, it is seen that it is necessary to add only two more categories to the above (namely, categories SS and S4V) to obtain all categories making significant contributions to core melt frequency. Hence, 6 out of 13 release categories were found to be significant with respect to their contribution to risk or to frequency of occurrence according to the following key: l l

Significance Relative to: Release Category Risk Occurence Early Latent Frequency Effects Effects 56V Major Major Minor 52V Minor Major Major S3V None Major Major 53 None Major Major 54V None Minor Major S5 None None Major 3.2.2 Dominant Plant Damage States Identification of the dominant plant damage states is accomplished by using the matrix tabulations of release category frequency of occurrence versus plant damage state. Tables 7a and 70 show these matrices for early fatalities and latent cancer fatalities. By examining the columns of these matrices corresponding to the risk dominant release categories identified in Section 3.2.1, the dominant plant damage states are identified. (Not yet Complete) 3.3 Comparative Results The purpose of this section is to provide a perspective for viewing the absolute value of the risk levels calculated for Seabrook through the use of limited comparisons. Four types of comparisons are made, as follows: o to risk levels calculated in PRAs of other nuclear power plants o to risk four sources ofenergy other than nuclear power l l

l O to sources of risk other than energy production to the provisional safety goals currently under evaluation by the o NRC Each of these comparisons is described in modest detail below. 3.3.1 Comparison to the Risk from Other Nuclear Plants f PRAs have been performed on about 20% of the nuclear power plants in operation or under construction in the U.S. These are wide variations in the scopeof these PRAs, which includes variations in the extent ofthe analysis, the type It is nevertheless of plant, thesite, as well as manyother types of differences. possible to divide these PRAa into two general classes: limited scopeand full Limited scope is defined as not including consideration of external events scope. and limited to an evaluation of the core melt frequency. There are 13 of these. Full scope is defined as generally including consideration of external events and evaluating health risk. There are four of these. The limited scope PRAs have mean values of core melt frequency ranging from 0.1 to about 3 core melt events per year, with an average of about 0.5 events per 1000 reactor years of operation. The seabrook result is 0.23 events per 1000 years of reactor operations near the lower end of the range. The full scope PRAs have mean values of core melt frequency ranging The Seabrook result from less than 0.1 to about 1.6 events per 1000 reactor years. again compares favorably. An examination of the contributions to health risks and core melt frequency in the fuller scope PRAs reveals that there are general similarities in the results, such as the frequent appearance of fires and seismic events as risk contributors, and small LOCAs as important core melt contributors. In addition, two t i t

cases (including Seabrook) have the interfacing systems LOCA as the dominant contributor to early fatality risk. 3.3.2 Comparison to the Risk from Other Sources of Energy The issue of comparative risks of alternative energy sources is based on the final report of the National Research Council Committee on haclear and Alternative Energy Systems (CONAES). In essence, the CONAES report states that for routine operation nuclear power poses smaller risks to the public than other alternative energy sources except for natural gas. For accidents, the large range of uncertainty that still attaches to nuclear risk calculations makes itdifficult to provide a confident assessment of catastrophic reactor accidents, even thoug') the projected mean number of fatalities is probably less than the risk from routine operation of the nuclear fuel cycle. 3.3.3 Comparison to Sources of Risk Other than Energy Production This comparison also based on the CONAES report, indicates that energy sources, in general, pose accidental death rates less than 11 of the rate dueto all other causes of accidents such as automobil accidents, drownings, etc., and that this result is probably independent ofthe uncertainties in the calculation of nuclear accident risks. 3.3.4 Comparison to NRC's Provisional Safety Goals A comparison to NRC's provisional safety goals is presented in Table 6. As can be seen in this table, the risk of early fatalities to the 4,435 individuals within 1 mile of the plant was found to be a factor of between 5 and 6 below the individu8al risk goal, and the4.2 million people witnin 50 miles ofthe plant were found to have an individual risk oflatent cancer fatalitymore than two orders of magnitude below thesocietal risk goal. Note that the values calculated in Table 6 for Seabrook Station are mean values. The mean values are used to obtain the best match with the statement of the risk goals in celu9 5. U911kE thf media" l

values, the mean values are significantly influenced (increased) by the uncertainties that were quantified. With regard tothe core melt frequency design objective, the results of this study of 1.9 x 10-4 events per reactor year (median) are within a factor of two of the design objective. The use of median values for this core melt frequency comparison has been suggested by the NRC forthe trial use of the safety in view of the facts that the Seabrook results include contributions from a full spectrum of external events and that the NRC has indicated that care should be taken in the apportionment of external events to the design objective, the SSPSA results are viewed as comparing favorably with the design objective. If the contributions from seismic events, fires, and other external events were not included, the SSPSA results for the median core melt frequency would have been about 1.3 x 10-4 events per reactor year. In light of the underlying uncertainties, there is not a significant difference between 1.3 x 10-4 and 1.0 x 10-4 events per reactor year. i 4.0 Key Findings and Insights in general, insights from the PRA are presented in the Sumary Report. l beginning on page 17. A preliminary listing of insights believed to be important is l provided below, in our words. o Risks are low, but the core melt probability is higher than the proposed safety goal. o A very large number of sequences contribute to the total core melt probability. The single most dominant sequence contributes less than 15% of the total, and the top 27 sequences contribute just over half to the total. n. .,--.n e- - - - - --, -n-- c-, -, -, - - ----

P The V-sequence accident totally dominates the risk of early o fatalities. External events are not important risk contributors (this result is o not consistent with other recent PRA results: Zion, Indian Point, Oconee (NSAC/ utility assessment), Millstone Unit 3). o The most important initiating event in terms of core melt probability is loss of off-site power. 1 I

The V-sequence accident totally dominates the risk of early o fatalities. External events are not important risk contributors (this result is o not consistent with other recent PRA results: 2109 Indian Point, Oconee (hSAC/ utility assessment). Millstone Unit 3). o The most important initiating event in terms of core melt probability is loss of off-site power. _.,,_...,.--~,-.---+---w

