ML20214B241

Containment Event Analysis for Postulated Severe Accidents: Sequoyah Power Station,Unit 1.Draft Report for Comment

ML20214B241
Person / Time
Site:	Sequoyah
Issue date:	02/28/1987
From:	Behr V, Benjamin A, Kunsman D, Lewis S, Murfin W SANDIA NATIONAL LABORATORIES
To:	NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
References
CON-FIN-A-1322 NUREG-CR-4700, NUREG-CR-4700-V2-DRF, SAND86-1135, NUDOCS 8705200197
Download: ML20214B241 (208)

Text

. . _ . - _ . _ _ _ . . _ _ _ _ . - _ _ _ . . _ _ . - - _ - - _ _ _ . . _ _ - _ _ _

,_4 .

NUREG/CR-4700 DRAFT REPORT FOR COMMENT SAND 86-1135 Volume 2 Printed February 1987 Containment Event Analysis for Postulated Sovere Accidents: Sequoyah Power Station, Unit 1 l V. L. Behr, A. S. Benjamin, D. M. Kunsman,

> S. R. Lewis, W. B. Murfin Prepared by Sandia National Laboratones Albuquerque, New Mexico 87185 and Livermore, Cahfornia 94550 for the United States Department of Energy under Contract DE-AC04-76DP00789 s

i Prepared for U. S. NUCLEAR REGULATORY COMMISSION

< x.~ m, 8705200197 870228 PDR ADOCK 05000327 OS PDR

1 l

s O.

l l

NOTICE This report was prepared as an account of work sponsored by an agency of the United States Government. Neither the United States Government nor any agency thereof, or any of their em.

ployees, makes any warranty, expressed or implied, or assumes any legal liability or responsibility for any third party's use, or the results of such use, of any information, apparatus product or process disclosed in this report, or represents that its use by such third party would not infringe privately owned rights.

Available from Superintendent of Documents U.S. Government Printing Office Post Office Box 37082 Washington, D C. 20013-7082 and National Technical Information Service Springfield, VA 22161 l

l

\

I l

l l

- = . _ .

i DRAFT REPORT FOR COMMENT NUREG/CR-4700 SAND 86-ll35 Volume 2 CONTAINMENT EVENT ANALYSIS FOR POSTULATED SEVERE ACCIDENTS: SEQUOYAH POWER STATION, UNIT 1 .

by l V. L. Behr

! A. S. Benjamin, D. M. Kunsman S. R. Lewis *

, and W. B. Murfin**

Printed: February 1987 Sandia National daboratories Albuquerque, New Mexico 87185 Operated by Sandia Corporation for the U.S. Department of Energy Prepared for Division of Reactor System Safety Office of Nuclear Regulatory Research U. S. Nuclear Regulatory Commission Washington, D.C. 20555 i

Under Memorandum of Understanding DOE 40-550-75 NRC FIN No. A-1322 l

• Safety and Reliability Optimization Services. Inc.,

Knoxville, TN

• • Technadyne Engineering Consultants, Inc., Albuquerque, NM l

i

l

/

. PREFACE Because of the time constraints imposed on this work to meet the Nuclear Regulatory Commission's schedule for publication of NUREG-llSO, this draft report has not yet received the full level of peer and management review customarily accorded to reports issued by Sandia National Laboratories. The reviews will be completed and corrections made, if necessary, prior to I

final publication.

1 J

4

, /*

l.

1

/I ..

i RE0 VEST FOR COMMENT This report, NUREG/CR-4700, " Containment Event Analysis for Postulated Severe Accidents," was prepared for the U.S. Nuclear Reoulatory Commission by the Sandia National Laboratories and its subcontractors.

The methods and results set forth in this report are being used by the NRC to support the development of the Reactor Risk Reference Document (NUREG-1150) and will be used in areas of broad public interest such as probabilistic risk analyses, emergency response planning, siting, NPC safety goal applications, and cost / risk / benefit analyses--indeed, wherever risks to public health need to be considered in regulatory applications. Thus, it is considered imperative that an opportunity for public comment on the results as presented in the report be provided.

Comments should be sent to the U. S. Nuclear Regulatory Commission, Washington, D.C. 20555, Attention: Joseph Murphy, Division of Reactor System Safety. These comments will be most useful to the staff if they are received by June 1,1987.

Of particular interest to uskis the receipt of comments on the methodology, and results, related to uncertainty analysis. One criticism voiced with respect to the Reactor Safety Study was its lack of an uncertainty analysis. We have included an uncertainty analysis, but we are sure that its nature,will be the subject of lively debate.

We welcome this, and solicit constructive advice and criticism.

The NRC hereby expresse'sitsgreatappreciakiontoallparticipantsin this study for their considerable efforts, as well as to all who will take the time and effort to provide it with comments on this report.

/

T D. F. Ross, Deputy Director Office of F t car Regulatcry Research i

i I

3 y

3 iii

, E s

ABSTRACT A study has been performed as part of the Severe Accident Risk Reduction Program (SARRP) to investigate the response of a particular pressurized water reactor with an ice-condenser containment (Sequoyah Unit 1) to postulated severe accidents.

A detailed containment event tree for the Sequoyah plant has been devised to describe the various possible accident pathways that can lead to radioactive releases from containment. Data and analyses from a large number of NRC and industry-sponsored programs have been reviewed and used as a basis for quantifying the event tree, i.e., determining the likelihood of each pathway for a variety of accident sequence initiators. A generalized containment event tree code, called EVNTRE, has been developed to facilitate the quantification. The uncertainty in the results has been examined by performing the quantification three times, using a different set of input each time to represent the variation of opinion in the reactor safety community. In the so-called " central" estimate, the likelihood of early containment failure (occurring before or at the time of reactor vessel breach) was found to be high for station blackout sequences but very low for other accident sequence initiators. Unavailability of igniters and air return fans was the principal reason for the high failure probability for station blackouts. The analysis also showed that melting or bypass of the ice before or within a short time after vessel breach can be expected to occur with moderate to high likelihood during station blackouts and during sequences initiated by very small LOCAs with failure of emergency core cooling in the recirculation phase after success in the injection phase. This work supports NRC's assessment of severe accident risks to be published in NUREG-ll50.

t iv

TABLE OF CONTENTS l

Paae

1. INTRODUCTION . . . . . . . . . . . . . . . . . . . . . . . 1

2. METHOD . . . . . . . . . . . . . . . . . . . . . . . . . . 4 2.1 General Description and Observations . . . . . . . . 4 2.2 Specific Example - Sequoyah S HF 3 . . . ... . . . . 9 2.3 Treatment of Verbal Descriptors . . . . . . . . . . . 10 2.4 Summary of Plant Features and Accident Sequences . . . . . . . . . . . . . . . . . . . . . . 10 2.4.1 Plant Features Important to Containment Response . . . . . . . . . . . . . . . . . . . 10 2.4.2 Description of Accident Sequence . . . . . . . 13

3. RESULTS . . . . . . . . . . . . . . . . . . . . . . . . . 20 3.1 Sequence Results for Sequoyah . . . . . . . . . . . . 20 3.1.1 Sequence S 2H2 . . . . . . . . . . . . . . . . 21 3.1.2 Sequence S 2H 2F . . . . . . . . . . . . . . . . 22 3.1.3 Sequence S 3H2 . . . . . . . . . . . . . . . . 23 3.1.4 Sequence S 3H 2P . . . . . . . . . . . . . . . . 24 3.1.5 Sequence S 3D1 . . . . . . . . . . . . . . . . 24 3.1.6 Sequence T I1i L D F . . . . . . . . . . . . . . . 24 3.1.7 Results for Plant-Damage States . . . . . . . 26 3.2 Sensitivities . . . . . . . . . . . . . . . . . . . . 30

4.

SUMMARY

AND CONCLUSIONS . . . . . . . . . . . . . . . . . 31 REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . 34 TABLES . . . . . . . . . . . . . . . . . . . . . . . . . . 39 APPENDICES A. Sample Containment Event Tree Problem . . . . . . . . A-1 B. Containment Event Tree for Sequoyah . . . . . . . . . B-1 C. Computer Input for Sequoyah Containment Event Tree . . C-1 D. Representative Computer Output'for Sequoyah Containment Event Tree. . . . . . . . . . . . . . . . D-1

[

y u . . . .

1 l

LIST OF TABLES i Page 2.1 Use of Information Sources in Addressing the Issues in the Containment Event Tree . . . . . . . . . . . . . 39 2.2 Event Descriptions for PWR Ice Condenser Containments . 40 2.3 Event Descriptions and Likelihoods for Sequoyah S 3 HF . . 42 2.4 Containment Loadings for Sequoyah S3 HF . . . . . . . . . 53 2.5 Sequoyah Containment Capacity Estimates . . . . . . . . 55 2.6 Basis for Pressure Estimates for Concurrent Direct Heating / Steam Spike . . . . . . . . . . . . . . . . . . 56 2.7 Alternative Assignment of Values to Verbal Descriptors . 57 2.8 Description of Plant-Damage States . . . . . . . . . . . 58 3.1 Results for Sequoyah S H22 - - - - - - - - - - - - - - - 59 3.2 Results for Sequoyah S 22 H F . . . . . . . . . . . . . . . 60 3.3 Results for Sequoyah S H32 . . . . . . . . . . . . . . . 61 3.4 Results for Sequoyah S 32 H P . . . . . . . . . . . . . . . 62 3.5 Results for Sequoyah S D13 . . . . . . . . . . . . . . . 63 3.6 Results for Sequoyah T ttt L D F . . . . . . . . . . . . . . 64 3.7 Likelihood of Containment Failure Modes by Plant Damage State (Optimistic Walkthrough) . . . . . . . . . 65 3.8 Likelihood of Containment Failure Modes by Plant Damage State (Central Walkthrough Without Direct Heating and In-Vessel Steam Explosion) . . . . . . . . . . . . . 67 3.9 Likelihood of Containment Failure Modes by Plant Damage State (Central Walkthrough with Direct Heating and In-Vessel Steam Explosion) . . . . . . . . . . . . . . . 69 3.10 Likelihood of Containment Failure Modes by Plant Damage State (Pessimistic Walkthrough without Direct Heating and In-Vessel Steam Explosion) . . . . . . . . . . . . . 71 vi

4-LIST OF TABLES Pace i

3.11 Likelihood of Containment Failure Modes by Plant Damage

, State (Pessimistic Walkthrough with Direct Heating and In-Vessel Steam Explosion) . . . . . . . . . . . . . . . 73 1

3.12 Sensitivity of Sample Case Results to Alternative Assignments of Numerical Values . . . . . . . . . . . . 75 4

i r

1 i

I 1

l l

i i

i i

4

,1 f

t 1

i f

i vii

.,,... . , . - - , , - .. , , . - . - , , . , - . - , . - - - - . . - - . . - - -,-,., -,-, .-.-- - -.-... - ... .- ~---.- --., - ,-,

T

1. INTRODUCTION This report documents analyses performed under the Severe Accident Risk Reduction Program (SARRP) to investigate the response of containments at commercial nuclear power plants to i

severe accidents. These analyses focus on identifying the various_ pathways that could lead to the release of fission products beyond the containment boundaries and on estimating j

~

their frequencies of occurrence for four reference power plants.* The analyses are part of an overall effort aimed at developing a comprehensive and up-to-date characterization of nuclear plant risk (Ref. 1) which will serve as a principal

input to a forthcoming report by the U.S. Nuclear Regulatory Commission (NRC), NUREG-1150 (Ref. 2). It is anticipated that NUREG-1150 will provide a technical basis upon which to consider future decisions regarding changes to regulations and regulatory practices.

{

The analyses described in this report also provide a broader perspective on the significance of the fission product source terms estimated in NUREG-0956 (Ref. 3), in BMI-2104 I (Ref. 4), and in subsequent analyses performed for SARRP (Ref.

! 5), by examining the ways in which containment failures leading to source terms of different magnitudes can occur.

i' One might ask, "Why is there a need for such a perspective on the various accident pathways that are possible?" To answer this question, let us consider as an example the BMI-2104 analysis of the sequence S 2 D at Surry. Two calculations were made. In the first, a break of a specified size (2-inch diameter) was presumed to occur in the cold leg through the reactor coolant pump seals, and the primary system coolant was presumed to end up in the containment sump. The emergency 4 coolant injection system was assumed not to function. The i containment was assumed to fail from rapid pressurization due i to events occurring just after meltthrough of the reactor vessel, and the containment sprays were assumed to fail at that time. The release from containment was assumed to bypass the auxiliary building.

In the second calculation, a break of the same size was I

assumed to occur in the hot leg piping. Both the containment 4

• This is the second in a series of reports, each describing the analyses and results for one of the reference plants.

i The introductory material and description of the methodology (Section 2.1) are virtually identical to that in the report t for Surry. The reader who is familiar with that report may wish to skip these sections and begin with Section 2.2.

and the containment sprays were assumed to survive the events following vessel meltthrough, but the water flow to the reactor cavity was assumed not to prevent the core-concrete interaction from occurring. Containment remained intact until the calcula- l tion was terminated.

i While the events assumed in the BMI-2104 calculations are  !

plausible, one might ask whether they are the most likely series of events that could occur following sequence S 2D, or whether they produce the highest source terms that could occur within the realm of reasonable probability. Is it more likely, for example, that the size and location of the break would be different from what was assumed in BMI-2104; that the contain-ment sprays would fail prior to containment failure, because of debris in the containment sump pluggins the pump intakes; that containment might fail by some means other than early overpres-surization; or that the release pathway would be through the auxiliary building, where further reduction of the source term would take place?

These questi0ns raise the need for a systematic identifica-tion of the various pathways that the accident can take and assessment of the likelihood, or probability, of each. Such an analysis provides a basis for evaluating whether the source terms developed for a particular accident sequence cover the range of risk-significant source terms for that accident sequence.

The overall objectives of this study are (1) to identify accident pathways (i.e., combinations of accident sequences and containment events) that delineate source terms which may be important to risk, (2) to estimate the frequencies of those pathways and hence the frequencies of the source terms they attend. (3) to ascertain how well the BMI-2104 source terms cover the accident pathways that are important to risk, (4) to identify accident pathways for which additional source term calculations are needed, and (5) to provide important input for the SARRP assessment of risks (Ref. 1). These objectives have been implemented so far in assessments of four reference plants: Surry (PWR subatmospheric containment), Sequoyah (PWR ice condenser), Peach Bottom (BWR Mark I), and Grand Gulf (BWR Mark III). This volume describes the analyses for the Sequoyah plant. Each of the other plants is treated in a separate volume of this report.

With regard to the nature of the sequences considered, two aspects should be noted. In the current analysis, we have not explicitly treated accidents initiated by external events (e.g., earthquakes, fires, flooding, etc), since no analyses of the frequencies of these events or of the capabilities of plant systems and structures to withstand them have been performed.

In addition, we have not performed an in-depth evaluation of operator actions that might be taken to respond to core melt

. . . = . . . - . - . .- - _ . . . - = - _ - - -- ..- - _ _ _

b i #

accidents to mitigate their consequences. The most serious obstacle to the performance of such an analysis is that plants do not currently have procedures which provide instruction to the operators past the onset of melting"of the core: it is 4

beyond the scope of the present analysis to devise such proce-i dures and to evaluate their effects.

In Section 2, we will describe the method we used to 4

achieve the objectives stated just above. First we will provide a general description of our containment event trees and the procedure we use for quantifying the branches. We will then illustrate this procedure by applying it to a specific j sequence for Sequoyah. We will then discuss some of the j special considerations we made for other sequences. In Section f

3, we will present the results, in terms of frequencies of containment failure modes for the various accidents that were

. assessed to be potential contributors to risk for Sequoyah. In 2

Section 4 we will summarize the results and comment on those that are particularly significant.