2.3 MAJOR CONTRIBUTORS TO RISK There are many different ways to dissect the results of a PRA to detemine The risk the major contributors to risk and to core malt frequency. model that was developed and quantifled for Seabrook Station contains several billion accident scenarios. One aspect of finding the principal risk contributors is the ranking of these scenarios with respect to risk The approach that was followed to detemine the relative contribution. importance of the scenarios was to first rank the major groups of accident scenarios as defined by the pinch points of the rf sk model; 1.e., the initiating events, plant damage states, and release A systematic way to quickly focus on the most important categories. groups of accident sequences determined by the pinch points is afforded d._..'_,matgf,omgism of risk.!" '-{q.;; 0QQ yg;{. ._..., ucr m eo on um ....,,m Using this systematic procedure and working backwards Sw U un 13. through the risk model, a detemination is made of the release categories that control the risk curves for each damage index, the plant damage ) states that significantly contribute to the important release categories l and to the risk curves, and the initiating events that give rise to the most important paths through these plant states and release categories to ^ The results of this systematic l their ultimate impact on the risk curves. matrix analysis of the Seabrook Station risk model,_9'" b h.er.^cd l are sumarized below. s Of the 13 categories of radioactf ye release utilized in the study, release category HV was found to completely dominate the public risk of 'ategory NV is typified by a core melt with early early health effects. containment bypass with an opening greater than 3 inches in diameter, no filtration by the containment butiding sprays, and a vaporization compo-nent of the radioactive release source tem as are all the categories The dominance of category NV to early fatality i l with a V designator. The only other release category risk is illustrated in Figure er3-+e. that was found to make a noticeable contribution to early fatality or illness risk is release category UV. Release category 37V is similar to category EV except that the opening is less than 3 inches in diameter and containment overpressurization faildre eventually occurs. With respect to latent health effects, the relative contributions of l sequences grouped by release category were found to be dependent on the level of damage or portion of the risk curves under investigation as seen in Figure 0.0 3 In the upper left-hand portion of the risk curves, release categories UV, U, and MV make significant contributions, whereas categories UV and EV dominate the lower right-hand portion. Category RV makes a small contribution across the full spectru J the containment without filtration by the containment building sprays. Release category RY denotes core melt scenarios with containment failure associated with core debris penetration of the basemat concrete of the l On an expected value basis category UV ranks t containment structure. first followed by release categories U, UV, and 56V with respect to latent health risk. I r l

The only additional release category out of the set of 13 in the risk model found to have a significant f requency of occurrence and, hence, is important with respect to core melt frequency, is category 55. Category $5 is typified by a core melt with a containment that success-fully isolates and remains intact throughout the accident, and because of this results in no potential for early health effects and a very low potential for latent effects. Hence, of the 13 release categories in the model, 6 were found to be significant with respect to risk or occurrence frequency according to the following key: Significance Relative To: Release Category Risk Occurrence Early latent Frequency Effects Effects 15Y Major Major Minor T77 Minor Major Major T37 None Major Major 37 None Major Major T47 None Minor Major 55 None None Major The search for the most important accident sequences is narrowed further The plant by grouping them with respect to the plant damage state. damage states are the pinch points that group accident sequences emana-ting from the plant model by similarity with respect to their expected behavior in the case and containment response model. A total of 39 plant damage states was used in the Seabrook Station risk model and these are Given specifications of the plant damage state, defined in Table t:t t. the probability of progressing down each path in the containment event Since the search for dominant accident tree can be uniquely detemined. sequences has already been narrowed down to six release categories, the next logical step is to determine which plant damage states dominate the These dominant plant damage states are the same as those i risk curves. that respectively dominate the risk dominant release categories. To obtain a final ranking of these plant damage states, it is necessary to examine their contributions to the risk curves without reference to However, prior to this ranking, the entire set of release category. plant damage states that riake significant risk contributions is deter-mined by ranking the plant danage states in each of the six release At ec e 1* Se:t4a Q categories detemined alone to be significant. The the tratria forralisr of the ris6. rodel f acilitates tre search. results cf tMs ste; of the ris6. u"avelin; pro:ess are descrite: telo,. 2.2 : l i

Upon ranking the plant damage states, it was found that of the containnent, were highly correlated to specific releaseThe "F states, particularly IF and 3F, which are typified by a large, unfiltered containment bypass, were found to dominate release categories. The "FP" states, category EV, and hence, the risk of early fatalities. particularly 3FP and 7FP, which are typified by a small, unfiltered Release containnent bypass, were found to dominate release category UV. categories found earifer to be important with resgect *o latent healthD" states, p risk (U, UV and T4Y) were all dominated by the 3D, 70, and 8D. The "D" states are typified by an isolated containment with no sprays and With no containment heat removal, contain-no containment heat removal. ment failure would eventually occur because of the decay heat en generation. risk significance for "D" states are basemat melt-through (RV) and In sunaary, of the 39 plant damage overpressurization (U and UV). states employed in the risk model, 9 respect to risk or core melt frequency according to the following key: Significance Relative To: Risk Plant Damage Occurrence State Early Latent Frequency Effects Effects IF Major Major Minor 3F Major Major None 7FP Minor Major Minor 3FP Minor Major Minor 8D None Major Major 70 None Major Major 3D None Major Minor 8A None None Major 4A None None Minor The final step described here in the process of detemi A total of to examine sequences with respect to the initiating event. 58 initiating events was fully quantified in the risk model and a listing The list of initiating of these events is presented in Table Mevents includes 6 loss 4 support system faults, 8 seismic-induced loss of coo The initiating 3 aircraf t crashes; 4 flooding events; and a truck crash.eve f re;.et:y can te viewed fron several differert pers:ectives a9d cc e Pfit f or-aii s-pren " !Mt4cc=hh C9e w!y te a:::-:' * ; :: t*f at:-in

in view the significance of particular initiating events is illust events, shows the frequency of all accident sequences associated with Table M-h Uncertain-each initiating event and risk significant release category. ties in selected initiating event frequencies are illustrated in melt frequency was calculated to be low, the uncertainty in quantifying Figure M-lhr the frequencies of individual events that were propagated through the risk model was high. Release category WY, which was identified as the dominant contributor early health risk, is seen to be driven by sequences associated with interfacing systems loss of coolant accident and to a lesser extent by Seismic events dominate category UV and make seismic-induced events. significant contributions to categories RV, MV, U and, henc latent health risk. with respect to latent health risk over much of the r power, which also appears as a significant contributor to categ 33, and $5 and the first ranking contributor to core m

Finally, significant contributions to the latent health risk categories.

transient events and small LOCAs collectively health risk categories, especially category U.su sne v isi, cur c eri of-v;cident ;;;; nce; gr;@cd by fin us6vr pre:erted e Sectica 13 1 The final column of Table erS-S provides a mean melt of the resulting accident sequences. is presented in condensed form in Table >+4 which also indicates th A total of 58% of the core melt importance of groups of initiators. frequency is attributed to com which cause both a plant transient condition and a f ailure or degradatio of one or more systems needed to recover the plant to a stable conditio Comon cause initiating events found to be particularly important include Transient initiators loss of offsite power, seismic events and fires.are seen to with the remaining coming from the loss of coolant inventory group, especially small LOCA. Having narrowed the search for what is driving t sense to find out which of the several billions of accident sequences within these groups are the most important and are deserving of focus A computer-aided, systematic subsequent risk management activities. search of the risk model was performed to find these risk dominant pwurJ.nc sen,ibed in Sc;tien'.? end sequences.;cc;rdi ; te ihr p'"G"PMhe dCtGils M the m-A 4 tM' WlY'"* d;lin;nted O!ia0 t? A surr;8ry of the results is treercn are presented ID Sistien 13.2.This table describes a total of 22 scenarios presented in Table M-f.in:1uding the first 20 in the list of scenarie

.: i

....+

1 e core melt frequency and the single scenario found to rank first with The respect to early health risk, an interfacing systems i among those with significant contributions to latent health risk. The first sequence in Table Sv3=fi ranks first with respect to core melt Station blackout is postulated to frequency and latent health risk. occur in the sequence following a loss of offsite power initiating event, failure of the onsite AC power system, and failure to recover electric The core damage that is postu-power before the onset of core damage. lated to occur along this se The seal damage inventory through damaged reactor coolant pump seals. and failure of high and low pressure makeup capability is a conseq effect of the loss of the primary component cooling system, which cann Note that in describing the failures that occur along the scenarios, Table ih9=1E distinguishes between indepe run without electric power. Independent system failures and human actions and dependent failures. are those that occur in addition to the initiating event, and hence, D define the sequence frequency. have a conditional frequency of 1 given occurrence of the initiating event and independent failures. The second sequence in Table tr6-5 also begins with a loss of offsite power and is followed by a failure to provide servi successfully start along this sequence.The sequences ranked eighth and ninth scenario sistlar to the first one. with respect to core melt are variations on the above with diesel generator failures on one train and service water failure on the train, and the same end result. core melt is similar to the first sequence except that offsite power is lost as a result of a turbine building fire and, hence, is scenarios resulting from a loss of offsite power an recovered. There are seven additional of AC power, an unmitigated pump seal LOCA. scenarios in thf s table that also involve a loss of PCC and, hence, an I unmitigated pump seal LOCA, but for which AC electric power is These sequences are ranked 4th, 7th,10th,11th,12th,15th, I and 16th with respect to core melt frequency and include those in wh available. the PCC system is lost due to internal system failure, as well as

Thus, dependent failures caused by fires and the loss of service water.

a total of 12 of the 22 scenarios in Table ered invol and a resulting pump seal LOCA.the table are dependent failures r ) f of service water, or damage done by a fire. Three scenarios were identifled in the top 20 among core melt j contributors that were initiated by a small LOCA and subsequent residual heat renoval capability. da resid;al heat re.cval (RHR) problems in the case of ir

• r e re air.' ; s:c a <:s i

cases. tM c:::s'te t adr. in Ve re 3'*; t=: O GA

4 in Table W include two sequences which are classified as transients without scram, three that involve operator failure to establish long tem heat removal after transient, a loss of one DC bus scenario with failure of the emergency feedwater system, and the interfacing systems LO;A The latter sequence ranks first out of all sequences in tems scenario. Me,e Jeteft.,rg.ce:ni, the of its contribution to early health risk. ethers cen be fe.nd irc analy>is ef these wmd sequences nd imuy Sectic-inh A concise sumary of the infomation presented above on contributors to In this risk and to core melt frequency is presented in Table M-6. table, the 13 release categories have been arranged into three groups: one with potential for early and latent health effects; one with potential for latent health effects, and, at most, small potential for early health effects; and the third with the damage level limited to damage to the core with at most small number of latent health effects. This table shows that these consequence categories are highly correlated The primary concern with early health to the containment response. effects is with early containment failure, bypass, or a large open These release categories have a total mean frequency of penetratipn.or about li of the total core melt frequency. Delayed 2.4 x 10-0, overpressurization, basemat melt-through and small open penetration are seen to be the predominant causes of latent hepith risk, and are or about 731 of the estimated to occur at a frequency of 1.7 x 10-Consequences are largely confined te core total core melt frequency. damage for releases from an intact containment, at a frequency of 6.0 x 10-5 or about 26% of the core melt frequency. 2.

3.1 REFERENCES

Kaplan, S., and B. J. Garrick, "On the Quantitative Definition of 2.3-1. Risk," Risk Analysis _, Vol. 2 No.1. March 1981. i

TAE'.E 2.3-1. DEFlh! TION CF PLA* T DAMA3E STATES LSED IN SEA 5R00s: STAT 10', F!5K MSDEL Conditioes at Time of Restter EI'EI Vessel Melt inrosgh cam;e Certainmett Co*ditice State y,gg,3 g,,, Cose Time pp,gg,,, g,,ggy i 1D tarly Los try 1s:1sted, he $p'ays, h: heat Re3:.at IF !arly Low Dry Snassed. Large Opening. No Filtratice IFF te ly Low Dry tnessed. Small Opedng, ho Filtration 1FA terly Low Dry Aircraf t Crash. No Filtration 2A tarly Lou Wet Isoisted. Sprays. Heat Rttval 2C terly Lov Wet 1solated. Sprays, ha heat Rencval fi; tarly Lov Wet Isolated. No Sprays, ho Heat Rexval 21 ta*1y Low Wet S nassed. Large 0;teing. F11tratice 2F tarly Lou Wet ly;assed. Large Opening. No Filtratice 2FP.tarly Lov Wet ty;assed. Small C;edeg. No Filtratise. 2FA Lasty los Wet Aircraft Crash, ho Filtration 3: Early Hig% Cry 1sciated. No $peays ho West Fe,cial 3r ta*1y High cry ty;assed. Large Opentes. Mc Filtratice ; 3rP ta*1y Hig% Dry By;assed. Sna11 Opening No Filtratic9 4A tarly High Wet Isolated. Sprays. Heat teievat 40 tacly High Wet u Isolated $pesys, ha > eat Reisvat 4: taey Hig* Wet 15:1sted, tic Sprays. %: Fest fea:.a1 8 4! 'tarly High Wet Inessed. Large Cpening. Filtration of Early High Wet typassed. Large Opentag, ho Filtration 4FP tarly High Wet typassed. Sea 11 Opening, ho Filtration 64 Late Low Wet Isolated. Sprays. Heat Removal 6C Late Lov Wet Isolated. Sprays. No Neat Remval 53 Late Low Wet Isolated. No Sprays. No Heat Aereval 6! Late Low Wet typassed. Large Opening. Filtration 6F Late Low Wet ty;assed. Large Openirs. No Filtration i l 6FF Late Lo. Wet ly;assed. Small Cperirg. No Filtratio9 i 6FA Late Lee Wet Aircraft Crash, ho Filtration [ 7C Late High Dry 1solated. he Sprays. No Heat Rep. vat 7F Late High Dry typassed. Large Openteg, ho Flitration 7FP Late High Dry typassed. Small Opening. No Filtratica 8A Late High Wet Isolated. Sprays. Heat Removat 8 Late High Wet isolated. Sprays. No Heat Remeral SD Late High Wet Isolated. No Sprays, ho Heat Remval ( I 8t Late High Wet Inassed. Large Opening. Filtratter. 8F Late High Wet Snessed. Large Opening, ho Filtration 8FP tate High Wet Spassed.-$nall Openteg he Filtration Intued. Sprays. Wat Rmal IA Special States for $ team > g: cererator 1.te R otre inassed. 5,ra,s..o nai.e-vai I 9: !;;;ngeti;,;'"""' pnaat. h: 55an.ai~> > ~.' i 9 5 1

T AB.E 2.3-2. Ih111 ATlH3 EVEhiS SELECTED F0F. QUA'i11FICAT10N.0F THE SEABROOK STATION RISK M3 DEL Initiating Event Categories Selected Code Group for Separate Quantificatier. Designator e Loss of Coolant 1. Excessive LOCA ELOCA Inventory 2. Large LOCA LLOCA 3. Medium LOCA M.0CA 4. Small LOCA SLOCA 5. Interfacing Systems LOCA Y 6. Steam Generator Tube Rupture SGTR RT e General 7. Reactor Trip Transients 8. Turbine Trip TT 9. Total Main Feedwater loss TLMFW