We have provided in Appendix A a sample problem for the

containment event tree to facilitate the explanation of the

{ manner in which the input for the event. tree is structured.

2 Appendix B presents our evaluation of the events in the I containment event tree for Sequoyah. Appendix B also provides j the details of the example sequence treatment we discuss-in '

Section 2. Appendix C includes the actual computer input to i the containment event tree while Appendix D presents the

j. detailed computer output for three damage states.

]

The containment event trees developed by SARRP have been

! reviewed by a committee commissioned by NRC and consisting of' 4 representatives from the University of Wisconsin, Battelle 1 Columbus Laboratories, Oak Ridge National Laboratory, Brook-

. haven National Laboratory, and Pickard, Lowe and Garrick, Inc.

i The review group investigated the reasonableness of the event

! trees (scrutability, completeness, consistency, and level of i detail) and the reasonableness of the quantification (trace-

! ability, documentation completeness, coverage of sources, and l consistency of approach). The review uncovered a few errors of 3

logic and omission which have subsequently been corrected.

Their-detailed comments are provided in Reference 39.

1 i

1,

! l

\

. i

! l l

i l

, , - ~ , - - - , . , - - - - + - - - . . , - - , - ---.-,-..,-,,---,..n~_.,n_., -

.,-....,--,-,.,.,.,.n.,,,m __-.en,vn,---..,,n,----,-

I

2. METHOD 2.1 General Description and Observations l In traditional probabilistic risk assessments (PRAs), the l

accident pathways that contribute to risk are described by two types ofievent trees. " System event trees" are used to define

the spectrum of accident sequences (i.e., the. combinations of accident initiators and subsequent system failures) that can

, lead to core melting. " Containment event trees" are used to j define the containment failure modes which lead to fission product releases beyond the containment boundary.

4 Our analyses are based upon the core-melt sequences that 4

were developed and whose frequencies were estimated through'the Accident Sequence Evaluation Program-(ASEP) (Ref. 6). For each of the reference plants, a PRA has been performed previously.

Sequoyah was one of the plants studied under the Reactor Safety Study Methodology Applications Program (RSSMAP, Ref. 7). The 4 sequence analysis from that study has been updated by ASEP to reflect the current plant configuration, accounting for changes

! that have been made to the plant since completion of the j original study, and to incorporate improved risk assessment

methods and data bases. A few planned, but not yet installed,

} plant modifications have been considered. These are modifica-tions that are mandated by rulemaking and that have a specific time period by which they must be implemented.

! In order to promote coordination between the sequence and

containment analyses, the sequence results were also cast. in

! the form of plant damage states. These plant damage states group sequences that would be expected to have similar effects on containment response. Although the same results could be obtained by evaluating containment response for the accident

sequences themselves, the plant damage states reflect explicit consideration of the sequence conditions that could produce unique effects on the likelihood of various containment failure modes (e.g., the status of containment safety features). They also provide a systematic mechanism for combining sequences that were~ distinguished for convenience in estimating their

, frequencies, but that need not be subjected to separate evalua--

] tion in the containment event tree. The definitions of the l plant damage states for Sequoyah are provided in Section 2.4.

i Our primary focus is upon the containment event tree for Sequoyah. We apply the event tree to particular sequences by selecting the appropriate branches within the event tree.

Our containment event trees are considerably expanded 1 beyond those considered previously in PRAs. We ask the j following types of questions:

i t

4 1

J'

1. Entry States. What.are the conditions in the reactor

! . coolant system and containment prior to melting of the core that could influence the accident progression?

2. Phenomenoloolcal events. What are the phenomena that could i affect the progression of severe accidents, at what points j during the accident timeline do they occur, and what are

their subsequent effects on the accident development?

3. Reactor coolant system failure modes. What is the size and i location of the reactor coolant system breach and the ,

pressure in the system at the time of breach?

4. Containment system survivability. Do the containment sprays, fans, and suppression systems survive the conditions occurring during severe accidents that exceed their design bases?

i

5. Containment failure modes. What are the loads that l challenge containment, does containment survive these j loads, what is the nature of the failure (approximate size i and location), and what is the subsequent pathway for fission product release to the environment?

The questions in the containment event tree are posed in j ways that require the answers to be expressed in terms of like-

lihoods. For a loss-of-coolant accident in a PWR, for example,

, we might ask how likely it is that the reactor coolant system j breach will be in the cold leg piping as' opposed to the hot leg

piping, or how likely it is that containment will fail due to a

! hydrogen burn following reactor vessel failure. Answers to j such questions require information about the reactor design, the phenomenology of reactor accidents, and the capabilities of i containment. For example, to answer the two likelihood ques-

! tions just posed, one would'need to know about the characteris-l tics of the cold leg versus. hot leg piping, the amount of hydrogen generated prior to vessel breach, the availability of ignition sources, the status of ice in the ice condenser, and

the failure pressure of the containment.

! Some of the issues addressed by the containment event trees

! are listed in Table :2.1. We point out that these-are not the events themselves, but rather some of the issues that must be addressed in order for the event trees to be quantified.

We have utilized a large number of sources to obtain the needed information, including the following:

(1) Containment Loads Working Group (CLWG), Refer-ences 8-10.

l .(2) Containment Performance Working Group (CPWG),

Reference 11.

I i

t . _ -

(3) Battelle calculations for Accident Source Term Project Office (BMI-2104) and for SARRP, References 4-5.

(4) Quantitative Uncertainty Estimation for the Source Term (QUEST), Reference 12.

(5) Industry Degraded Core (IDCOR) program, I Reference 13.

(6) Severe Accident Sequence Analysis (SASA) program, References 14-18.

(7) Severe Accident Uncertainty Analysis (SAUNA),

Reference 19.

(8) Accident Sequence Evaluation Program (ASEP),

Reference 6.

, (9) SARRP Phenomena Assessment Task Force (PATF),

Reference 20.

(10) Steam Explosion Review Group (SERG), Reference 21.

(11) High pressure ejection test series (HIPS),

Reference 22.

(12) Available probabilistic risk assessments (PRA),

References 3, 23-26.

(13) Final Safety Analysis Reports (PSAR), Reference 27.

(14) Architect-engineer (AE) and other estimates of containment failure pressure, Reference 28.

(15) Others, References 29-38.

The NRC provided general guidance with respect to the priorities to place upon the use of this variety of sources in evaluating the events in the containment event trees. To the extent possible, the information developed by the Containment Loads and Performance Working Groups and other analyses performed or sponsored by NRC were to be used. To provide a fuller treatment of issues raised in the containment event tree, other references, including analyses sponsored by the utility industry, were to be considered.

! We encountered several difficulties in attempting to utilize information from these various sources. One of the most sig-nificant was incomplete coverage. Table 2.1 illustrates the relationship between some of the issues addressed by the event

-w. . . _ - - - . . , ., - . , , , , . ,

trees and the information provided by the two containment working groups and a number of the other programs. It is clear from the table that the results from the NRC-sponsored working groups and review groups do not address all of the questions asked.

When a question was addressed by one of the studies, the information provided often required us to make extrapolations.

For example, the Containment Loads Working Group provided estimates of the size of steam spikes and direct heating for only the large-dry PWR reference plants, and then with preconditions appropriate for only one accident sequence. We had to extrapolate this information to other plants and other sequences. The same was true for the analyses of global hydrogen burns, diffusion flames, and containment temperatures achieved from core-concrete interactions. Similar statements apply to other studies.

Furthermore, the information provided to us often did not include a best-estimate characterization of a particular issue, but rather a range of possible values. In particular, the CLWG provided low, medium, and high estimates for containment

] loading, whereas the CPWG provided only high estimates for containment leakage. In the CLWG, consensus was generally reached more often on the low and high estimates than on the medium estimates.

When we quantified our containment event trees, therefore, we propagated three separate estimates -- optimistic, central, and pessimistic. Thus, we derived three sets of accident outcome probabilities for each sequence. For each of the three estimates, we formulated a set of branch point probabilities and parameter values that were different but each internally consistent. The set labeled " pessimistic" tends to provide higher probabilities for the pathways that lead to higher source terms and lower probabilities for the lower source term pathways. The ones labcled " central" and " optimistic" are analogously interpreted. In this report we do not propose weighting factors or distributions for these estimates, nor purport that one is better than another. We present the estimates as a reflection of the information that is available.*

We have attempted to define the three sets of values in such a way that the central estimate represents the median of the reactor safety community on any particular issue. That is,

• In the SARRP risk report (Ref. 1), we explore the question of reasonableness by developing and implementing a second procedure of uncertainty ev'aluation called " limited Latin hypercube." That procedure uses expert opinion to establish 4 weighting factors, which are used as a basis for sampling a number of uncertain issues.

~

t if a large number of experts were to be polled with regard to their views, a substantial fraction would respond that the l actual situation would be better than that indicated by the central estimate, and an approximately equal number would respond that it is worse. Of those who responded that the ,

situation is actually better than that represented by the '

central estimate, we would expect that a substantial number would view our optimistic estimate as an appropriate represen-tation of the issue; likewise, those who viewed the situation as being worse than the central estimate implies would agree that reality is or could be as unfavorable as that reflected by our pessimistic estimate. It should be pointed out that we were not able to apply these principles rigorously in assessing some questions because, for some issues, there is a very small community of experts. In some cases, it was necessary for us i to define the optimistic, central, and pessimistic estimates according to the views of one or two experts. For a few cases in which there was a very wide variety of opinions, indicating particularly large uncertainties, the issues were treated as sensitivities. This was specifically true for the assessments of direct heating and of an in-vessel steam explosion capable of causing containment failure (the a-mode). For these ques-tions, we performed the sensitivity analysis by defining opti-mistic, central, and pessimistic values twice, once as the

optimistic group would view them, and again as the pessimistic would. We then conducted two corresponding sets of optimistic /

central / pessimistic walkthroughs. Generally, the optimistic 1 views were that these phenomena were not physically possible, so that the sensitivity cases were effectively reduced to 4 occurrence or non-occurrence of the phenomena.

Finally, our containment event analysis is based on the

plant configurations that currently exist. We have not attempted to account for plant changes that are planned but not yet implemented, with the exception that changes mandated by rulemaking are included even if they have not yet been installed. This exception had a larger impact for the BWR plants than for the PWRs. For example, in our analysis for Peach Bottom, we account for the symptomatic emergency pro-cedures that include containment venting, alternate injection sources, and procedures for anticipated transients without scram, and we do credit the increased standby liquid control capacity, since it is mandated by the ATWS rulemaking.

It is worth noting that in this report we describe only the accident pathways leading to fission product releases, and do not calculate the source terms. The source terms for these

accidents are being documented in a separate report (Ref. 5).

The pathways and corresponding source terms are integrated i through the performance of consequence analyses to provide a characterization of risk: this process and the results, includ-

ing an assessment of the impact on risk of various potential l safety measures, are described in the SARRP risk rebaselining and risk reduction report (Ref. 1).

i l

-- -,iww r,,v+, , - , - - , - w- v7-g --- r,,--+ +-w->-w-e%-g ,, ww+--,-eze-m-g- ---r- -r,,, , % e- m

2.2 Specific Example: Sequoyah Sequence S 3 HF A containment event tree was developed specifically to address the potential responses of ice condenser containments to core-melt accidents. The 49 questions which comprise the containment event tree provide the means to assess the outcomes for all of the accident sequences that we analyzed. Because of the large number of questions, and the fact that many of them have more than two possible branching options, it was not practical for us to provide a graphical representation of the event tree. Each of the questions, its possible outcomes, and their probabilities are described in Appendix B.

Sequence S 3 HF involves an accident initiated by a small LOCA due to the failure of reactor-coolant pump seals (event S3), followed by successful operation of the emergency core cooling system in the injection mode (in this case, requiring the high-pressure injection systems), but failure of both core cooling and the containment spray system in the recirculation mode (events H and F, respectively). The application of the event tree for the assessment of Sequoyah sequence S3 HF is illustrated in Tables 2.3 through 2.6.

The containment event tree questions relevant to the assessment of sequence S 3 HF are delineated in Table 2.3. For each question, the specific conditions that affect the choice of branches (i.e., preconditions associated with the accident sequence itself and conditions established by the outcomes of prior questions in the event tree) are noted, as are the branching options themselves. Our estimates of the relative likelihood of the branches are also provided for the optimistic, central and pessimistic cases. As we noted above, many of the questions have more than one possible branching option: for these, the question posed requires an estimate of severity of damage, or choice of size or location of breaks, rather than a yes/no answer.

Note also that the likelihoods for many of the branches are indicated by verbal descriptors, rather than numerically. This distinguishes questions for which data provides a reasonably direct means to estimate probabilities from those for which it is possible to estimate probabilities only in a qualitative sense. Later, we assign numerical values to the qualitative descriptors to enable us to estimate the relative frequencies of the various containment failure modes. However, It should be recognized that the choice of numerical values for these questions is particularly subjective; the sensitivity of the results of this analysis is addressed in Gections 2.3 and 3.2.

For several of the questions addressed in Table 2.3, information is needed about the pressure loadings that the containment would be expected to experience and the capacity of the containment to withstand such loadings. The results of our

_9_

evaluations of containment loadings under the various accident conditions of interest are summarized in Table 2.4. In Table 2.5, we summarize containment capacities for both structural failure and pressure-induced leakage. Table 2.6 presents the information we used to estimate containment loading from direct heating and steam spikes following breach of the reactor vessel.

A more detailed discussion of our rationale for the assign-ment of values in Tables 2.3 through 2.6 is provided in Appendix B.

2.3 Treatment of Verbal Descriptors Interpretation of words such as "likely," " indeterminate,"

"unlikely," or " remotely possible" in Table 2.3 is subjective.

In cases where we have used these words, we did so because there was no clearcut way to quantify the likelihoods of the questions being asked. Still, some assignment of numerical values is necessary if the frequencies of the outcomes are to be estimated.

Table 2.7 shows four plausible assignments of values for l the verbal descriptors we have used. In most cases, we used Alternative 1 to quantify the outcome frequencies; however, we also investigated the sensitivity of some of the results to the choice of quantification alternatives. The results of the sensitivity study are described in Section 3.2.

2.4 Summary of Plant Features and Accident Sequences In this section, we will discuss specific reactor plant features which shape the form of the containment event tree and will describe the various accident sequences which have been quantified for Sequoyah. Details of the event tree construction and quantification are provided in Appendix B.

2.4.1 Plant Features Important to Containment Response The Sequoyah containment is a cylindrical steel vessel with a hemispherical dome and a flat bottom, surrounded by a concrete shield building. It has a much smaller free volume than the typical PWR containment; as shown in Figure 2-1, to compensate for this smaller volume in accomaodating steam pressures that could be generated during accident conditions, the upper portion of the containment has an annular compartment containing ice. As steam is blown down from the primary system during an accident, it is directed up through the ice, where it is condensed. Water then drains back into the lower compartment of the containment.

Three active containment safety features are provided: a containment spray system, an air return fan system and a hydrogen ignition system. The spray system consists of two

l DOME CoWPARTMENT

,_,,,'s N

• [ \

/

/ CONT AINMENT/

PRAVs

\g

1. \

/ \

/ \

j / Top oEck POLAR CRANE g J l doors v n y

l I

1.,

,s-i

\

t -

x

_ _.\

l cont AINME NT UPPER l l l wAtL CR ANE W ALL l PLENUM

/

g l l u

y 7.2sQ .

i a- - 0i l

i i w , j INTE R M Eot A T E l a l oEck oooRs I____1

=

i 2E 8

j l I ._.

ll l

I Ja l I i

l I

y o

{a l l

CONTROL Roo oRIVE Missile sniELo l

so l

y i l ICE coNoENstm l ,,

l l l l 'NLETLOWER PLENUM cooms l l g

! F?; _ _ _ _ _ _ _ _ _ _ J1g /[ ,

I goggg I gMg -

l PLENUM o

i -l ..[ i ml_l. l k.___j --,

L___

I i

M N ACC h qcp l' l h

J

)I l

VENTIL ATING EAN ANo l $

,N

[ _

e e -

EcuiPMENT suuP' l

I S b) bl l 4 ,

'# Il st s

v (*. / L I I .

,c L - 1 I

l ky-w d X *E^CT0"  ;/.

O \ l l

{ L____i___ _______I

/

/

l tG .

g }

1 l tower Rep = RE actor cool ANT PuWe l l coWPARTMENT l g so = sTE AM ctNER Atom l Acc. AccuMuL AToR L_ _ _ ._ _ .l I

Figure 2.1. Schematic of Sequoyah Containment i 1

i

- . . - - . . ~ - .- -- .

1 J

t

pump trains capable of drawing suction from the refueling water l storage tank (RWST) and discharging through spray headers in the dome of the containment building. Water sprayed into con-tainment passes through drains in the upper compartment floor to the containment sump. When the RWST reaches a low level, the pump suction is transferred by operator action to the l l sump. In this mode of operation. heat is removed from the con-tainment atmosphere by a heat exchanger in each of the pump trains: the heat exchangers are in turn cooled by a service water system. ItHis worth noting that the failure to remove the upper compartment drain covers following refueling opera-tions was assessed in RSSMAP (Ref. 7) to be an important source l of failure for both the spray and core cooling systems in the

recirculation phase, since water from spray flow would be l trapped in the upper compartment and would never reach the sump. Recent improvements in maintenance procedures have significantly reduced the likelihood that the drain covers

! could be left in place.

The primary purpose of the air return fan system, which is i located entirely within containment, is to ensure that air, displaced into the upper containment by the blowdown of steam

from the primary, is returned rapidly to the lower containment.

The fan system also serves to provide mixing of the containment i atmosphere, which can help to prevent the formation of pockets ,

j of hydrogen that could lead to local detonations.

i l The hydrogen igniters are provided to help preclude large hydrogen burns by burning relatively small quantities of hydro-gen as it is generated. Unlike the spray and air return fan j systems, which are both actuated automatically when containment 4

pressure reaches 3 psig (FSAR, Ref. 27), the hydrogen igniters

! must be initiated by the operators. Availability of the l igniters is dependent upon availability of ac power.

i The compartmentalization of the containment's free volume l is one reason for the added concern regarding hydrogen burns.

! If hydrogen were to' collect in either the upper or lower con-tainment, the likelihood of a burn capable of leading to con-tainment failure might be increased. This is particularly true i for burns occurring in the upper compartment, since the doors ,

1 at the entrance and exit of the ice condenser are designed to open only to flow from the lower to the upper compartment.

Thus, the pressures from a hydrogen burn in the upper compart-J ment would not be relieved by flow through the ice condenser.

i The design of the reactor cavity is such that it is essentially a large room, with a keyway located some distance

from the reactor vessel. For sequences in which the RWST ,

contents are injected into containment and there is melting of

~

the ice, the reactor cavity would invariably be flooded at the l

i i

I 12-i i

~--- - -_ -.__ ,.,. .--_,m-,,_,,-,,-.m,m. ,m,,-.- .-.,-_,,y---._-..-.-m.,,,..~_m.,.-..m r,- .-

time of vessel failure. Only for sequences involving failure of both emergency coolant injection and containment spray injection would it be likely that the cavity would be dry.

This design is particularly important in limiting the potential for direct heating of the containment atmosphere following vessel breach: if the RWST inventory were to be discharged, l there would be so much water in the reactor cavity (including a l substantial amount from the ice melting) that even the hot legs would probably be covered. For direct heating, the ejected core debris would have to displace perhaps as much as 15 ft of water before dispersing into the containment atmosphere. We felt that it was unlikely that the debris would not be quenched in the process of displacing this amount of water.

2.4.2 Description of Accident Sequences In this section, we describe the accident sequences assessed by ASEP to dominate the core melt frequency for Sequoyah and to be potentially important with respect to risk. As we mentioned previously, the accident sequences were developed within the context of plant damage states to promote coordination between the sequence and containment response analysts. These plant damage states reflect consideration of several parameters that could affect the progression of a core melt accident, including the following:

• The size and location of any breach in the reactor coolant system;

• The quantity of water injected it.to containment;

• The status of the containment spray system for both containment heat removal and reduction of fission products in the containment atmosphere;

• The availability of the other containment safety features, including the air return fans and the hydrogen igniters; e Whether or not the core is successfully cooled by rejecting decay heat to the containment for some time prior to the loss of core cooling, such that the ice in the ice condenser may be largely melted as the core melt progresses; and

• Whether or not the sequence results in a bypass of containment.

The plant damage states were developed in a two-step process. In the first step, a set of damage states was defined that reflected the major choices arising from the considera-tions noted above and that also produced a manageable number of outcomes for evaluation by ASEP. Five categories of conditions affecting containment response were identified. The categories and corresponding conditions are as follows:

1. Type of initiating accident, and in particular, the size of the initial breach in the reactor coolant system:

A Large or intermediate LOCA, such that the reactor coolant system is at relatively low pressure just prior to the time that the core debris breaches the reactor vessel.

S Small LOCA, leading to high or intermediate pressure in the reactor coolant system prior to vessel breach.

T The reactor coolant system is essentially intact, with no significant source of leakage other than cycling of the pressurizer relief valves.

V The sequence involves an interfacing systems LOCA, such that the containment is directly bypassed by the core melt accident itself.

2. The status of emergency core cooling systems (which is relevant both to the determination of the amount of water in the reactor cavity at the time of vessel breach and the amount of ice melting prior to vessel breach).

I Emergency core cooling is available in the injection phase, but not in the recirculation phase:

B Emergency core cooling succeeds in both the injection and recirculation phases, but may fail as a consequence of containment failure; and N Emergency core cooling fails in the injection phase, and is therefore also unavailable for the recirculation phase.

3. Availability of containment heat' removal:

Y Containment heat removal is available; or N Containment heat removal is not available.

4. Availability of containment spray system:

i

A i

t Y Containment spray system is available in the injec, tion and recirculation phases; or N Containment spray system is unavailable in the recirculation phase.

e

5. Availability of ac power. This is signifi-l cant immediately in determining whether or

. not the hydrogen igniters and air-return fans

! are available-(both are assumed to fail only in the event of station blackout,_with the exception that the operator action to actuate the igniters is reflected in the containment event tree), and also presents the possibil-ity of recovering containment safety features

! (sprays, fans, and igniters) late in the j accident if power is restored.

l Y Ac power is available; or l

j N The sequence involves a station black-4 out, and ac power is therefore not available.

t-The plant-damage states were therefore delineated by taking j all of the combinations of conditions from the five categories, leading to five-character designators for the states. For example, AIYYY would include sequences involving a large LOCA, with successful emergency coolant injection (but failure in recirculation), and availability of both containment heat removal via the spray system and ac power.

In order to characterize some of the questions in the

~

i containment event tree appropriately, we developed a finer discrimination for the plant-damage states, accommodating the following considerations:

l

• We subdivided the small LOCA in category 1

j. into LOCAs that would reduce RCS pressure to approximately the point at which injection of

the cold leg accumulators occurs prior to ,

i vessel breach (designated as S2 LOCAs); and l very small LOCAs that would cause the RCS I i

pressure to remain high up to the point of

vessel breach (referred to as S3 LOCAs).

e We broke the category identifying the status

. . of the containment spray system into whether l or not sprays failed during injection or recirculation. This breakdown, in

!' combination with the status of emergency core cooling, identifies whether or not the f

.L

contents of the refueling water storage tank (RWST) have been injected. The breakdown for this category is therefore as follows:

I Containment sprays operate in the-injection phase, but are unavailable in recirculation:

R Sprays are available in recirculation

,but not in injection (note that spray l flow in this case would be from the low-pressure recirculation system,

rather than the spray system itself, since the implication is that the spray system is unavailable)

B Sprays are available in both injection and recirculation; and N Sprays are unavailable in both the injection and recirculation phases.

I e We added a sixth category to include conditions that could affect the potential for an induced LOCA during core melt. This

category is only of interest for LOCAs smaller than the S2 category. The conditions for this category are as follows

A Auxiliary feedwater is available, so that an induced steam generator tube rupture is precluded, but seal cooling for the reactor coolant pumps (RCPs) is not available; R RCP seal cooling is available, but auxiliary feedwater has failed:

B Both auxiliary feedwater and seal cooling are available; and

, N Neither auxiliary feedwater nor seal cooling is available.

Many of the combinations of the six categories suggest damage states that are not logically possible (e.g., if ac power is unavailable, emergency core cooling, containment heat removal, containment spray, and RCP seal cooling are also not

functioning). For those damage states'that remain, we j interacted with ASEP to ensure that the sequences they t delineated could be appropriately assigned. The damage states that were assessed to have a frequency contribution'are defined in Table 2.8. The sequences assigned to them are described in the following discussion.

l Because the V sequence results directly in a release of i

fission products outside containment, the other parameter cate-l gories are not of interest. For this sequence, the only rele-vant question.is whether or not the location of the break is '

submerged by water from the reactor coolant system and/or the RWST.

Plant-damage state ABNIY implies that core cooling is successful for some period, but eventually fails as a conse-quence of containment overpressurization in the absence of con-

! tainment heat removal. ASEP identified four sequences that l contribute to the frequency of this damage state: (1) sequence i

SGt t, which is initiated by an intermediate LOCA, and

, includes failure of cooling of the containment heat exchanges 2

(2) sequence S 1F, an intermediate LOCA with failure of con-  !

1 tainment heat removal due to the unavailability of the spray system (3) sequence AG1, a large LOCA with no containment.

4 heat removal; and (4) AF, a large LOCA with failure of I containment sprays.

Damage state AIYBY also is comprised of four sequences, i Two are intermediate LOCAs with successful coolant injection,

{ one with failure of high pressure recirculation (sequence j SH),

12 and the other with failure of low pressure recir-

! culation, which is needed to support high pressure recircula-

! H ). The third sequence is AD S , a large LOCA tion (S 44 l with failure of the cold-leg accumulators to inject (low pres-

sure injection will provide flooding of the reactor cavity, but

] will not he able to prevent core melt). The fourth sequence, f sequence AHt , is a large LOCA with successful low pressure 4 injection but failure of low pressure recirculation. i s

i Damage state AINBY includes sequences that involve failure of core cooling in the recirculation phase and unavailability j of containment heat removal, but with spray flow available.

1 The sole sequence contributing to this plant-damage state is H F, an intermediate LOCA with failure.of high pressure S t2  ;

recirculation and containment spray recirculation.(but success-ful spray flow from low pressure recirculation).

l f Plant-damage state AINIY is similar, except that the con-tainment spray system operates only in the injection phase.

Sequence S t4 H F, an intermediate LOCA with failure of both

, low pressure recirculation and containment spray recirculation, j dominates the frequency of this state.

l Sequence AD6 is a large LOCA with failure of low pressure j injection, but with containnent heat removal provided by the

spray system. This sequence is assigned to damage state ANYIBY.

l.

i l

I j  ! l

.i

Two sequences dominate the frequency of plant-damage state S21YBY: (1) SH, 22 a small LOCA with failure of high pressure recirculation; and (2) SH, 2 3a small LOCA with failure of low pressure recirculation. In both cases, 1 containment heat removal is provided by the spray system. l Plant-damage state S NNNY 2 encompasses sequences in which there is no core cooling or containment spray operation, but ac power is available. The sequences that contribute to this damage state involve failure of component cooling water, which is a vital support system for both the emergency core cooling and containment spray systems. A loss of RCP seal cooling is assessed to result. One such sequence is an accident initiated by a loss of component cooling water, sequence Teew. The other sequence assessed to contribute is TDCIgD3WDt3 H P, which is a loss of a de bus and subsequent failure of component cooling water.

Plant-damage state S 2 NNNN also includes sequences involving transient-induced LOCAs, but they result from the total loss of ac power. The two contributing sequences are Tt3D WD t P. which is a seal LOCA that results from loss of seal cooling and seal injection as a consequence of station blackout, and T tQt Dt P, which is a station blackout followed by a stuck-open pressurizer relief valve.

Damage state S2 INIY is comprised of sequences in which core cooling and the containment spray system both fail in recirculation. The only contributing sequence was assessed to be S 2II3 P, failure of low pressure recirculation and containment spray recirculation.

Small LOCA sequences in which core cooling was not available in the injection phase but containment heat removal is functioning are represented by plant-damage state S NYBY. 2 The only sequence assigned to this damage state was TK t2 K Ze which is a failure of the scram system at a time when the moderator temperature coefficient is not sufficiently negative to prevent gross overpressurization of the RCS, Damage state S3 IYBYB includes sequences initiated by very small LOCAs, with emergency core cooling succeeding in injection but not in recirculation, and with containment heat removal, ac power, steam generator feedwater, and RCP seal cooling all available. The two sequences assigned to this damage state are S 3 II2 and S 311 3 , which involve failure of high pressure recirculation and low pressure recirculation, respectively.

Plant-damage state S3 INIYU is similar, except that the spray system fails in the recirculation phase. The dominant ,

sequence is S 3H3 P, which is a very small LOCA with failure of both low pressure recirculation and containment spray recirculation.

-la-

I 1

l

)

Sequences in which spray flow is available but containment heat removal is failed (for very small LOCAs) are assigned to damage state S3 INBYB. Sequence S 32 H F, a very small LOCA with failure of high pressure recirculation and the containment spray system contributes the frequency for this plant-damage state.

Station blackout sequences in which a LOCA does not occur appear in plant-damage state TNNNNN. The sequence dominating its frequency is T tL ti D F, which involves failure of all ac power and failure of the turbine-driven auxiliary feedwater l pump following depletion of the batteries, causing loss of all control and indication.

Plant-damage state TNYBYR includes transient sequences involving a total loss of core cooling but with containment heat removal available. Three sequences contribute to its frequency. Two of these, TDCI LPt I and TDCII bPt.

l are initiated by a failure of a de power bus, followed by failure of auxiliary feedwater. The de failure prevents the pressurizer pilot-operated relief valves (PORVs) from opening, precluding feed-and-bleed cooling as an option for core heat removal. The third sequence is T 2L IMP1, which is a loss of main feedwater followed by failure of auxiliary feedwater, with the PORVs unavailable.

The final plant-damage state assessed to have a non-negligible frequency is TNYBYB, which is a transient with no core cooling, but with containment heat removal available and with both auxiliary feedwater and RCP seal cooling func-tioning. The only sequence that contributes to this damage state is TK K Dt 2 4, a sequence with failure to scram and unavailability of the high pressure injection system to provide boration and eventual shutdown of the reactor.

3. RESULTS The results of the containment event analysis are presented in this section. The results for particular sequences are described in some detail and tabulated; each of these sequence tables has three parts. The first is the sequence frequency as estimated by ASEP. The second part of the table gives the containment failure mode probabilities for the optimistic, central, and pessimistic walkthroughs. Two cases are presented for the central and pessimistic analyses. Case 1 excludes in-vessel steam explosion and ex-vessel direct heating phenom-ena, and case 2 includes them. (In the optimistic walkthrough, these either do not occur or have negligible effect so the two cases are not differentiated.) The third part of each table describes the principal containment pathways and whether a calculation was made for them in BMI-2104.