10. Partial Main Feedwater Loss PLKPW

11. Excessive Feedwater Flow EXFW

12. Loss of Condenser Vacuur LCV

13. Closure of One Main Steam Isolation Valve (MSIV)

IMSIV

14. Closure of All MSIVs AM51V

15. Core Po.er Excursion CFEXC

16. Loss of Primary Flow LOPF

17. Steam Line Break Inside Containment SLBI

18. Steam Line Break Outside Containment SLB0

19. Main Steam Relief Valve Opening MSRV

20. Inadvertent Safety Injection SI e Common Cause Initiating Events

- Support

21. Loss of Offsite Power LOSP System Faults

22. Loss of One DC Bus L1DC

23. Total Loss of Service Water LOSW

24. Total Loss of Component Cooling LPCC Water

- Seistic

25. 0.7g Seismic LOCA E.7L I

Events 26, 1.0 Seismic LOCA E1.0L

27. 0.2 Seismic Loss of Offsite Power E.2T

28. 0.3 Seismic Loss of Offsite Power E.3T

29. 0.4 Seismic loss of Offsite Power E.4T

30. 0.5 Seismic Loss of Offsite Power E.5T

31. 0.7 Seismic Lots of Offsite Power E.7T

32. 1.0g Seismic loss of Offsite Power E1.0T

,S'.I

.3-2,::.:".!:.

Initiating Event Categories Selected Code Group for Separate Quantification Destgrator FSRCC - Fires

33. Cable Spreading Room - PCC Loss

34. Cable Spreading Room - AC Power Loss FSRAC FCRCC

35. Control Room - PCC Loss

36. Control Room - Service Water loss FCRSW FCRAC

37. Control Room - AC Power Loss

38. Electrical Tunnel 1 FET1 FET2

39. Electrical Tunnel 3 FPCC

40. PCC Area

41. Turbine Building - Loss of Offsite FT8LP Power TMSLB

- Turbine

42. Steam Line Break TNLL Missile

43. Large LOCA TNLCV

44. Loss of Condenser Yacuum

45. Control Room Impact TMCR

46. Condensate Storage Tank Impact TNCST TNPCC

47. Loss of PCC

- Tornado

48. Loss of Offsite Power and One MELF Missile Diesel Generator MPCC

49. Loss of PCC

50. Control Room Impact MCR APC

- Aircraft

51. Containment Impact ACR Crash

52. Control Room Impact APAB

53. Primary Auxiliary Building Impact FLLP

- Flooding 54 Loss of Offsite Pcwer

55. Loss of Offsite Pcwer and FLISG One Switchgear Rocm

56. Loss of Offsite Power and FL2SG Two SwItchgear Rooms

57. Loss of Offsite Power and Service FLSW Water Pumps

- Others

58. Truck Crash into Transmission Lines TCTL

!!!:.T' ::!

_ _ - = _ _ - - _ _ PERSPECTIVE ON THE RESULTS_ i 2.4 In the previous section, infomation was presented to quantify the ris to public health, safety, and property due to potential reactor accide This infomation has included an enumeration of at Seabrook Station. potential accident scenarios, estimates of the frequencies and dama i levels of these scenarios, and a qUantification of the un 2 use of some limited comparisons, a perspective for viewing the abso these estimates. This perspective will enable the risk levels that were calculated. reader to compare, to the extent that such comp plants other than Seabrook Station, energy sources other than nuc Further power, and sources of risk other than energy production. perspective will be provided by comparing the results of the PRA the provisional safety goals currently under evaluation by the Nuclea Regulatory Comission (NRC) as part of a 2-year trial periodA l between the results of this report and those of the Phase I study wh (Reference 2.4-1). l was completed earlier in this project (Reference 2.4-2). COMPARATIVE RISKS OF NUCLEAR PLANTS 2.4.1 To date, roughly 20% of the nuclear power plants f The differences in scope of these least a limited scope PRA evaluation. i studies range from full scope as in the cases of Seabrook Station, of the RSSMAP and IREP studies. There are eve i The scope limitations among between the $$PSA and Indian Point 2 and 3. all the completed PRAs in some cases are due to excluding external l events, internal fires, floods and comon cause failures; in other cases, In several cases, i they are due to excluding uncertainty quantification. i all or part of the damage calculations (i.e., core tven if all the differences in scope were eliminated, risk estimates that include consequence assessment would still exhibit w l

scope, f

variations because of the df fferences in the site characteristics, meteorology, demography, topography, etc., that would tend to maskEve l differences due to the plant design and operations. i out differences in consequence assessments an l site independent. i l In light of these differences, we have elected not to attempt com of risk curves from different PRAs, but some lim l ( These limited comparisons for Seabrook Station into better perspective,are made o This is accomplished l in scope among those PRAs that have been published.The first category, whichi by separating the PRAs into two categories. cco:ste ef tt:te stweies Wet er:1ude er cely cse eitt e1 c.c ts e : irte s* f eat at: fi:::: 4 > 4ct ccestit.

• t't l

I I i

A total of 13 published Vast najority of the PRAs cenpleted thus f ar. limited scope PR follcwing plants. Arkansas One 1 (! REP) Biblis B (German Reactor Safety Study) o e Browns Ferry 1 (IREP) e Calvert Cliffs 1 (! REP) e Calvert Cliffs 2 (RSSMAP) e Crystal River 3 (! REP) e - Grand Gulf (RSSMAP) e 1.imerick (Industry Sponsored) e Millstone 1 (IREP) e Oconee 3 (RSSMAP) e Peach Bottom 2 and 3 (RSS) e Sequoyah 1 (RSSMAP) e Surry 1 and 2 (RSS) e The distribution of the mean values of core melt frequency estimated in the limited scope PRAs is illustrated in Figure 2.4-1. 1 The results for the mean value of core melt frequency range from less than 0.1 to about 3 core melt events per 1,000 reactor y per 1,000 reactor years and the distribution of the mean includes operation. Considerable analysis would be required to fully explain the differences in these results in tems of and 9 below the computed average. differences in plant design, operations and maintenance procedu l possible range of values of the core melt frequen analysis methods, and data. !l It would not be unreasonable to in methods and data notwithstanding. expect an even wider range of values if the analyses were expanded include all reactors in operation and under construction, even if differences in scope and methods were isolated out and the evaluations were more complete. of 0.23 events per 1,000 years of reactor operation This result f s in spf te of the fact that a full spectrum of internal and i external events was included in the results for Seabrook St l The second category of PRAs includes the published full scope PRAs on Zion, Indian Point, Seabrook Station, and the PRA carried I j Big Rock Point.however, it did include an analysis of internal fires which w l dominant contributors to core melt frequency, scope between the Of g Rock Pofnt PRA and the full scope PRAs are l l judged to be significant with respect to core melt frequency, it was I found to be appropriate to categorf te f t with the full scope PRAs for t A comparison of the results is purpose of this limited comparison. presented in Table 2.41 an i Hence, with respect to the l 1.6 core melt events per 1,000 reactor years." fuller" scope PRA j i i 2.i*2 l p::- q i