The results are also tabulated in a more condensed format for all of the plant damage states. These tables present, for each damage state, the containment failure modes that were identified in our meetings with Battelle and other laboratories aimed at grouping (or binning) the failure modes for source term considerations.

3.1 Sequence Results for Sequoyah Tables 3.1 through 3.6 summarize the results of the containment failure analyses for the following six sequences:

• SH2, 2 a small LOCA followed by successful emergency core cooling in the injection phase but failure in the recirculation phase;

• S 2H 2F, a sequence similar to S 2H, except that the containment spray system also fails to function;

• SD, 3 i a sequence initiated by a LOCA caused by failure of the reactor coolant pump seals, followed by failure of emergency coolant injection; e SH, 3 2 a very small LOCA followed by successful emergency core cooling in the injection phase but failure of recirculation;

• S 3H 2F, a sequence similar to S 3H2 , except that the containment spray system also fails; and l

• TLDF, 1 1i which is a transient with loss of all core cooling due to a total loss of ac power (i.e., a station blackout), followed by failure of the turbine-driven auxiliary feedwater pump to provide core cooling. Note that this sequence implies that all active containment safety features (e.g., the containment spray system, fan return system and hydrogen igniters) are unavailable at least up to the onset of core melt.

The important pathways for each of the sequences are also described in the following sections.

3.1.1 Sequence S H22 Table 3.1 summarizes the optimistic, central, and pessimistic containment failure modes for sequence S H2 2 at Sequoyah. The modes of containment failure range from no containment failure in the optimistic case to almost assured failure at the time of vessel breach due to steam spike, RCS depressurization, and hydrogen combustion in the pessimistic case. In the central case, there is about an 80% chance that no containment failure will occur; the majority of the remaining frequency results in a late overpressurization of containment due to the generation of steam and non-condensible gases. There is a small (about 2%) chance that containment failure will occur at or before the time of vessel breach.

For this sequence, the likelihood that the ice would be melted or bypassed by the time of vessel breach is quite low in the central estimate (0.1%), although it is substantially higher in the pessimistic walkthrough (about 10%). The operation of core cooling in the injection phase and the containment spray system ensure that the contents of the RWST will be injected into containment, so that the reactor cavity would be heavily flooded prior to vessel breach.

The differences in modes of' containment failure among the three walkthroughs result from different estimates regarding hydrogen production and combustion. In the optimistic analy-sis, the hydrogen production is minimized and continuous localized burning near the igniters occurs. This results in very low pressure rises due to hydrogen combustion, and the containment does not fail early. Also in this walkthrough a coolable debris bed forms. This, coupled with the availability of containment heat removal, prevents late containment failure. Thus, the majority of the outcomes result in no containment failure.

In the central walkthrough nearly half of the zirconium is predicted to oxidize before vessel breach. The sudden release of hydrogen from the RCS at vessel breach results in some burns of a global nature, leading to the small chance of early con-tainment failure. the chance that the overpressurization will occur can result from two different causes with approximately equal likelihood: (1) the potential for the containment sprays to fail following vessel breach (e.g., from debris in the recirculation sump), causing loss of containment heat removal; or (2) the failure of a coolable debris bed to form, leading to core-concrete interactions and the associated release of non-condensible gases.

Pessimistically, all of the zirconium oxidizes before vessel breach. This results in a very large release of hydrogen at the time of the breach. A large hydrogen burn then occurs which is nearly certain to fail containment early. (It is worth noting that the combustion model matters very little here because the amount of hydrogen released from the RCS determines the magnitude of the burn.) Also, a small chance of contain-ment failure is predicted due to vessel depressurization and steam spike alone.

We should further note that direct heating has no effect on the results for this sequence, because of the massive amounts of water present in the reactor cavity and lower containment.

In addition, we estimate for the pessimistic case that an in-vessel steam explosion fails containment a fraction of the time (about 10%).

3.1.2 Sequence S 2H2 P Table 3.2 summarizes the results of the three walkthroughs

'for sequence SgH2F at Sequoyah. The predominant differ-ences in containment response for this sequence relative to the previous sequence (S H22) is that containment sprays are unavailable in the recirculation phase, leading to'the unavailability of containment heat removal. For the optimistic and central walkthroughs, the chance that containment will not fail is shifted to late overpressurization. In the central G estimate, the containment is still most likely to fail due to a ,

combined steam spike / hydrogen burn at vessel breach.

For this sequence, the likelihood that the ice will be melted or bypassed prior to vessel breach is identical to.that in the previous sequence. However, the chance that the ice function will be impaired prior to the onset of core-concrete interactions in the central estimate (about a 10% probability) can be important. Once again, the reactor cavity will be heavily tlooded due to the injection of the RWST contents.

t N

In the optimistic walkthrough, hydrogen burn considerations are essentially identical to those for sequence S 22 H , and j the containment is virtually assured to fail late due to steam

( ove'rpressurization. Because Sequoyah has a free-standing steel l containment and the Containment Performance Working Group l provided us with information dismissing the possibility of i le'akage which would arrest the pressure buildup, we concluded that containment failure would always be a gross rupture.

In the central walkthrough, the production of hydrogen once again leads to a small (2%) probability of containment failure due to a hydrogen burn in combination with the steam spike at the time of vessel breach. Otherwise, late overpressurization is nearly certain. The pessimistic estimate is nearly identical to that for sequence S H22 As in the previous sequence, the water in the reactor cavity is sufficient to preclude direct heating.

3.1.3 Sequence S H32 As shown in Table 3.3, the results for sequence S 3 H2 indicate trends similar to those obtained for sequence SH.

2 2 They range from near certainty that containment will not fail in the optimistic case to near certainty that early containment failure will occur prior to or at the time of vessel breach in the pessimistic case. In the central case, the most likely outcome is no failure, with a slightly higher chance of late steam overpressurization (about 23%). The likelihood of early containment failure is higher than for SH22 but still quite low (about St).

\

For this sequence, a large induced LOCA may occur due to high temperatures and pressures during the period of core Aegradation. (In the previous case for S2 H2 , the reactor coolant system pressure at the time of vessel breach was judged to be sufficiently low as to preclude the development of a

.large induced LOCA.). Because the failure of core cooling occurs in the recirculation phase, a significant amount of energy may have already been released to containment during the injection phase. When the induced LOCA occurs, the accumula-tors will discharge and will quench the core, causing a large additional release of steam. Recent MARCH calculations by Battelle indicate that this will cause melting of as much as 90% of the ice. There is, therefore, a significant chance of ice melting or bypass prior to vessel breach in the central (13%) and pessimistic (about 23%) walkthroughs.

In the optimistic case, no containment failure is very likely. There is a small (about 3%) chance that the meltthrough of the basemat will occur. This potential arises if there is a large induced LOCA, leading to an increased chance that the core debris will be confined to the reactor cavity following vessel breach.

i The chance for late containment failure due to failure to i form a coolable debris bed is somewhat larger in the central i walkthrough, leading to the 23% chince-of late overpressurization.

In the pessimistic walkthrough,'early containment failure is again very likely. However, in this case there is a substantial possibility that containment failure will occur prior to vessel breach due to a hydrogen burn. This occurs.

when there is a large induced LOCA, leading to early releases of large quantities of hydrogen. It would be likely that the

. ice would be unavailable at the time of vessel breach for this case. Otherwise, containment is most likely to fail at vessel  !

breach, due to a combined hydrogen burn / steam spike or (in the

,l second set of walkthroughs) due to an in-vessel steam explosion.

L 3.1.4 Sequence S 32 H F il Table 3.4 summarizes the results for the three walkthroughs for sequence S 32 H F. The results and considerations,are very similar to those described above for sequence G 32 H e

! except that the fraction that led to no containment failure j will instead be assigned to late overpressurization due to unavailability of containment heat removal.

3.1.5 Sequence S 3D1 f

! The optimistic.' central, and pessimistic results for

! sequence S 3 D at Sequoyah are summarized in Table 3.5. The results are virtually identical to those for sequence S H32

! with the important exception that, because core cooling fails j in injection, melting or bypass of the ice is much less likely.

3.1.6 Sequence T tti L D F

! Table 3.6 summarizes the results for the optimistic,'cen-

tral, and pessimistic walkthroughs for sequence T t1i L D F .'

at Sequoyah. The outcomes range from a significant chance that containment will not fail in the optimistic walkthrough to 4

virtual certainty that containment will fail prior to or at the i time of vessel breach in the pessimistic walkthrough. In the central walkthrough, the most likely outcome is failure at the time of vessel breach due to a hydrogen burn and the steam j spike (about 70s probability). Early failure may also occur with a much smaller likelihood due-to'an earlier hydrogen. burn

{' or due to direct heating. If early failure does not occur, late steam overpressurization is the most probable outcome.=

. In this case, there is again a substantial chance for a l' large induced LOCA to occur. If this does occur, the steam generated by quenching of the core'when the accumulators discharge can result in substantial melting of the ice. Bypass 4

I i

l '

of the ice may also occur for this sequence due to detonation of hydrogen in the ice condenser. These factors combine to produce about a 15% likelihood of ice melt or bypass in the central walkthrough, and a 55% chance in the pessimistic case.

There is about an additional 40% likelihood of ice melt or bypass following vessel breach in both the central and -

pessimistic walkthroughs.*

For this sequence it fi:s also important to note that the reactor cavity will be dry at the time of vessel breach, since no systems are available to inject the RWST water. If no induced LOCA occurs, the accumulators will provide some water i

to the cavity following vessel breach, contributing to the steam spike but not sufficient to prevent direct heating.

The results for the optimistic walkthrough are strongly affected by the recovery of ac power late in the accident, such that containment heat removal can be restored, preventing late overpressurization. This leads to the likelihood (68%)~that no containment failure will' occur. If power is not available, steam and non-condensible gases will eventually overpressurize containment.

In the central walkthrough, there is a large probability of containment failure at the time of vessel breach due to a steam spike and hydrogen burn in the first case, or with a small (71) contribution from direct heating in the second case. The early failure is particularly significant since the ice may already have melted at the time of vessel breach, and is very likely to be melted or bypassed by the time that releases due to core-concrete interactions are generated.

,_

• Another situation which could produce a relatively early con-l tainment failure during station blackout with simultaneous I defeat of the ice condenser function has recently been sug-gested by researchers from Brookhaven and Sandia National Laboratories (Ref. 41). The scenario involves the high pres-sure ejection of core debris from the reactor vessel resulting

< in the accumulation of core debris against the containment steel shell. Subsequent meltthrough of the shell would produce a pathway by which fission products could escape to the outside environment without passing through the ice con-denser. Since severe flooding of the reactor cavity would-preclude debris dispersal in this manner, this scenario is far more likely for station blackouts than for other sequences. The possibility of shell meltthrough has not been included in the present analysis but will be treated in the SARRP uncertainty evaluation when the Sequoyah risk rebaselining is performed (Ref. 40).

I i i

In the pessimistic walkthrough, the containment is nearly certain to fail due to a hydrogen burn prior to vessel breach if a large induced LOCA occurs (such a LOCA is induced approximately 50% of the time). Otherwise, the outcomes are very similar to the central walkthrough. It should also be noted that the ice is virtually assured to be melted or bypassed very soon after vessel breach.

For this sequence, the steam generators will be dry, introducing the potential for a steam generator tube rupture to be induced. This was assessed to occur with a probability of 14% for all three walkthroughs for this sequence.

3.1.7 Results for Plant-Damage States The likelihoods of the various containment release modes for each of the plant-damage states assessed to have non-negligible frequencies are presented in Tables 3.7 through 3.11 for the three walkthroughs and for the cases with and without direct heating and in-vessel steam explosion considered. The release modes into which these outcomes were sorted were based on the considerations important to the source-term binning.

The release modes reported in these tables may be further grouped to establish the source term bins for risk assessment.

That process will be described in the SARRP risk report (Ref.

1). The current release modes are as follows:

1. No containment failure.

2. Basemat meltthrough.

3. Late overpressurization due to the generation of steam and non-condensible gases, with the ice condenser available through the time of vessel breach, and with the containment spray system operating through the period of core-concrete interactions.

4. Late overpressurization due to a hydrogen burn, with the ice condenser available through the time of vessel breach, and with the containment spray system operating through the period of core-concrete interactions.

5. Late overpressurization due to the generation of steam and non-condensible gases, with the ice condenser available through the time of vessel breach, and with the containment spray system failing prior to the release corresponding to core-concrete interactions.

6. Late overpressurization due to a hydrogen burn, with the ice condenser available through the time of vessel breach, and with the containment spray system failing prior to the release corresponding to core-concrete interactions.

7. Late overpressurization due to the generation of steam and non-condensible gases, with the ice 4 condenser not available at the time of vessel breach, and with the containment spray system operating through the period of core-concrete interactions.

8. Late overpressurization due to a hydrogen burn, with the ice condenser not available at the time of. vessel breach, and with the containment spray system operating through the period of core-concrete interactions.

9. Late overpressurization due to the generation of steam and non-condensible gases, with the ice condenser not available at the time of vessel breach, and with the containment spray system i failing prior to the release corresponding to j core-concrete interactions.

l 10. Late overpressurization due to a hydrogen burn, with the ice condenser not available at the time of vessel breach, and with the containment spray system failing prior ~to the release corresponding to core-concrete interactions.

11. Early containment failure due to a hydrogen burn prior to the time of vessel breach, with the ice condenser available and with the containment spray system operating at least through the time of vessel breach.

12. Early containment failure due to a hydrogen burn prior to the time of vessel breach, with the-ice ccadenser available and with the containment spray system failing prior to vessel breach.

l 4 13. Early containment failure due to a hydrogen burn prior to the time of vessel breach, with the ice condenser not available and with the containment spray system operating at least through the time of vessel breach.

14. Early containment failure due to a hydrogen burn prior to the time of vessel breach, with the ice condenser not available and with the containment spray. system failing prior to vessel breach.

15. Containment failure at vessel breach due to a s steam spike, with the ice condenser available through the time of vessel breach and with the containment spray system continuing to operate for a significant time after vessel breach.

16. Containment failure at vessel breach due to the combined effects of a hydrogen burn and a steam spike, with the ice condenser available through the time of vessel breach and with the containment spray system continuing to operate for a significant time after vessel breach.

17. Containment failure at vessel breach due to the combined effects of direct heating and a steam spike, with the ice condenser available at the time of vessel breach and with the containment spray system continuing to operate for a significant time after vessel breach.

18. Containment failure due to an in-vessel steam explosion.

19. Containment failure at vessel breach due to a steam spike, with the ice condenser available at the time of vessel breach and with the containment spray system unavailable, or failing soon after vessel breach.

20. Containment failure at vessel breach due to the combined effects of a hydrogen burn and a steam spike, with the ice condenser available at the time of vessel breach and with the containment spray system unavailable, or failing soon after vessel breach.

21. Containment failure at vessel breach due to the combined effects of direct heating and a steam spike, with the ice condenser available at the time of vessel breach and with the containment spray system unavailable, or failing soon after 4

vessel breach.

22. Containment failure at vessel breach due to a steam spike, with the ice condenser not available at the time of vessel breach and with the containment spray system continuing to operate for a significant time after vessel breach.

23. Containment failure at vessel breach due to the combined effects of a hydrogen burn and a steam spike, with the ice condenser not available at the time of vessel breach and with the containment spray system continuing to operate for a significant time after vessel breach.

24. Containment failure at vessel breach due to the combined effects of direct heating and a steam spike, with the ice condenser not'available at the time of vessel breach and with the containment spray system continuing to operate for a significant time after vessel breach.

25. Containment failure at vessel breach due to a steam spike, with the ice condenser not available at the time of vessel breach and with the containment spray system unavailable, or failing soon after vessel breach.

26. Containment failure at vessel breach due to the combined effects of a hydrogen burn ~and a steam spike, with the ice condenser not available at the time of vessel breach and with the containment spray system unavailable, or failing soon after vessel breach.

27. Containment failure at vessel breach due to the combined effects of direct heating and a steam spike, with the ice condenser not available at the time of vessel breach and with the containment spray system unavailable, or failing soon after vessel breach.

28. Containment failure prior to core melt, with the ice condenser available at the time of vessel breach and with the spray system operating through the time of vessel breach and for a significant period thereafter.

29. Containment failure prior to core melt,.with the ice condenser available at the time of vessel breach and with the spray system not available at the time of vessel breach.

30. Containment failure prior to core melt, with the ice condenser not available at the time of vessel breach and with the spray system operating through the time of vessel breach and for a significant period thereafter. j i

l 4

31. Containment failure prior to core melt, with the ice condenser and the spray system not available ,

at the time of vessel breach. -

32. Containment bypass with the point of release submerged by a substantial quantity of water. l

33. Containment bypass with the point of release not submerged by a substantial quantity of water.

34. Pre-existing leakage or containment isolation failure.

35. Induced steam-generator tube rupture (steam generators must be dry).

3.2 Sensitivities As mentioned in Section 2.3, the numerical values assigned to verbal descriptors such as "unlikely" or " remotely possible" are somewhat arbitrary, and the results could be sensitive to these choices. Accordingly, we performed a sensitivity study for an example sequence similar to Surry S D1 2 using the 4

four alternative numerical sets in Table 2.7. The results, depicted in Table 3.12, indicate that the variation of condi-i tional probability within each walkthrough (optimistic, central, pessimistic) is small compared to the difference in results between walkthroughs.

Although the results are not very sensitive to the choice

of numerical values, they are sensitive to the choice of verbal descriptors, i.e., whether phenomena are considered "likely" or I

"unlikely" to occur. This sensitivity has been covered by the choice of optimistic, central, and pessimistic walkthroughs.

4.

SUMMARY

AND CONCLUSIONS To interpret our results in the most illuminating way, it is necessary to clarify the meanings of the words " optimistic",

" central", and " pessimistic". For this purpose, let us offer the following observations. Our central estimates are intended to provide a glimpse of the median of the reactor safety community. That is to say, we have a reasonable expectation that if the community were polled on the subjects treated in this document, a substantial fraction would respond that the reality of the situation is better than that depicted by our central estimates, and a comparably substantial fraction would respond that it is worse. Among those who think that reality is better (a group that we could refer to as " optimists"), a substantial fraction would claim that it is or could be as benign as our optimistic estimates imply. Similarly, many of the opposing group (the " pessimists") would claim that reality is or could be as unfavorable as our pessimistic estimates imply.

Again, it should be pointed out that we were not able to apply these guidelines rigorously as a consequence of the small number of experts whose views were available on particular issues. The effects in these circumstances are illustrated by sensitivity studies performed for issues with very large uncertainties.

Viewed in this context, the results of our containment event analyses indicate quite clearly that there are large uncertainties in the reactor safety community's understanding of containment loading and response. The differences between the optimistic and pessimistic results attest that for many accident sequences, one cannot state the probabilities of the containment failure modes within narrow limits. However, one can identify the factors that drive the results and understand the reasons for the differences.

Fortunately, there are a limited number of factors that are governing, and they can be readily identified from our analysis. In the paragraphs below, we shall summarize the principal features of our results and the factors that drive them. In the process, we shall attempt to identify the sources of uncertainty which are principally responsible for the differences between the views of the optimist and the pessimist. Of course, the discussion will be limited only to those sequences which we have analyzed.

In the central walkthroughs, the likelihood of early containment failure (i.e., before or soon after vessel breach) is very low (less than 5%) for all sequences except station blackout. For station blackout, the likelihood of early failure is predicted to be about 80%. The principal factors affecting the potential for early failure of containment are whether or not the. igniters and air return fans are operating and.whether the reactor cavity-is sufficiently flooded to preclude direct heating and large steam spikes.

If the igniters and fans are not available, hydrogen will I not be burned in small quantities as it is released, leading to a substantial chance for a large burn at vessel breach in the-central walkthrough. This condition is.only realized for the station blackout sequence. This is also a sequence in which there will not be a large quantity of water injected into containment, providing an' opportunity for direct heating and l larger steam spikes. '

In the pessimistic walkthroughs, for all sequences that we analyzed, early containment failure nearly always occurs. For large_LOCAs, this early failure is predominantly the result of the burning of the large quantity of hydrogen produced during

{ core degradation and released prior to vessel breach. For small LOCAs, the hydrogen tends to be retained in the RCS until j the time of vessel breach, at which point the containment will

fail due to the combined effects of the hydrogen burn and the

steam spike.

a Another important issue for ice condenser containments is whether the ice will be melted or bypassed during periods when fission products are being released to the containment. If-melting or bypass has not occurred, the ice condenser can be-

! quite effective as a fission product scrubber. Our analyses indicate that melting or bypass of the ice before vessel breach or shortly after vessel breach can be expected to occur with moderate to high likelihood during station blackouts and during sequences initiated by very small LOCAs with failure of emergency core cooling in the recirculation phase after success in the injection phase.

The specific ways in which early ice melting or bypass can occur are the following:

e A transient initiating event in which a large LOCA l could be induced by high temperatures in the RCS

! during the period of core degradation. In this 4 case, the water in the RCS has been boiled off l prior to the developmentoof the large breach. For i

sequences initiated by large LOCAs, the water is lost much more quickly due to initial blowdown, rather than being converted to steam. Thus, the amount of ice melted prior to vessel breach is significantly less.

l i

- _ . _ ~ - .

• A small LOCA with success of emergency core cooling ,

in injection but failure in the recirculation phase, with a large LOCA again induced by high RCS temperatures. Note'that, as in the previous case, the key to whether or not ice is melted is the  !

timing of water boil-off from the RCS. In both  !

cases, the accumulators must discharge prior to '

vessel breach, resulting in temporary recooling of the core, and the generation of additional steam.

• A local detonation in the ice condenser. Detonable concentrations are predicted to occur only when the air-return fans are not operating and only after the core has slumped into the lower plenum of the ,

reactor vessel. Since the ice is already predicted to be about 50% melted at this point, it is postulated that the detonation may sufficiently disrupt the ice condcaser geometry to produce bypass paths.

4 Another situation which could produce a relatively early containment failure during station blackout with simultaneous defeat of the ice condenser function has recently been suggested (Ref. 41). The scenario involves the high pressure l

ejection of core debris from the reactor vessel resulting in the accumulation of the debris against the containment _ steel shell with subsequent melting of the shell. This possibility has not been included in the present analysis but will be included in the SARRP risk uncertainty analyses (Ref. 1).  ;

For transient initiating events or very small LOCAs, there is a significant chance that a steam generator tube rupture may-result from_the generation of very high temperatures in the reactor coolant system during core degradation (if the steam generators are dry). This could constitute an important path for bypassing containment, particularly for sequences in which no other early failure mode is likely. In addition, in-vessel steam explosions that fail containment are also important in the pessimistic walkthroughs; as uncertainties regarding this phenomenon are resolved, it could have a more significant influence than we have credited.

4 i

REFERENCES

1. (SARRP). A. S. Benjamin, et al., " Evaluation of Severe 4

Accident Risks and the Potential for Risk Reduction: Surry Power Station Unit 1," NUREG/CR-4551, SAND 86-1309, Volume l 1 Sandia National Laboratories, September, 1986.

l 4

2. (1150). " Nuclear. Power Plant Risks and Regulatory Implications," NUREG-1150 U.S. Nuclear Regulatory Commission, to be published.

3. (0956). " Reassessment of the Technical Bases for Estimat-ing Source Terms," U. S. Nuclear Regulatory Commission.

Report NUREG-0956, August 1985.

I

4. (BMI). J. A. Gieseke, et al., "Radionuclide Release Under Specific Accident Conditions," BMI-2104, Battelle Columbus Laboratories, Volumes II through VI, July, 1984.  !

5. (BMI). R. S. Denning, et al., "Radionuclide Release Calculations for Selected Severe Accident Scenarios: PWR. )

i Ice Condenser Containment Design," NUREG/CR-4624, BMI-2139, l Volume 2, Battelle Columbus Laboratories, July, 1986. '

i.

j 6. (ASEP). R. Bertucio, et al., " Analysis of Core Damage

, Frequency from Internal Events: Sequoyah Unit 1,"

5 NUREG/CR-4550, SAND 86-2084, Volume 2 (Draft Report). Sandia j National Laboratories. April, 1986.

l

7. (RSSMAP). " Reactor Safety Study, Methodology Applications Program," NUREG/CR-1659,- SAND 80-1897, Volume 1, Sandia

]

National Laboratories, February. 1981.

8. (CLNG). " Estimates of Early Containrent Loads from Core

Melt Accidents, Report of the Containment Loads Working Group," NUREG-1079 (Draft for Comment), U.S. Nuclear j Regulatory Commission. December, 1985.

J

9. (CLWG). Consensus Summaries for Standard Problems 1 through 6. Letter Reports Addressed to J. Telford, U. S.

Nuclear Regulatory Commission, May-June 1984.

10. (CLWG). F. E. Haskin, et al., "Combusti~on-Induced Loads in

! Large Dry PWR Containments," Proceedings of the 2nd Containment Integrity Workshop, NUREG/CP-0056, SAND 84-1514, A ugust 1984. Also, letter to M. Silberberg, U. S. Nuclear Regulatory Commission, March 9, 1984.

{ 11. (CPWG). Containment Performance Working Group, "Contain-ment Leak Rate Estimates," NUREG-1037 (Preliminary Draft),

U. S. Nuclear Regulatory Commission, 1984.

t

__ . _ , _ _ . , - . _ - - _ - . _ _ - _ _ ~ - __ , _. _ - - .

12.-(QUEST). R. J. Lipinski, et al., " Uncertainty in Radio-nuclide Release Under Specific LWR Accident Conditions "

l SAND 84-0410, Sandia National Laboratories. February 1985.

13. (IDCOR). IDCOR Program Reports, IDCOR Tasks 21.l'(Risk Reduction Potential) and 23.1 (Integrated Containment Analysis), Technology for Energy Corporation, November 1984,

14. (SASA). F. E. Haskin, et al., " Analysis of Hypothetical Severe Core Damage Accidents for the Zion Pressurized Water Reactor," NUREG/CR-1989, SAND 81-0504, Sandia National Laboratories, October 1982.

15. (SASA). A. L. Camp, et al., " MARCH-HECTR Analysis of Selected Accidents in an Ice-Condenser Containment,"

NUREG/CR-3912, SAND 83-0501, Sandia National Laboratories, j December 1984,

16. (SASA). S . .E . Dingman and A. L. Camp, Pressure-Temperature Response in an Ice Condenser Containment for Selected Accidents," Transactions of the 13th Water Reactor Safety Research Information Meetino, NUREG/CP-0071, October, 1985.

17. (SASA). R. D. Gasser, et al., " Analysis of Station Blackout Accidents for the Bellefonte Pressurized Water Reactor," NUREG/CR-4563, SAND 86-0576, Sandia National Laboratories. September, 1986.

18. (SASA). D. B. King and A. C. Peterson, " Gas Transport Calculations for a Large Dry PWR Containment (Bellefonte) for Arrested Sequences," NUREG/CR-45-99, SAND 86-0972, Sandia National Laboratories, to be published.

19. (SAUNA). J. B. Rivard, et al., " Identification of Severe i Accident Uncertainties," NUREG/CR-3440, SAND 83-1689, I Sandia National Laboratories, September 1984. ,

4 i

20. (PATF). A. S. Benjamin, et al., "SARRP - Risk Rebaselining and Risk Reduction Analysis," lith Water Reactor Safety

] Research Information Meetina, NUREG/CP-0048, Vol. 3 l January, 1984. '

i i

21. (SERG). Steam Explosion Review Group. "A Review of the Current Understanding of the Potential.for Containment Failure Arising from In-Vessel Steam Explosions " NUREG-

! 1116, U.S. Nuclear Regulatory Commission, February 1985.

22. (HIPS). High Pressure Ejection Test Series, Sandia National Laboratories, 1985.

j 23.-(RSS). " Reactor Safety Study." WASH-1400, NUREG-75/014, U.S. Nuclear Regulatory Commission, October 1975.

. _ . - - - . - _ , _ ~ _ . .__ _ _ _ ~ _ _ _ _ . . _ _ _ _ . . . _ _ _ _ _ _ _ . _ . . _

24. (ZPSS). " Zion Probabilistic Safety Study," Commonwealth Edison Co., 1981.

25. (SPSS). "Sizewell-B Probabilistic Safety Study," WCAP-9991, Westinghouse Electric Co., 1982.

26. (SPSA). "Seabrook Plant Probabilistic Safety Analysis,"

Yankee Atomic Electric Co., December 1983.

27. (FSAR). Final Safety Analysis Report for Sequoyah Unit 1.

I

28. (AMES). L. Greimann, et al., " Final Report, Containment Analysis Techniques, A State-of-the-Art Summary," NUREG/

CR-3653, SAND 83-7463, Ames Laboratory and Sandia National Laboratories, March 1984. '

29. (NRR). M. B. Weinstein, " Primary Containment Leakage Integrity: Availability and Review of Failure Experience," N uclear Safety, 21, 1980.

30. (BNL). W. T. Pratt, et al., " Containment Response During Degraded Core Accidents Initiated by Transients and Small Break LOCA in the Zion / Indian Point Reactor Plants,"

NUREG/CR-2228, BNL-NUREG-51415, Brookhaven National Laboratories, July 1981.

31. (A-43). A. W. Serkiz, " Containment Emergency Sump Perfor-mance, Technical Findings Related to Unresolved Safety Issue A-43," NUREG-0897, U. S. Nuclear Regulatory Commission, April 1983.

32. (NRR). W. G. Lyons, Presentation at NRC/IDCOR Meeting on Containment Loads and Fission Product Behavior. U. S..

Nuclear Regulatory Commission, May 15-17, 1984.

33. (ZIP). W. B. Murfin, et al., " Report of the Zion / Indian

! Point Study," NUREG/CR-1410, SAND 80-0617, Sandia National Laboratories, August 1980.

34. (HYDR). A. L. Camp, et al., " Light Water Reactor Hydrogen Manual," NUREG/CR-2726, SAND 82-ll37, Sandia National Laboratories, August 1983.

35. (IVSE). M. Berman, et al., "An Uncertainty Study of PWR Steam Explosions " NUREG/CR-3369, SAND 83-1438, Sandia National Laboratories, May 1984.

36. (LPS). S. J. Niemczyk, et al., "The Consequences from d

Liquid Pathways After a Meltdown Accident," NUREG/CR-1596, i SAND 80-1669, Sandia National Laboratories, June 1981.

37. (EPRI). " Loss of Offsite Power at Nuclear Power Plants:

Data and Analysis," EPRI NP-23Ol. Interim Report, Electric Power Research Institute, March 1982.

38. (HPE). G. A. Greene, M. Pilch, and W. W. Tarbell, Letter to A. S. Benjamin, 19 March 1986.

39. (UWISC). "A Review of the Severe Accident Risk Reduction Program (SARRP) Containment Event Trees," NUREG/CR-4569, University of Wisconsin, May, 1986.

I i

n s

4 x . , , . - - - - , -

l 1

i INTENTIONALLY LEFT BLANK 4

4

Table 2.1. USE OF INFORMATION SOURCES IN ADDRESSING THE ISSUES IN THE CONTAINMENT EVENT TREE INFORMATION SOURCES

• ISSUE CLWG CPWG BMI-2104 QUEST IDCOR SASA SAUNA PTAP SERG HIPS PRA AE OTHERS

1. Size of Preexisting Containment Leakage --

X -- -- X -- -- -- -- -- X -- X

2. Size and Location of the Primary System -- -- -- -- -- -- -- -- -- -- -- -- --

Break During the Melt Release (e.g., Hot Leg vs. Cold Leg).