~ Sin:e the scope be neither extremely high nor lom but near the average.of th j in this comparison, if the scope variations were isolated out, the i Seabrook Station results would still compare favorably with the average, A final comparative perspective on the SSPSA results in relation to other nuclear plants for which

• fuller
• scope PRAs have been carried out is provided by examining the contributors to risk and core melt freqsencySuch identified in the respective studies.As can be seen in this table, there are som In Table 2.4-2.

arities such as the frequent appearance of fires and se contributor and two cases including Seabrook Station with interfacing Despite systems LOCA as the dominant contributor to early fatality risk. the similarities, there are enough differences to prove since the ordering of risk and core melt contributors is dependent on plant and site specific factors. COMPARATIVE RISKS OF ALTERNATIVE ENERGY SUPPLIES 2.4.2 In the previous section, a limited perspective was provided for cocparing the frequency of a serious accident estimated for Seabrook Station with Whereas 1 estimates that have been made for other nuclear power plants. this provides one element of a relative safety perspective for nuclear plants, it does not provide a basis for judging the acceptability orSuch unacceptability of the results.For example, in the absence of a decision to decision-making framework. use energy, the need to consider the risks of energy production alongSim with the costs and benefits does not exist. a decision to select among alternative energy sources, the need to consider the comparative risks, costs, and benefits of the alternatives For this reason, it would not make sense to judge the does not exist. acceptability or unacceptability of the risk levels calculated for Seabrook Station without specifying the context in terms of a decision and the risks, costs, and benefits of each of the decision alternatives. For example, it would not make sense to decide whether to build a nuclear The alterna-power plant based on a consideration of its risks alone.tiv risk in relation to costs and benefits. The issue of comparative risks of alternative energy sources was taken up by the National Research Council in their final report of the Comittee on Nuclear and Alternative Energy Systems (CONAES) (Reference 2.4 3). ) Their assessment was that: I In tems of pubite risks from routine operation of electric power plants (including fuel production and delivery), l Tulic ' sec;, rut in:1utt a alysis cf enternal c.c ts. l 1 1

risk, with oil-fired and r;. clear 1s-era'f:n censideraMy safer With respect to accidents, the and natural gas the safest. generation of fossil fuels presents a very Icw risk ofThe projected n catastrophic accidents. associated with nuclear accidents is probably less tha from routine operation of the nuclear fuel cycle (including mining, transportation, and waste disposal) but the large range of uncertainty that still attaches to nuclear safety calcula-tions makes it difficult to provide a confident assessment ofThe spread the probability of catastrophic reactor accidents.of uncertaint and nuclear power is such that ranges of possible risk overlap High level nuclear waste management does not present catastrophic risk potential, but its long-tern low-level threat somewhat. demands more sophisticated and comprehensive study and planning than it so far has received, particularly in view of the acute public sensitivity to this issue. Comparative risk data from the above study are rep above) are regarded by the CONAES as probably less than the rf sks forI routine operation of the nuclear fuel cycle. comprehensive PRA evaluations of high consequence accidents hav been carried out for nuclear power plants. of hydroelectric dams and liqueff ed natural gas facilities have indicated potential for catastrophic accf dents; indeed, such accidents have alrea occurred and no other energy source has been subjected to the analysis of hypothetical, low probability accidents to any extent approaching the case of nuclear power plants. risks in Table 2.4-3, the CONAES found energy sources in general to pose of accidents such as automobile accidents, falls, drown found coal and oil to pose greater risk than natural gas and uranium Hence, the oxide, the source of fuel for If ght water fission reactors. CONAES results would indicate an increase in accidental de Seabrook Station were replaced by a fossf1 plant not the uncertaintfes in the calculation of nuclear accident risks. gas. l I 2.4.3 SAFETY GOAL.5 Numerous proposals have been made in recent ye l l Recently, the Nuclear risk levels calculated for nuclear pcwer plants. l Regulatory Comission has published preliminary safety goals and preliminary numerical design objectives for a 2-year trial per l i (Reference 2.4 4). deterministic rules, regulations, and criteria used by the NRC to license j and regulate nuclear power plants, but instead, will serve as interim j goals to augment the regulatory decision making process. I I 2.8 4 P :... '. : : l

Tne NR"'s safety goals consist of qualitative and qsantitative eleNnts. The two qualitative goals are: 4 Individual members of the public should be provided a level of 1. protection from the consequences of nuclear power plant operation such that individuals bear no significant additional risk te life and l health. 4 Societal risks to life and health from nuclear power plant operation 2. should be comparable to or less than the risks of generating electricity by viable competing technologies (particularly coal-fired plants) and should not be a significant addition to other societal risks. The NRC also provides a quantitative interpretation of the numerical risk levels that might be regarded as "significant" in proposing the following quantitative design objectives. 1. Individual and Societal Mortality Risks Individual Risks. The risk to an average individual in the a. vicinity (within 1 mile of the site boundary) of a nuclear power plant of prompt fatalities that might result from reactor accidents should not exceed one-tenth of one percent (0.11) of the sum of prompt fatality risks resulting from other accidents to which members of the U.S. population are generally exposed. Societal Risks. The risk to the population in the area near b. (within a 50-mile radius of) a nuclear power plant of cancer fatalities that might result from nuclear power plant operation should not exceed one-tenth of one percent (0.1t) of the sum of l cancer fatality risks resulting from all other causes. l The benefit of an incremental reduction of Benefit-Cost Guideline. 2. societal mortality risks should be compared with the associated costs on the basis of $1,000 per man-rem averted. l Plant Performance Design Objective. The likelihood of a nuclear f 3. reactor accident that results in a large scale core melt should noma 11y be less than 10,000 per year of the reactor operations. The NRC further states that item 2, the benefit-cost guideline should be applied only if one of the other quantitative objectives is not met. l l A comparison of the results of this study against the quantitative design l objectives for individual and societal risks is presented in As can be seen in this table, the risk of early fatalities Table 2.4-4. to the 4.435 individuals within 1 mile of the plant was found to be a factor of between 5 and 6 below the individual risk goal, and the 4.2 rillion people within 50 miles of the plant were found to have an individual risk of latent cancer fatality more than two orders of hote that the values calculated mar.itude below the societal risk goal, O e g