3. Timing of Accumulator Discharge Relative -- -- X --

X X -- -- -- --

X -- X to Timing of Reactor Vessel Breach.

4. Occurrence of In-Vessel Steam Explosion -- -- -- -- -- -- X -- -- -- -- -- X Large Enough to Fall the Reactor Vessel.

5. Occurrence of In-Vessel Steam Explosion -- ,-- -- --

X -- X --

X --

X --

X Large Enough to Fail the Containment.

6. Timing and Magnitude of Early Hydrogen X -- X --

X X X X -- --

X -- X Burns.

7. Magnitude of the Ex-Vessel Steam Spike X -- X X X X X X -- X X --

X

8. Extent of Direct Heating of the X -- -- -- X -- -- -- --

X -- -- --

Atmosphere Following Vessel Breach.

I LJ 9. Containment Structural railure Pressure -- -- -- -- X -- -- X --

W -- X X --

I 10. Size of Containment Leakage Induced by --

X -- --

X -- -- -- -- -- -- -- --

Temperature or Pressure.

11. Survivability of Cont,ainment Sprays and -- -- -- --

X -- -- -- -- --

X -- X Air Return Fans at Various Times During the Accident.

12. Functionality of Ice Condenser at Various -- -- -- -- -- -- -- -- -- -- -- -- X Times During the Accident

13. Extent of Core-Concrete Interaction X --

X --

X X X -- -- -- X -- --

14. Timing and Magnitude of Late Hydrogen Burns X --

X -- X X -- -- -- -- -- --

X

15. Potential for Direct Core Debris Attack X -- -- -- -- -- -- -- -- -- -- -- --

on Containment Structures

16. Potential for Cor6 Debris to Melt Through -- -- X --

X -- X -- -- -- X -- X the Basemat

17. Potential for Effluent to Pass Through -- --

X --

X -- -- -- -- -- -- -- --

Adjacent Structures, Such as the Auxiliary Building

'Information sources used in evaluating a particular issue are denoted by an "X".

Y

Table 2.2 EVENT DESCRIPTIONS FOR PWR ICE CONDENSER CONTA1KMENTS Prior Question Event Tree Question Dependencies

1. Is ac power available early? None

2. Is there pre-existing leakage or isolation failure? None

3. What is the initial break location? None

4. What is the initial break sise? None l l

5. Is containment initially bypassed? 3

6. Are the steam generators wet or dry? None 7 Are the air return fans operating initially 1

8. Does emergency core cooling operate in the injection mode? I

9. Do the sprays operate in the injection mode? 1

10. Do the sprays operate in the recirculation mode? 9

11. To what degree is there bypass of the auxiliary building? 2,3,6

12. What is the location of any induced failure in the primary system? 4,6

13. What is the effective site of any termperature-induced primary system failure? 12

14. What is the level of early containment bypass during meltdown? 3,4,12,13

15. Do the ignitors operate early in the scenario? 1

16. Is the lower compartment inert during core degradation? 7

17. What is the magnitude of the early baseline pressure in containment? . 7

18. Does a local detonation occur in the ice condenser? If so, what are the effects? 7

19. Is there a deflagration in the lower compartment before vessel breach? Also, what is 4,5,12,13, the magnitude of the pressure rise from such a burn? 14,15,16

20. Is there a deflagration in the upper compartment before vessel breach? Also, what is 4,5,13,14, the magnitude of the pressure rise from such a burn? 15,19

21. Does containment fall due to an early deflagration? Also, what is the mean 17,19,20 containment failure pressure and the standard devialtion?

22. To what degree is there bypass of the auxiliary building early in the scenario? 11,16,21

23. Has the ice melted out of or have bypass paths developed in the ice condenser before 3,4,8,13,18 vessel breach?

24. What is the status of the air return fans after hydrogen burns? 7,19,20

25. What is the status of the containment sprays after hydrogen burns? 10,18,21 26 What is the primary system pressure during core degradation? Also, what is the 4,12,23,24 pressure rise in containment due to primary system depressurization?

27 What is the mode of reactor vessel breach? 26

28. Is the reactor cavity wet at vessel breach? 8,9

29. Does direct heating occur? Also, what is the pressure rise in containment due to 23,26,27,28 direct heating and/or a steam spike?

30. Does a detonation occur in the ice condenser at the time of vessel breach? Also, 18,23,24 what are the ef fects?

31. Has the ice melted out of or have bypass paths developed through the ice condenser 3,4,8,13, at the time of vessel breach? 23,28,30

32. Does a hydrogen deflagration occur at the time of vessel breach? Also, what is the 4,13,14, pressure rise due to the deflagration? 15,19,20 l 33. Does containment fail at vessel breach? Also, what is the mean failure pressure and 17,26,29,32 standard deviation?

34. What is the mode of intermediate containment failure? 2,18,21,27, 29,30,32,33 l

l 40 l

l I

l - _ _ - - , -

_ _ ~ , =_ _

Table 2.2 EVENT DESCRIPTIONS FOR PWR ICE CONDENSER CONTAINMENTS Prior Question Event Tree Question Dependencies

35. To what degree is there intermediate auxiliary building bypass? 22,30,34

36. What is the status of the air return fans after vessel breach? 24,32

37. What is the status of the containment sprays after vessel breach? 25,34

38. Is ac power restored af ter vessel breach? 1

39. Are the igniters on late? 1,15,38

40. What is the status of the air return fans late? 36,38

41. What is the status of the containment sprays late? 37,38

42. What is the late baseline pressure in containment? 31,40,41

43. Is a coolable debris bed formed and maintained after vessel breach? 26,27,28

44. Is there a late deflagration? Also, what is the pressure rise associated with 2,19,20,21 the deflagration? 32,34,39,43

45. Does the containment fail due to a late deflagration? 2,21,34, 42,44

46. What is the status of the containment sprays very late? 41,45

47. Does containment failure occur late due to non-condensible gas and/or steam buildup? 2,18,21,34 43,46

48. To what degree is there late auxiliary building bypass? 35,45,47

49. Does basemat meltthrough occur? 43,46 f

41

l l

i l

l l

TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOODS FOR SEOUENCE S 3HF LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC 2- Pre-existing containment (a) None (1) Tech. specs., or less 0.983 0.965 0.938 leakage or isolation failure (ii) Greater than (1), but not 0.017 0.033 0.058 sufficient to preclude l gradual overpressurization later in accident l

(iii) Greater than (11), but not 0.0 0.0025 0.004 sufficent to depressurize containment in about 2 hours2.314815e-5 days <br />5.555556e-4 hours <br />3.306878e-6 weeks <br />7.61e-7 months <br /> (iv) Greater than (iii) 0.0 0.0 0.0

3. Initial reactor coolant S3 initiator (1) Hot leg 0.0 0.0 0.0 system break location (ii) Cold leg 1.0 1.0 1.0 t, (iii) other 0.0 0.0 0.0 w

7. Initial operation of air Ac power available (1) Fans operate 1.0 1.0 1.0 retur 'a n s (ii) Fans available but not 0.0 0.0 0.0 operating (iii) Fans fail 0.0 0.0 0.0

8. ECC operation in the S3 HF Sequence (1) ECC injection 1.0 1.0 1.0 injection phase (ii) No ECC injection 0.0 0.0 0.0

9. Containment spray S3 HF Sequence (1) Sprays operate 1.0 1.0 1.0 operation in injection (ii) Sprays available but 0.0 0.0 0.0 not operating (iii) Sprays fail 0.0 0.0 0.0

TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

10. Sprays operate in S3 HF Sequence (1) Sprays operate 0.0 0.0 0.0 recirculation mode (ii) Sprays available but 0.0 0.0 0.0 not operating (iii) Sprays fail 1.0 1.0 1.0

11. Degree of auxiliary (a) Pre-existing leakage (1) Complete bypass or Inpossible Impossible Impossible building breakthrough or isolation failure breakthrough or bypass greater than (i) above (ii) Partial bypass Unlikely Unlikely Unlikely (iii) No bypass Likely Likely Likely (b) Otherwise (1) Complete bypass Impossible Impossible Impossible k

ca (ii) Partial bypass Impossible Impossible Impossible (iii) No bypass Certain Certain Certain

12. Location of induced S3 initiator seal (1) No induced failure 0.74 0.74 0.74 failure in primary cooling available, steam generators wet (ii) Small induced steam Impossible Impossible Impossible generator tube rupture (iii) Large induced steam Impossible Impossible Impossible generator tube rupture (iv) Large induced hot leg LOCA 0.26 0.26 0.26 (v) Small cold leg LOCA Impossible Impossible Impossible (vi) Large cold leg LOCA Impossible Impossible Impossible

13. Size of induced (a) Failure in cold leg (1) Equivalent to S2 break Indeterm. Indeterm. Indeterm.

primary system or tube rupture failure (ii) Equivalent to Si break Indeterm. Indeterm. Indeterm.

(b) Failure in hot leg (1) Equivalent to large break Certain Certain Certain (ii) Other break sizes Impossible Impossible Impossible

TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

14. Early containment bypass (a) Induced tube rupture (1) No bypass Impossible Impossible Impossible during meltdown (ii) Small bypass Certain Certain Certain (iii) Large bypass Impossible Impossible Impossible (b) Otherwise (1) No bypass Certain Certain Certain (ii) Small bypass Impossible Impossible Impossible (iii) Large bypass Impossible Impossible Impossible

15. Hydrogen igniters operate None (1) Igniters operate 1.0 0.997 0.970 early (ii) Igniters do not operate 0.0 0.003 0.03

x. 16. Lower compartment inert Air return fans on (1) Lower compartment inert Certain Certain Certain n

(ii) Lower compartment not Impossible Impossible Impossible inert

18. Local detonation in ice Air return fans on (i) No damage from early Certain Certain Certain condenser detonation (ii) Ice condenser function Impossible Impossible Impossible fails due to detonation (iii) Containment fails due Impossible Impossible Impossible detonation

19. Deflagration in lower Lower compartment inert (1) No early deflagration Certain Certain Certain compartment before vessel breach (ii) Early deflagration Impossible Impossible Impossible

20. Deflagration in upper (a) Large induced LOCA, (1) No upper compartment Certain Certain Impossible compartment igniters on deflagration (ii) Upper compartment Impossible Impossible Certain

~ - .- __. _ - - -.- - - _ _ _ - . .. . - - - . - . . . .. - _ - _ - - - - , -_ . - . - . - - - - - - - - - - - - - - - - - - - - - -

1 TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD i EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC l

20. (Continued) (b) No induced LOCA, (1) No upper compartment Certain Certain Certain

. deflagration (ii) Upper compartment Impossible Impossible Impossible deflagration (c) Large induced LOCA, (1) No upper compartment Certain Certain Unlikely igniters not on deflagration

(ii) Upper compartment Impossible Impossible Likely a

deflagration T

21. Centainment failure due Calculated from containment to early deflagration pressure loading and capacity (see Tables 2.4 and 2.5) '

{ b w

22. Bypass of auxiliary (a) Containment failure (1) No bypass Impossible Impossible Impossible .

building early due to deflagration

(ii) Partial bypass Indeterminate Unlikely Remotely Possible 4

l (iii) Complete bypass Indeterminate Likely Almost Certain  ;

(b) Containment failure (1) No bypass Likely 0.25 Impossible.

due to detonation in ice condenser (ii) Partial bypass Unlikely 0.25 Remotely .;

Possible (iii) Complete bypass Impossible Indeterminate -Almost ,

certain (c) Containment failure (1) No bypass Impossible Impossible Impossible due to detonation

, in ice condenser, (ii) Partial bypass Certain Indeterminate Remotely prior partial bypass Possible (iii) Complete bypass Impossible Indeterminate Almost Certain i

(d) Otherwise (1) No bypass Certain Certain. Certain 4

3

TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOCDS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

22. (Continued) (ii) Partial bypass Impossible Impossible Impossible (iii) Complete bypass Impossible Impossible Impossible

23. Ice melted or bypassed (a) Early failure of (1) No failure of ice Impossible Impossible Impossible before vessel breach ice condenser due to condenser detonation (11) Bypass path established Certain Certain certain through ice condenser (iii) Ice melted Impossible Impossible Impossible (b) Induced large LOCA (1) No failure of ice Certain Indeterminate Unlikely condenser (ii) Bypass path established Impossible Indeterminate Likely through ice condenser s

os (iii) Ice melted Impossible Impossible Unlikely (c) Induced S2 LOCA (1) No failure of ice Certain Almost Likely condenser Certain (ii) Bypass path established Impossible Remotely Unlikely through ice condenser Possible (iii) Ice melted Impossible Impossible Impossible (d) Otherwise (1) No failure of ice Certain Certain Almost condenser Certain (ii) Bypass path established Impossible Impossible Remotely through ice condenser Possible (iii) Ice melted Impossible Impossible Impossible

24. Status of air return (a) Upper compartment (i) Fans operating following Certain Certain Almost fans after hydrogen burn deflagration deflagration certain (ii) Fans failed following Impossible Impossible Remotely deflagration Possible

TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

24. (Continued) (b) No upper compartment (1) Fans operating Certain Certain Certain deflagration (ii) Fans failed following Impossible Impossible Impossible deflagration

27. Mode of reactor vessel (a) No induced LOCA, or (1) Steam explosion failing Impossible 0.0001 0.1 breach induced S2 LOCA upper head (n-mode)

(ii) Steam explosion failing Impossible 0.21 0.47 bottom head (iii) Pressurized ejection Certain 0.79 0.43 (iv) Meltthrough Impossible Impossible Impossible (b) Induced large LOCA (1) Steam explosion failing Impossible 0.0001 0.1

-a upper head (n-mode)

(ii) Steam explosion failing Impossible 0.21 0.47 bottom head -

(iii) Pressurized ejection Impossible Impossible Impossible (iv) Meltthrough Certain 0.79 0.43

28. Water in reactor cavity successful ECC and spray (1) Reactor cavity wet Certain Certain Certain at vessel breach injection (ii) Reactor cavity dry Impossible Impossible Impossible

29. Direct heating at vessel Reactor cavity wet (1) Direct heating Impossible Impossible Impossible breach (ii) No direct heating Certain Certain Certain

30. Detonation in ice condenser (a) Fans failed due to (1) No damage from detonation Certain 0.8 Impossible after vessel breach early deflagration (11) Detonation causes Impossible Unlikely Indeterminate failure of ice condenser (iii) Upper containment failed Impossible Unlikely Indeterminate due to detonation

TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

30. (Continued) (b) Fans operating (1) No damage from detonation Certain Certain Certain (ii) Detonation causes Impossible Impossible Impossible failure of ice condenser (iii) Upper containment failed Impossible Impossible Impossible due to detonation

31. Ice melted or bypass (a) Detonation fails ice (1) No failure of ice Impossible Impossible Impossible paths established less condenser function condenser than 3 hr after vessel breach (ii) Bypass path established Certain certain certain through ice condenser (iii) Ice melted Impossible Impossible Impossible (b) Induced large LOCA, (1) No failure of ice Likely Indeterminate Impossible g, no previous failure condenser os of ice condenser (ii) Bypass path established Unlikely Indetermin. Indetermin.