---

rww-s

---e.. -w-ee- - - = - --2---.--

De mean vahes are in Table 2.4-4 for Seabrock Station are mean vahes. 1 in used to obtain the best match ith the state :e9t of the risk influenced (increased) of the uncertainties that were quantified. column S. With regard to the core melt frequency design objective, the results i this study of 1.9 x 10-4 events per reactor year (median) :-e with n aTh factor of two of the design objective. core melt frequency comparison has been suggested by the NR results include contributions from a full spectrum trial use of the safety goals. of external events to the design objective, the SSPSA results are vie If the contributions as comparing favorably with the design objective. from seismic events, fires, and other external events were not include the SSPSA resulgs for the median core melt frequency would ha In light of the underlying there is not a significant difference between 1.3 x 10-4 events per reactor year. about 1.3 x 10-uncertainties and 1.0 x 10 g events per reactor year. RETROSPECTIVE REVIEW OF PHASE I RESULTS 2.4.4 ment was made as a means of optimizing project re the These early results were risk controlling factors of Seabrook Station. obtained about 15% into the project in terms of time and manpower resources and, therefore, the uncertainties in these estimates were mu i Further perspective from which to view the current Phase !! results can be obtained by viewing these results in greater than in Phase II. 119ht of those published after Phase I (Reference 2.4-2). l risk curves were developed for early and latent canc The core melt all results included a quantification of uncertainty. frequency was based on preliminary event trees and systems analy However, allowances were made in the j relatively incomplete plant model. Phase I core melt frequency calculation for contributors missing from th The containment model used in Phase I was based on the Indian Point PRA and the site model simply consisted of subjective risk model. estimates of accident consequences. Phase 11 estimates of core melt frequency is presented in Table 2.4-As indicated in this table, two dif ferent approaches had been used in Phase 1 to develop uncertainty distributions, denoted The upper Method 2. to upper and lower bound estimates of core melt frequency. bound was taken as the point estimate from the P power recovery and manual scram following an anticipated transient The without scram (ATWS); f.e., 4.4 x 10-3 events per reactor year. lower bound was a subjective estimate by the study team on the lower bound of core megt freq;ency believed attainable for a modern reactor (1 x 10' per reactor year). In Metnod 2, a dif ferent approach i i i l j

.a-i

was followed consisting of propagating uncertainties through the Fhase I risk model with an allowance for risk contributcrs such as ext events left out of the Phase I model., As illustrated in Figure 2.4-2, the Phase 11 results are represented by a narrower distribution than Method I; Phase II results are situated near This result is logical since the Phase 11 results repr state of knowledge and therefore the distribution should be narrower. The fact that the Phase II distribution is near the center indicates, retrospectively, that the degree of conservatism embodied in the estimate ~ of the upper bound was balanced by the degree of optimism in the estima By contrast, Method 2 of Phase I appears to have of the lower bound. understated the effects of uncertainty on the low side This can be explained by conser-shifted up scale relative to Phase II. vatisms in the Phase I model that were not factored into the quantifica-tion, the chief such conservatism being the Phase I treat to core uncovery in 30 minutes compared with nearly 4 hours calculated i seal LOCA. This led to an underestimate of the effects of operator Therefore, the Method I Phase II. recovery in Phase I relative to Phase II. approach to quantifying uncertainty seems to have Phase I. With regard to the qualitative insights developed in Phase I and the list of dominant risk contributors that were identified, a comparison with The similari-Phase II reveals many similarities and a few differences. ties include the appearance of station blackout sce overpressurization failure of the containment given core melt, and the prominence of sequences involving failure of the primary compo cooling water system. of the sensitivity of assumptions regarding the behavio The most significant difference between the two sets of results was the inability in Phase I to identify the risk significance of interfacing system LOCA. The importance of interfacing systems LOCA with respect to early fatality The Phase I model risk in Phase I was masked by a simplified site model. overestimated the consequences of delayed overpressurization in relation Since the frequency of to those resulting from containment bypass. estimated to be much greater than that of the interfa l the risk of the latter category of scenarios was masked by that of the former. In summary, Phase I provided a good perspective on core melt frequency, a first-cut at defining candidate risk contributors, and an enhanced On the other hand, allocation of resources for the conduct of Phase II.in terr.s rist levels a*,d risi ccatribute s f r. Ftase I, a first-cut, prelirinary fcr a full s:c e FT* be regardEC as a substitutt risk assess *f-t c!W 1 ~

\\ Such as t9!! c ;letej in Enase 11 57d re;;rted.e-et.. Farther cre, tre f act that the differe-:es bet.een Phase I and Phase II have been explained enhances our confidence in the risk assess :ent now that it has been completed. 2.

4.5 REFERENCES

2.4-1. U.S. Nuclear Regulatory Cornission, " Proposed Comission Policy Statement on Severe Accidents and Related Views on Nuclear Reactor Regulation - Part II Proposed Policy on Safety Goals," Federal Register, Vol. 48, No. 72, April 13,1983. 2.4-2. Pickard, Lowe and Garrick, Inc., "SSPSA - Phase I Preliminary Risk Model Development," PLG-0242, August 1982. 2.4-3. National Acadecry of Sciences, Energy in Transition 1985-2010, W. H. Freeman and Company, San Francisco,1980, p. 429, 2.4-4. USNRC, " Safety Goal Development Program," Federal Register, Vol. 48 No. 50, March 14,1983. m 2.a-i

.: :::I,.

4 COMPARISON OF MEAN CORE MELT TABLE 2.4-1. FREQUENCY RESULTS OF FOUR ENHANCED SCOPE PPAs Mean Core Melt Frequency (events per Plant / Unit 1,000 reactor years) 1.6 Big Rock Point 47 Indian Point 2 .19 Indian Point 3 .052 Zion 1 and 2 0.23 Seabrook Station 1 and 2 I e

TABLE 2.4-2. COMPARISON OF RISK AND CORE MELT FREQUENCY CONTRIBUTORS FROM FOUR ENHANCED SCOPE PRAs Major Contributors - Accident Sequences Grouped 1 by Initiating Event Risk Plant Core Melt Frequency latent Cancer Early Fatalities Fatalities Zion 1 and 2 Small LOCA Seismic Events. Seismic Events Interfacing Systems LOCA I Indian Point 2 Fires, Seismic Events Seismic Events, Seismic Interfacing Events, Fires Systems LOCA l. Indian Point 3 Small LOCA, Fires Interfacing Fires Systems LOCA Fires Fires Big Rock Point Seabrook Station 1 and 2 Loss of Offsite Interfacing Loss of Power Seismic Systems LOCA Offsite Power Seismic Events, Fires Events, Fires

• No potential for early health effects identified.

COMPARISON OF ACCIDENTAL DEATH RATES FROM ALTE ENERGY SOURCES DURING ROUTINE OPERATION (PER GI TABLE 2.4-3. (Reference 2.4-3) Energy Source Quantity (pergfgaa year)* 6 Tons 3 x 10 Coal 4.0 Deep 2.6 Surface 6 Barrels 0.4 12 x 10 011 9 ft3 0.2 Natural Gas 67 x 10 0.2 Uranium Oxide 150 Tons 500.** Accidental Rate Due to All Causes i

• Includes extraction, processing, transport, and power station; excludes catastrophic accidents.

1,000,000 of the general

• • The rate given is average per population per year which is the number of people served by a gigawatt plant.

l --9 m v. -c-

i COMPARISON OF INDIVIDUAL AND SOCIETAL RISKS CALCULATED TABLE 2.4-4. FOR SEABROOK STATION AGAINST NRC INTERIM SAFETY GOALS Risk Nonnuclear Calculated NRC Risk b Fatality Risk in SSPSA* Goal Basis Per ent of Component of Population Population (frequency of. (frequency (percent of Wonnuclear Risk Goal Segment facility per of fatality nonnuclear Risk risk) person per year) per person per year) Early Fatality 1 mile radius 4,435 S.0 x 10-4 8.6 x 10-8 0.1% of Non-0.017% nuclear , Accidental Fatality Risk Latent Cancer 50 mile radius 4,200,000 2.0 x 10-3 6.3 x 10-9 0.1% of Non-0.0003% nuclear Cancer Fatality Risk Fatali ty

• Based on mean values of uncertainty distributions.

,..........,,coi ^ a lOf

COMPARISON PHASL I AND PHASE II TABLE 2.4-5. RESULTS OF SEABROOK STATION CORE MELT FREQUENCY PARAMETERS Phase I Phase II Parameter Method 1 Method 2 4.4-3 3.9-3 5.1-4 Upper Bound 1.8-3 9.3-4 2.3-4 Mean Median 6.7-5 5.1-4 1.9-4 Lower Bound 1.0-6 1.4-4 7.2-5 t Exponer tial notation is indicated inabbreyteted form P NOTE: Y i Y l l 1 j i s + e t 0?!? 121753

SSPSA RESULTS FOR ALL RISK CONTRIBUTORS V a AVERAGE OF 13 PRAs FOR INTERNAL RISK CONTRIBUTORS ONLY f 3 a E l V 4 z I l l I l l l 1.0 2.0 10 CORE MELT FREQUENCY (EVENTS PER 1,000 YEARS OF REACTOR OPERATIONI FIGURE 2.4-1. DISTRIBUTION OF MEAN CORE MELT FREQUENCY (Values Estimated in 13 Limited Scope PRAs) i I

UPPE R

• LOWER
• ROUND ROUND MEDIAN MEAN 9

i i i l l METHOD 1 1 i PHASEI i I METHOD I y ~. s i 4 i i l PHASE 11 i I I i I I 10-6 10-5 10 10-3 10-2 CORE MELT FREOUENCY (EVENTS PER REACTOR YE AR) 1 LOWER AND UPPER ROUNDS CORRESPOND, RESPECTIVELY, WITH STH AND 95TH PERCENTILES OF THE UNCERTAINTY DISTRIRUTIONS. i FIGURE 2.4-2. COMPARISON OF CORE MELT FREQUENCY RESUl_TS OBTAINED IN PHASE I AND PHASE 11

Summary of Results Presented in the Seabrook PRA 1.0 Ouestions Answered by the PRA The three questions listed below provide a structure for the analytical work of the Seabrook PRA (SPRA) and a framework for organizing the numerical results. o What is the likelihood of core melt? o What is the likelihood of release of radioactive materials as a function of release magnitude? o What is the likelihood of damage to public health and property as a function of the level of damage? The answers to these questions developed in the SPRA are briefly described in Section 3.0 herein. 2.0 What was Considered in the Analysis 2.1 Initiating Events The SPRA included consideration and quantification of the 58 initiating events listed in Table 1. The table includes the code designator used for each l initiator in the study. 2.2 Plant Damage States The SPRA considered 39 plant damage states in the risk model. There PDSs are listed and defined in Table 2a. Their relationships to one another are illustrated in Table 2b. l l t

o 2.3 Release Categories The SPRA used 13 release categories to represent the spectrum of release These 13 release categories can be further categorized into three groups: states. one group for categories in which the containment structure remains intact and isolated, a second for release categories that involve a gradual, long term degradation of containment integrity, and a third which involves early containment failure or bypass. Definitions of the release categories in each of these three groups are presented in Table 3. Each release category is represented by a three-part code. The code designator consists of an S, which denotes applicability to Seabrook Station, a number 3.c inoicate the absence of containment failure mode or state, a bar (--) to indicate the absence of containment filtration by the containment building spray system, and a V to indicate the presence of an additional vaporization component of the source term for scenarios in which the molten core debris experiences sustained elevated temperatures. Certain symbol combinations corresponding to additional release categories which can be hypothesized are precluded by Seabrook Station design features. For example, there is no 55 because the containment building sprays must function with adequate heat removal to keep the containment structure intact. 3.0 Results 3.1 Classes of Results Presented The SPRA presents results for the 7 classes of damage indices listed below, with consideration of uncertainty for each. o Core melt frequency. Early fatalities; those occurring within a short time after exposure, l o injuries (radiation illness not leading to fatality). o l l

Thyroid cancer cases (total occurring over a 30-year period); those o resulting mainly from iodine ingestion and curable medically in about 90% of the cases. o Latent cancer fatalities; those from cancers other than thyroid cancers, occurring over a 30-year period. Total population dose, or man-rem (whole body gamma dose). o Total public property damage and evacuation costs in dollars. o The results are also presented in terms of single unit and station (2-unit) risk. 3.2 Numerical Results The results of the SPRA are presented h a probability of frequency format for each of the seven classes of damage indices. 3.2.1 Single Unit Risk The probability of frequency of core damage (melt) is presented in Figure la (probability density) and Figure Ib (cumulative probability). The results for the remaining six damage indices are presented as risk curves including uncertainty in Figure 2a through 2f. 3.2.2 Two-Unit Station Risk The evaluation of risk of the Seabrook Station with two operating reactors required that the following factors be taken into account. The increase in the likelihood of accidents because of the presence of two o reactor units and the potential for interactions between the units. The possibility of simultaneous accidents in response to initiating events o that affect both units. The possibility of comon cause failures between units in response to an l o initiating event that affects both units. The results for core melt frequency are presented in Figures 3a I

(probability density) and 3b (cumulative probability). The difference in the mean frequencies of single unit and station (2-unit) core melt events per year is less than a factor of two because of the elimination of double counting of those initiating events that cause a core melt in both units. In other words, Figures 3a and 3b illustrate the probability of frequency of one or more (two) core melts per station year. The contributions to the total event frequency from single and double unit events are illustrated below in terms of mean values. Event Mean Frequency (Events Per Station Year) Core Melt Involving 4.0 E-4 One Unit (2.0 E-4) x 2 Core Melt Involving 0.3 E-4 Both Units Total 4.3 E-4 The effect of two-unit operation on the risk curves is illustrated in Figures da and 4b for early fatalities and latent cancer fatalities respectively. These figures contain plots of the mean value risk curves for two-unit (station) l operations and the corresponding curves for single-unit operations. The two-unit station risk curves include contributions from single unit accidents as well as from I double unit accidents. Figure 4a is indicative of the damage indices for early health effects, whereas Figure 4b is indicative of latent effects. 3.3 Principal Contributors to Core Melt and Risk The matrix formation used in the SPRA allows ready identification of major contributors to the numerical risk results by a process of decomposition of the risk matrices. The procedure for systematically determining the principal risk l

contributors begins with the risk curves and works progressively backward through the risk model to determine the most important paths insofar as numerical contributions are concerned. 3.3.1 Dominant Release Categories The dominant contributions are identified by determining the relative importances of the release categories at various damage levels, in terms of the relative frequency of exceedance values. Tables 4a and 4b present the relevant matrices for early fatalf ties and latent cancer fatalities, respectively, and Figures Sa and 5b are plots of the information in the matrices. Summary tabulations of the dominant release category contributions to risk for these risk indices are shown in Tables 5a and 56, respectively. The results of Figure 5a and Table Sa clearly show that release category 5BV dominates the risk of early fatalities across the full range of damage level presented and S2V makes a small contribution. Release category 51 contributes to an extended tail of the risk curve at frequencies below 10-9 as seen in Table Sa. None of the remaining 10 release categories were found to make a significant contributions to early fatality risk. In Figure 5b and Table 5b, it seems that release categories 55V and 52V also contribute significantly to the risk of latent cancer fatality, especially in the low frequency range of the risk curve. In the low consequence relatively high 4 frequency range of the risk curve, release categories $5V and 53 make significant contributions. Category S4V makes a small contribution across the full range of the latent cancer fatality risk curve, and category SS has only a small contribution to the risk with no potential for large numbers of latent cancer fatalities.