through ice condenser (iii) Ice melted Remotely Unlikely Indeterminate Possible (c) Induced S2 LOCA, (1) No failure of ice Almost Likely 0.4 no previous failure condenser certain of ice condenser (ii) Bypass paths established Remotely Unlikely Indeterminate through ice condenser Possible (iii) Ice melted Impossible Remotely Unlikely Possible (d) Induced large LOCA, (i) No failure of Impossible Impossible Impossible early bypass of ice ice condenser condenser (ii) Bypass paths established Almost Likely Indeterminate through ice condenser certain (iii) Ice melted Remotely Unlikely Indeterminate Possible

TABLE 2.3 EVENT DESCRIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

31. (Continued) (e) Induced S 2 LOCA, (1) No failure of Impossible Impossible Impossible early bypass of ice ice condenser condenser (ii) Bypass paths established Certain Almost Likely through ice condenser certain (iii) Ice melted Impossible Remotely Unlikely Possible (f) No induced LOCA, (1) No failure of Impossible Impossible Impossible early bypass of ice ice condenser condenser (ii) Bypass paths established Certain Certain Certain through ice condenser (iii) Ice melted Impossible Impossible Impossible em (g) Otherwise (1) No failure of

4) Certain Likely Indete rminate ice condenser (ii) Bypass paths established Impossible Unlikely Indeterminate through ice condenser (iii) Ice melted Impossible Impossible Impossible

32. Hydrogen burn at vessel (a) No prior burning, (1) No deflagration at Impossible Impossible Impossib?.e breach igniters available vesse. breach (ii) Deflagration at vessel Certain Certain Certain breach (b) No prior burning, (1) No deflagration at Unlikely Unlikely Unlikely igniters failed vessel breach (ii) Deflagration at vessel Likely Likely Likely breach (c) No large induced (1) No deflagration at Impossible Impossible Impossible LOCA, some prior vessel breach burning, igniters on (ii) Deflagration at vessel Certain Certain Certain breach

TABLE 2.3 EVENT OESCRIPTICNS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

32. (Continued) (d) No large induced (1) No deflagration at Unlikely Unlikely Unlikely LOCA, some prior vessel breach burning, igniters off (11) Deflagration at vessel Likely Likely Likely breach

33. Containnent failure at calculated from containment vessel breach pressure loading and capacity (see Tables 2.4 and 2.5)

35. Bypass of auxiliary (a) Steam explosion (1) No bypass Impossible Impossible Impossible building a-mode failure

,4 (ii) Partial bypass Impossible Impossible Impossible (iii) Complete bypass Certain Certain Certain LA (b) Containment failure (1) No bypass Impossible Impossible Impossible

~

C) at vessel breach

• other than due to (11) Partial bypass Indeterminate Unlikely Remotely hydrogen detonation Possible or a-mode (iii) Complete bypass Indeterminate Likely Almost Certain (c) Detonation at (1) No bypass Likely 0.25 Impossible vessel breach, no prior bypass (ii) Partial bypass Unlikely 0.25 Remotely Possible (iii) Complete bypass- Impossible Indeterminate Almost Certain (d) Detonation at vessel (1) No bypass Impossible Impossible Impossible breach, prior partial bypass (ii) Partial bypass Certain Ir. determinate Remotely Possible (iii) Complete bypass Impossible Indeterminate Almost

_ , ,Certain (e) Otherwise (1) No bypass Certain Certain Cert'ain (ii) Partial bypass _

Impossible Impossible Impossible

?

u.

, TABLE 2.3 EVEMT DESCRIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD EVENT PRIOR EVENTS BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

35. (Continued) (iii) Complete bypass Impossible Impossible Impossible

36. Air return fans on (a) Hydrogen burn at (1) Fans operating following Certain Almost Likely after vessel breach vessel breach deflagration Certain (11) Fans failed following Impossible Remotely Unlikely deflagration Possible (b) No deflagration at (i) Fans operating Certain Certain Certain vessel breach (ii) Fans failed following Impossible Impossible Impossible deflagration

39. Igniters on late (a) Igniters not (1) Igniters actuated late Indeterminate Unlikely Impossible initiated early tn (ii) Igniters not on late Indeterminate Likely Certain H

(b) Otherwise (1) Igniters on late Certain Certain Certain (ii) Igniters not on late Impossible Impossible Impossible

43. Coolable debris bed (a) Vessel breach by (1) Coolable debris bed Likely Unlikely Remotely meltthrough Possible (ii) No coolable debris bed Unlikely Likely Almost Certain (b) Otherwise (1) Coolable debris bed Almost Likely Indeterminate Certain (ii) No coolable debris bed Remotely Unlikely Indeterminate Possible

44. Late hydrogen burn (a) No prior failure, (1) No late hydrogen burn Unlikely Unlikely Unlikely igniters off, no coolable debris bed (ii) Late hydrogen burn Likely Likely Likely (b) No prior failure, (1) No late hydrogen burn Impossible Impossible Impossible prior burning, igniters on late (ii) Late hydrogen burn Certain Certain Certain

.._-.,c . ~ . - - . . - . - . - . ~_. , .- . - , . . . - .- -

TABLE 2.3 EVENT DESCPIPTIONS AND LIKELIHOODS FOR SEQUENCE3 S HF (Continued)

LIKELIHOOD

. EVENT PRIOR EVENTS- BRANCHING OPTIONS OPTIMISTIC CENTRAL PESSIMISTIC

44. (Continued) (c) Otherwise (1) No late hydrogen burn Certain Certain Certain .g (ii) Late hydrogen burn Impossible Impossible. Impossible 2 .,

47. Late containment Sprays not available (1) Late containment failure Certain certain certain' .-c, . e-failure due to steam or non-condensible gases (ii) No late failure Impossible Impossible Impossible

48. Auxiliary building Late hydrogen burn or (i) Partial bypass Indeterminate Unlikely ' Remotely bypass late late overpressurization Possible l

(14) Total bypass Indeterminate Likely Almost

. Certain-

49. Basemat me).tthrough Sprays not available (1) Meltthrough certain Certain certain b (ii) No meltthrough- Impossible Impossible Impossible s

W.

1 I

k 1

J s

\

W W T-

TABLE 2.4 CONTAINMENT LOADINGS FOR SEQUENCE S3 HF PRESSURES (PSIG)

SOURCE OF PRESSURE PRIOR EVENTS OPTIMISTIC CENTRAL PESSIMISTIC

1. Containment pressure before (a) Containment air return f ans 18 19 20 vessel breach operating (b) Containment air return fans 19 21 23 not operating l

l

2. Containment pressure increment (a) Large or intermediate LOCA, with 1 5 25 due to hydrogen burn in the no containment bypass, igniters

, lower containment before operating, and lower compartment vessel breach not inert l (b) Large or intermediate LOCA, with 0 25 37 I

no containment bypass, igniters not j operating, and lower compartment i not inert (c) Small LOCA, with no containment 1 4 6 bypass, igniters operating, and lower compartment not inert (d) Small LOCA, with no containment 0 6 8 bypass, igniters not operating, and lower compartment not inert (e) Otherwise 0 0 0

3. Containment pressure increment (a) Air return fans do not operate 0 0 0 due to hydrogen burn in the and no lower compartment upper containment before deflagrations occur vessel breach a

('? Large or intermediate LOCA, with 0 0 25 nv containment bypass, ig-iters operating (c) Large or intermediate LOCA, with 0 0 32 no containment bypass, igniters not operating (d) Small LOCA 0 0 0 (e) Otherwise 0 0 0

4. Total contt Lnment pressure Sum of pressures in 1, 2 and 3 following early hydrogen burn

5. Containment pressure increment (a) High RCS pressure (no induced 37 37 37 due to blowdown from reactor failure), ice condenser function coolant system defeated due to bypass or melting (b) High RCS pressure (no induced 5 5 5 failure), ice condenser effective and air return fans operating (c) Intermediate RCS pressure (induced 19 19 19 S2 LOCA), ice condenser defeated (d) Intermediate RCS pressure, ice 5 5 5 condenser effective and air return fans operating (e) Low RCS pressure (inouced large 0 0 0 LOCA)

(f) Effective ice condenser, air 0 0 0 return fans not operating

6. Pressure increment due to Loss of ice condenser function 0 20 34 steam spike at vessel breach 53

TABLE 2.4 CONTAINMENT LOADINGS FOR SEQUENCE3 S HF (Continued)

PRESSURES (PSIG)

SOURCE OF PRESSURE PRIOR EVENTS OPTIMISTIC CENTRAL PESSIMISTIC

7. Containment pressure due to (a) High RCS pressure prior to 1 31 100 combined steam spike and direct vessel breach heating, given loss of ice condenser function (b) Intermediate RCS pressure 0 15 50 (c) Low RCS pressure 0 0 0

8. Containment pressure increment (a) Prior burning of hydrogen 1 20 45 due to hydrogen burn at vessel breach (b) No prior burning, wet reactor 5 30 100 cavity (c) No prior burning, dry reactor 5 60 100 cavity

9. Containment pressure following Sum of pressures in 5,6,7 and 8 vessel breach

10. Containment baseline pressure (a) Containment heat removal and 18 19 20 late air return fans operable (b) Containment sprays operating, but 19 21 23 no air return fans (c) No containment heat removal 20 30 40 (d) Ico condenser functions, but 20 25 35 no containment sprays or air return fans

11. Containment pressure (a) No igniters operating 5 35 185 increment due to late hydrogen burn (b) Igniters operating, no prior burns 5 25 55 (c) Igniters operating, prior burns 5 18 25

12. Late containment pressure Sum of pressures in 10 and 11 i

54 l \

l

kl Table 2.5 SEQUOYAH CONTAIdMENT CAPACITY ESTIMATES (a) Leakage Area Versus Pressure PRESSURE (psig) LEAK AREA (in2 )

Optimistic Central Pessimistic 0 .004 .004 .004 12 .004 .004 .004 (linear 50 .004 .004 .4 variation)

(b) Structural Failure Pressure PRESSURE (psig)

Optimistic Central Pessimistic Mean Structural Failure Pressure 60 50 36 Standard Deviation 7.5 6 6 55

Table 2.6 BASIS FOR PRESSURE ESTIMATES FOR CONCURRENT DIRECT HEATING / STEAM SPIKE optimistic Central Pessimistic (1) Percent of Core Malt 50% 75% 100%

. Ejected at the Time of Vessel Breach (2) Percent of Amount in 80% 75% 50%

(1) Quenched Thermally in Water (3) Percent of Amount in 0% 30% 30%

(2) which Reacts with Steam 4

(4) Percent of Amount in 2% 25% 50%

(1) Quenched Thermally in Atmosphere (5) Percent of Amount in 0% 50% 0%

(4) which Reacts with Steam (6) Percent of Amount in 0% 0% 50%

(4) which Reacts with Oxygen 4

I l

56

TABLE 2.7. ALTERNATIVE ASSIGNMENT OF VALUES TO VERBAL DESCRIPTORS LIKELIHOOD VERBAL ALT. 1 ALT. 2 ALT. 3 ALT. 4 DESCRIPTOR (BASE CASE) certain or 1.0 1.0 1.0 1.0 Almost Certain Likely 0.9 0.9 0.9 0.9 Indeterminate 0.5 0.5 0.5 0.5 Unlikely 0.1 0.01 0.01 0.1 l

i Remotely Possible 0.001 0.001 0.0001 0.01 Impossible 0 0 0 0 i

57

- . - . , - - ~.

1 Table

2.8 DESCRIPTION

OF PLANT DAMAGE STATES Designation Description

]

l

.V(WET) LOCA outside containment, with the break location submerged. l t

-V(DRY) -LOCA outside containment, with the break location not submergod. I AIYBY LOCA leading to low pressure in the RCS prior to vessel breach, with i core melt due to failure of emergency coolant recirculation,

'; containment heat removal provided by operation of the containment sprays,'and ac power available.

AINBY Same as'AIYBY, but with containment heat removal unavailable.

AINIY Large LOCA, .with success of both core cooling and containment sprays in the injection phase, but failure of both in recirculation (ac power available).

ANYBY Large LOCA with failure of core cooling in the injection phase, ~ but with containment' heat removal available via the spray system.

ABYNB Large LOCA with failure of emergency core cooling systems, but eventual overpressurization of containment due to the absence of containment heat removal, causing loss of core cooling.

i S2IYBY Small LOCA leading to intermediate pressure prior ot vessel breach, J

with emergency core cooling succeeding in injection but'failing in

, recirculation, and with containment heat removal provided by the spray i system.

S2NNNY Small LOCA with failure of both emergency core cooling and containment sprays in the injection phase, but with ac power. available.

S2NNNN Small LOCA with failure of core cooling and containment sprays due to unavailability of ac power.

S2INIY Small LOCA with success of both emergency core cooling and containment sprays in injection, but failure of both in recirculation.

S2NYBY Small LOCA with failure of emergency coolant injection but success of containment sprays and containment heat removal.

S3IYBYB Very small LOCA, with RCS pressure. remaining high prior to vessel breach, and with emergency core cooling failing in the recirculation phase but containment sprays and containment heat removal succeeding,.

and with ac power, RCP seal cooling, and steam generator feedwater available.

S3INIYB Very small LOCA with both core cooling and containment sprays failing l in the recirculation phase, and with ac power, RCP seal cooling, and steam generator feedwater available.

4~

S3INBYB Very small LOCA with core cooling failing in recirculation but with containment heat removal provided by the spray system and with ac power, RCP seal cooling, and steam generator feedwater available.

TNNNNN Transient with cycling relief valves and with core cooling, i containment sprays, and RCP seal cooling unavailable due to failure:of l ac power, and with the steam generators dry.

TNYBYR Transient with loss of core cooling, including loss of feedwater, but with containment heat removal, RCP seal cooling, and ac power available.

TNYBYB. Transient with loss of core cooling in injection but with success of containment heat removal, RCP seal cooling, steam generator feedwater, and ac power.

58

Table 3.1 RESULTS FOR SEQUOYAH S2 H2 (Small Pipe Break LOCA with Failure of Emergency Core Cooling Recirculation)

I '. SEQUENCE FREQUENCY OPT. CENTRAL PESS.

ASEP Baseline value 2x10-6 II. CONTAINHENT FAILURE MODE PROBABILITIES OPT. CENTRAL PESS.

CASE 1* CASE 1* l CASE 2*

l CASE 2*

No Containment Failure 1.0 .79 .79 .05 .04 Basemat Meltthrough Only .001 .009 .009 -- --

Late Overpressurization (Rupture) .001 .18 .18 .08 .07 Late Hydrogen Burn -- -- -- .10 .09 Early Steam Spike and/or Vessel Depressurization --

Early Steam Spike /Depres./H2 Burn -- .02 .02 .77 .70 Early Steam Spike /Depres./H 2Burn / Direct Heating * -- -- -- -- --

Hydrogen Combustion Before Vessel Breach -- --

In-Vessel Steam Explosion (Missile Breach) -- -- -- -- .10 Large Isolation Failure or Preexisting Leak -- .002 .002 .004 .004 1.00 1.00 T!UU- 1.00 1.00 Fraction of Above Affected by (a) Preexisting Small Containment Leak +* .02 .03 .03 .06 .06 (b) Induced Small Steam Generator Tube Leak ** -- -- -- -- --

(c) Ice Melting / Bypass Before Vessel Breach -- .001 .10 .10 .10 (d) Ice Melting / Bypass Before Core-Concrete Interaction .001 .10 .10 .54 .54 III. FRINCIPAL CONTAINMENT PATHWAYS Generale Initial RCS failure occurs in hot or cold leg with comparable frequency. Induced RCS f ailures do not occur. The vessel usually fails by pressure ejection into a water filled cavity.