The importance of a release category, as measured by its influence on the risk curves, depends not only on its frequency but on its potential for producing various levels of damage. The singular effect of each release category's potential for producing damage is measured by the elements of the S matrix whose results have been plotted in Figure 6a for early fatalities and Figure 6b for latent cancer fatalities. In figure 6a, it is seen that over much of the damage level range, category 51 has the highest potential to produce early health effects given its postulated occurrence, followed by categuries $6V and S1 with category 52V having a much lower potential to produce damage. However, as shown in the previous section, categories S1 and 5I have an extremely low frequency of occurrence relative to categories 56V and 52V; thus, they make extremely small contributions to risk. Hence, upon completing the first step of the risk unraveling process, it has been determined that of the 13 release categories utilized in the risk model to group the billions of accident sequences that were analyzed, only 4 release categories make significant contributions to risk. These are categories 33V, 37V, 53V, and S3. Of these, only categories S6V and S2V make significant contributions to the risk of early fatalities. The results of an uncertainity analysis performed in the evaluation show that it is necessary to add only two more categories to the above (namely, categories SS and 53V) to obtain all categories making significant contributions to core melt frequency. Hence, 6 out of 13 release categories were l found to be significant with respect to their contribution to risk or to frequency of occurrence according to the following key: l l l i 1

i Significance Relative to: Release Category Risk Occurrence Early Latent Frequency Effects Effects $6V Major Major Minor SIV Minor Major Major SlV None Major Major 53 None Major Major 54V None Minor Major 55 None None Major 3.3.2 Dominant Plant Damage States Identification of the dominant plant damage states is accomplished by using the matrix tabulations of release category frequency of occurrence versus plant damage state. Tables 7a and 7b show these matrices for early fatalities and latent i. cancer fatalities. By examining the columns of these matrices corresponding to the risk dominant release categories identified in Section 3.2.1, the dominant plant damage states are identified. (Not yet Complete) 3.4 Compardtive Results The purpose of this section is to provide a perspective for viewing the absolute value of the risk levels calculated for Seabrook through the use of limited comparisons. Four types of comparisons are made, as follows: to risk levels calculated in PRAs of other nuclear power plants 0 to risk from sources of energy other than nuclear power o l \\ I

o to sources of risk other than energy production o to the provisional safety goals currently under evaluation by the NRC Each of these comparisons is described in modest detail below. 3.4.1 Comparison to the Risk frmn Other Nuclear Plants PRAs have been performed on about 20% of the nuclear power plants in n operation or under construction in the U.S. There are wide variations in the scope of these PRAs, which includes variations in the extent of the analysis, the' type of plant, the site, as well as many ottar types of differences. It is nevertheless possible to divide these PRAs into two general classes: limited scope and full scope. Limited scope is defined as not including consideration of external events and limited to an evaluation of the core melt frequency. There are 13 of these. Full scope is defined as generally irxluding consideration of external events and evaluating health risk. There are four of these. The limited scope PRAs have mean values of core melt frequency ranging from 0.1 to about 3 core melt events per year, with an average of about O'.5 events per 1000 reactor years of operation. The Seabrook result is 0.23 events per 1000 years of reactor operations, near the lower end of the range. The full scope PRAs have mean values of core melt frequency ranging from less than 0.1 to about 1.6 events per 1000 reactor years. The Seabrook result again compares favorably. An examination of the contributions to health risks and core i melt frequency in the fuller scope PRAs reveals that there are general similarities l in the results, such as the frequent appearance of fires and seismic events as risk contributors, and small LOCAs as important core melt contributors. In addition, two cases (including Seabrook) have the interfacing systems LOCA as the dominant s contributor to early fatality risk.

3.4.2 Comparison to the Risk from Other Sources of Energy The description of comparative risks of alternative energy sources is based on the final report of the National Research Council Committee on Nuclear and Alternative Energy Systems (CONAES). In essence, the CONAES report states that for routine operation nuclear power poses smaller risks to the public than other alternative energy sources except for natural gas. For accidents, the large range of uncertainty that still attaches to nuclear risk calculations makes it difficult to provide a confident assessment of catastrophic reactor accidents, even though the projected mean number of fatalities is probably less than the risk from routine operation of the nuclear fuel cycle. 3.4.3 Comparison to Sources of Risk Other than Energy Production This comparison, also based on the CONAES report, indicates that energy sources, in general, pose accidental death rates less than 1% of the rate due to all other causes of accidents such as automobile accidents, drownings, etc., and that this result is probably independent of the uncertainties in the calculation of nuclear accident risks. 3.4.4 Comparison to NRC's Provisional Safety Goals A comparison to NRC's provisional safety goals is presented in Table 6. As f can lie seen in this table, the risk of early fatalities to the 4,435 individuals within 1 mile of the plant was found to be a factor of between 5 and 6 below the individual risk goal, and the 4.2 million people within 50 miles of the plant were found to have an individual risk of latent cancer fatality more than two orders of magnitude below the societal risk goal. Note that the values calculated in Table 6 l for Seabrook Station are mean values. The mean values are used to obtain the best match with the statement of the risk goals in column 5. Unlike the median values, the nean values are significantly influenced (increased) by the uncertainties that were quantified. '

k'ith regard to the core melt frequency design objective, the results of' this study of 1.9 x 10~4 events per reactor year (median) are within a factor of two of the design objective. The use of median values for this core melt frequency comparison has been suggested by the NRC for the trial use of the safety goals. In view of the facts that the Seabrook results include contributions from a full spectrum of external events and that the NRC has indicated that care should be taken in the apportionment of external events to the design objective, the SSPSA results i are viewed as comparing favorably with the design objective. If the contributions from seismic events, fires, and other external events were not included, the SSPSA results for the median core melt frequency would have been about 1.3 x 10-4 events per reactor year. In light of the underlying uncertainties, there is not a significant difference between 1.3 x 10'4 and 1.0 x 10-4 events per reactor year. 4.0 Key Findings and Insights in general, insights from the PRA are presented in the Summary Report, beginning on page 17. A preliminary listing of insights believed to be important is provided below, in our words. o Risks are low, but the core melt probability is higher than the proposed safety goal, o A very large number of sequences contribute to the total core melt probability. The single most dominant sequence contrit/Jtes less than 15% of the total, and the top 27 sequences contribute just over half to the total.

f o The V-sequence accident totally dominates the risk of early fatalities. o External events are not important risk contributors (this result is not consistent with other recent PRA results: Zion, Indian Point, Oconee (NSAC/ utility assessment), Millstone Unit 3). o The most important initiating event in terms of core melt probability is loss of off-site power. 4 i I w l l l ,}}