The auxiliary building is nearly always bypassed after containment failure. Oxidation and vaporization releases rarely occur.

l BMI-2104 CALCULATION (1) No Containment Failures Containment sprays survive and a No BMI-2104 calculations coolable debris bed is formed. were made for S H22 Releases from the BCS should be similar to (2) Late overpressurization: Either the sprays fail after vessel those for S 2H2 F.

breach and containment failure occurs due to steam buildup or Reductions in source the debris is not coolable and containment failure occurs due should be similar to to non-condensible gas accumulation. those achieved in TML which was calculated.

(3) Early Steam Spike / Hydrogen Burn Containment fails at the time of vessel breach due to hydrogen combustion and the sprays usually (C) or seldom (P) survive.

• Case 2 for central and pessimistic walkthroughs includes the effects of in-vessel steam explosions and ex-vessel direct heating. Case 1 does not include these effects. Direct heating includes hydrogen burning if the atmosphere is flammable. See text for a discussion of limitations associated with the results for steam explosions and direct heating.

• Small containment leaks include those producing leakage greater than containment design but not large enough to preclude gradual overpressurization from steam and/or noncondensibles.

Likewise, small steam generator tube leaks include those producing primary-to-secondary leakage greater than design but not large enough to depressurize the primary system below the accumulator setpoint before vessel breach.

59

Table 3.2 RESULTS FOR SEQUOYAH S2H2P (Small Pipe Break LOCA with Failure of Emergency Core Cooling Recirculation and containment Spray Recirculation)

1. SEQUENCE FREQUENCY OPT. CENTRAL PESS.

ASEP Baseline Value 2x10-7 II. CONTAINMENT FAILURE MODE PROBABILITIES PFSS.

OPT. CENT RA L CASE 1*

CASE 1* l CASE 2* l CASE 2*

tio containment Failure -- -- -- -- --

Basemat Meltthrough Only Late overpressurization (Rupture) 1.00 .98 .98 .07 .06

-- -- -- .15 .14 Late Hydrogen Burn -- -- --

Early Steam Spike and/or Vessel Depressurization -- --

.02 .02 .77 .70 Early Steam Spike /Depres./H2 Burn --

Early Steam Spike /Depres./H2 Burn / Direct Heating. -- --

Hydrogen Combustion Before Vessel Breach -- --

.10 In-Vessel Steam Explosion (Missile Breach) -- -- -- --

.004 .004 Large Isolation Failure or Preexisting Leak -- .002 .002 1.00 1.00 1.00 1.00 1.00 Fraction of Above Affected by:

(a) Preexisting Small Containment Leak ** .02 .03 .03 .06 .06 (b) Induced Small Steam Generator Tube Leak * *

(c) Ice Melting / Bypass Before Vessel Breach -- .001 .001 .10 .10 (d) Ice Melting / Bypass Before Core-Conrete Interactions .031 .10 .10 .54 .54 III. PRINCIPAL CONTAINMENT PATHWAYS Generals Initial RCS failure occurs in hot or cold leg with comparable frequency. Induced RCS failure!

JL not occur. The vessel usually fails by pressure ejection into a water filled cavity. Sprays are defined to be failed in initiating sequencer no credit is given for recovery. A coolable debris bed usually forms. The auxiliary building is usually bypassed. Oxidation and vaporization releases rarely occur.

lBMI-2104 CALCULATION (1) Late Overpressurization: No hydrogen turns occur which No, but there was a calculati challenge containment. A coolable debris bed forms but the for late containment failure sprays are inoperable and the containment fails due to steam due to a hydrogen burn which buildup, may be applicable here.

(2) Ear t_y_St eam Spike /Hy_dr ogen Barn: A large hydrogen burn occurs at vessei breach Talling containment. The sprays have been inoperable since the beginning of the accident. The debris is sometimes (C) or usually (P) not coolable.

• Case 2 for central and pessimistic walkthroughs includes the ef fects of in-vessel steam explosiur.s and ex-vessel direct heating. Case 1 does not include these effects. Direct heating includes hydrogen burning if the atmosphere is flammable. See text for a discussion L tattations associated with the results for steam explosions and direct heating.

• • Small containment leaks include those producing leakage greater than containment design but not large enough to preclude gradual overpressurization f rom steam and/or noncondensibles.

Likewise, small steam generator tube leaks include those producing primary-to-secondary leakage greater than design but not large enough to depressurize the primary system below the accumulator setpoint before vessel breach.

l i 60

Table 3.3 RESULTS FOR SEQUOYAH S3 H2 (Reactor Coolant Pump Seal LOCA with Failure of Emergency Core Cooling Recirculation)

I. SEQUENCE FREQUENCY OPT. CENTRAL PESS.

ASEP Baseline Value 3x10-5 II. CONTAINMENT FAILURE MODE PROBABILITIES OPT. CENTPAL PESS.

CASE 1*

l CASE 2* CASEl'l CASE 2*

No Containment Failure .97 .70 .70 .04 .03 Basemat Meltthrough only .03 .02 .02 -- --

Late Overpressurization (Rupture) .001 .23 .23 .07 .07 Late Hydrogen Burn -- -- --

.07 .07 Early Steam Spike and/or Vessel Depressurization -- -- -- -- --

Early Steam Spike /Depres./H2 Burn -- .05 .05 .55 .50 Early Steam Spike /Depres./H2 Burn / Direct Heating. -- -- -- -- --

Hydrogen Combustion Before Vessel Breach -- -- -- .26 .26 In-Vessel Steam Explosion (Missile Breach) -- -- -- --

.07 Large Isolation Failure or Preexisting Leak --

.002 .002 .004 .004 N T.35~ T'UU~ N 1.00 Fraction of Above Affected by:

(a) Preexisting small Containment Leak ** .02 .03 .03 .06 .06 (b) Induced Small Steam Generator Tube Leak ** -- -- -- -- --

(c) Ice Melting / Bypass Before vessel Breach -- .13 .13 .23 .23 (d) Ice Melting / Bypass Before Core-Conrete Interactions .03 .35 .15 .40 .40 III. PRINCIPAL CONTAINMENT PATHWAYS General: Initial RCS failure occurs in cold leg. Induced RCS f ailures may occurr if they do. ice condenser bypass paths never (0), sometimes (C), and usually (P) develop before core degradation. The auxiliary building is nearly always bypassed after containment failure. Oxidation and vaporization releases rarely occur.

lBMI-2104 CALCULATION No BMI-2104 calculations were

~

(1) No Containment Failures Containment sprays survive and a coolable debris bed is formed. made for S H32 Releases from the RCS should be similar to (2) Late overpressurization Either the sprays fail af ter those for S 2H. Reductions vessel breach and containment failure occurs due to in source term due to containment steam buildup or the debris is not coolable and con- should be siellar to those achieved tainment failure occurs due to noncondensible gas accumu- in TML which was calculated.

lation.

(3) Early Steam Spike / Hydrogen Burn: Ice melting or bypass paths develop prior to or during the time of vessel breach. Containment fails at the time of vessel breach due to hydrogen combustion and the sprays usually (C) or seldom (P) survive.

• Case 2 for central and pessimistic walkthroughs includes the effects of in-vessel steam explosions and ex-vessel direct heating. Case 1 does not include these effects. Direct heating includes hydrogen burning if the atmosphere is flammable. See text for a discussion of limitations associated with the results for steam explosions and direct heating.

• • Small containment leaks include those producing leakage greater than containment design but not large enough to preclude gradual overpressurization f rom steam and/or noncondensibles.

Likewise, small steam generator tube leaks include those producing primary-to-secondary leakage greater than design but not large enough to depressurize the primary system below the accumulator setpoint before vessel breach.

61

Table 3.4 RESULTS FOR SEQUOYAH S32 HP (Reactor Coolant Pump Seal LOCA with Failure of Emerger :y Core Cooling Recirculation and Containment Spray Recirculation)

I. SEQUENCE FREQUENCY OPT. CENTRAL PESS.

ASEP Baseline Value 3x10-7~

II. CONTAINMENT FAILURE MODE PROBABILITIES OPT. CENTPAL PESS.

CASE 1* l CASE 2* CASE 1* l CASE 2*

No Containment Failure --

Basemat Meltthrough Only Late Overpressurization (Rupture) 1.00 .94 .94 .07 .06 Late Hydrogen Burn -- .001 -- .11 .10 Early Steam Spike and/or Vessel Depressurization -- -- -- -- --

Larly Steam Spike /Depres./Hy Burn - .05 .05 .55 .50 Early Steam Spike /Depres./H2 Burn / Direct Heating. -- -- -- -- --

Hydrogen combustion Before vessel Breach -- -- -- .26 .26 In-Vessel Steam Explosion (Missile Breach) -- -- -- -- .07 Large Isolation Failure or Preexisting Leak -- .002 .002 .004 .004 1.00 1.00 1.00 1.00 1.00 fraction of Above Affected by:

(a) Preextsting Small Containment Leak ** .02 .03 .03 .06 .06 (b) Induced Small Steam Generator Tube Leak ** -- -- -- --

i (c) Ice Melting / Bypass Before Vessel Dreach -- .13 .13 .23 .23 (d) Ice Melting /Dypass Before Core-Conrete Interactions .03 .15 .15 .40 .40 111. PRINCIPAL CONTAINMENT PATHWAYS Generate Initial RCS failure occurs in cold leg. Induced PCS failures may occur, and ice condenser bypass paths never (O). sometimes (C). and usually (P) develop before core degradation. Sprays are defined to be failed in initiating sequence. no credit is given for recovery. A coolable debris bed usually forms. The auxiliary building is usually bypassed. Oxidation and vaporization releases rarely occur.

lBMI-2104 CALCULATION (1) Late Overpressurization: No hydrogen burns occur which No, but there was a calculation cEaliensk containment. A coolable debris bed forms but the for late containment failure sprays are inoperable and the containment f ails due to steam due to a hydrogen burn which butLJup. may be applicable here.

(2) Early Steam Spige/ Hydrogen Burn Ice melting or bypass paths No develop prior to or during the time of vessel breach. A large hydrogen bury occurs at vessel breach failing containment. The sprays have been inoperable since the beginning of the accident.

The debris is sometimes (C) or usually (P) not coolable.

• Case 2 for central and pessimistic walkthroughs includes the effects of in-vessel steam explosions and ex-vessel direct heating. Case 1 does not include these effects. Direct heating includes hydrogen burning if the atmosphere is flammable. See text for a discussion of limitations associated with the results for steam explosions and direct heating.

• • Small containment teaks include those producing leakage greater than containment design but not large enough to preclude gradual overpreisurization from steam and/or noncondensibles.

Likewise, small steam generator tube leaks include those producing primary-to-secondary leakage greater than design but not large enough to depressurize the primary system below the accumulator setpoint before vessel breach.

62

Table 3.5 RESULTS FOR SEQUOYAH S3D (Reactor Coolant Pump Seal LOCA with Failure of Emergency Core Cooling Injection)

I. SEQUENCE FREQUENCY OPT. CENTRAL PESS.

ASEP Baseline Value < 10-7 II. CONTAINMENT FAILURE MODE PROBABILITIES OPT. CENTRAL PESS.

CASE 1* l CASE 2* CASE 1* l CASE 2*

No Containment Failure .97 72 .72 .04 .03 Basemat Meltthrough .03 02 .02 -- --

Late Overpressurization (Rupture) .001 .24 .24 .07 .07 Late Hydrogen Burn -- -- -- .07 .07 Early Steam Spike and/or Vessel Depressurization -- -- -- -- --

Early Steam Spike /Depres./H 2Burn -- .01 .01 .55 .50 Early Steam Spike /Depres./H 2Burn / Direct Heating * -- -- -- -- --

Early Hydrogen Burn -- -- -- .26 .25 In-Vessel Steam Explosion (Missile Breach) -- -- -- -- .07 Large Isolation Failure or Preexisting Leak -- .002 .002 .004 .004 N N T UD~ T TU- T 60-Fraction of Above Affected by (a) Preexisting Small Containment Leak ' * .02 .03 .03 .06 .06 (b) Induced Small Steam Generator Tube Leak +* -- .001 .001 .10 .10 (c) Ice Melting / Bypass Before Vessel Breach -- -- -- .001 .001 (d) Ice Melting / Bypass Before Core-Concrete Interactions --

.10 .10 .51 .51 III. PRINCIPAL CONTAINMENT PATHWAYS Generals Initial RCS failure is a pump seal in the cold leg. Induced RCS failure may occur in hot (C) and cold (O&C) legs, or a worsened seal LOCA or small SGTR (P) may occur. Vessel failure usually occurs with the RCS at high or intermediate pressure. The ice condenser is very seldom '

bypassed or void of ice until late. A cootable debris bed forms usually (0), sometimes (C), or very seldom (P). The auxiliary bu!1 ding is nearly always bypassed after containment failure.

Oxidation and vaporization releases almost never occur.

l BMI-2104 CALCULATION (1) No Containment Failures Sprays survive throughout the accident No BMI-2104 calculations and matntain the debrTo coolable. were made for S3D, but source terms for TML (2) Basemat Meltthrough Sprays survive but the debris is not may apply.

coolable, concrete attack occurs and produces meltthrough.

(J) Late Overpressurization: Sprays usually survive, the debris is not coolable, concrete attack proceeds, and containment fails due to non-condensible gas accumulation.

(4) Early Steam Spike / Hydrogen Burns Sprays fail at the time of vessel breach, the deEris is not coolable.

• Case 2 for central and pessimistic walkthroughs includes the effects of in-vessel steam explosions and ex-vessel direct heating. Case 1 does not include these effects. Direct heating includes hydrogen burning if the atmosphere is flammable. See text for a discussion of limitations associated with the results for steam explosions and direct heating.

• • Small containment leaks include those producing leakage greater than containment design but not large enough to preclude gradual overpressurization from steam and/or noncondensibles.

Likewise, small stear generator tube leaks include those producing primary-to-secondary leakage greater than design but not large enough to depressurize the primary system below the accumulator setpoint before vessel breach.

63

T ble 3.6 RESULTS FOR SEQUOYAH TI LIDgF (Station Blackout, Including Loss of Aux 111ery Fesdwstsr)

I. SEQUENCE FREQUENCY OPT. CEN"'RAL PESS.

ASEP Baseline Value 9x10-6

11. CottTAINMENT FAILURE MODE PROBABILITIES OPT. CFNTPAL PFSS.

CASE 1* l CASE 2* CASE 1* l CASE 2*

No Containment Failure .68 .02 .02 -- --

Basemat Meltthrough Only .07 .01 .01 -- --

Late Overpressurization (Rupture) .25 .14 13 .001 .001 Late Hydrogen Burn -- .01 01 .01 .02 Early Steam Spike and/or vessel Depressurization --

.006 -- .04 .005 Early Steam Spike /Depres./H 2Burn -- 71 .66 .44 .23 Early Steam Spike /Depres./H 2Burn / Direct Heating * -- -- .07 -- .19 Hydrogen Combustion Before Vessel Breach -

.10 .10 .50 .50 In-Vessel Steam Explosion (Missile Breach) -- -- -- -- .05 Large Isolation Failure or Preexisting Leak -- .002 .002 .004 .004 1.00 EUU T.T5 1.00 1.00 Fraction of Above Affected by:

(a) Preexisting Small Containment Leak ** .02 .03 .03 .06 .06 (b) Induced Small Steam Gene. at Tube Leak'* .14 .14 .14 .14 4 (c) Ice Melting / Bypass Befors essel Breach -- .15 .15 .55 .55 (d) Ice Melting / Bypass Before Core-Concrete Interactions -- .44 .44 .40 .40 III. PRINCIPAL CONTAINMENT PATHWAYS Generals Initial loss of coolant is through the PORV. Induced RCS failure in hot (C) and cold (06C) legs or small SGTR (P) may occur. Vessel failure usually occurs with the PCS at high or intermediate pressure. A coolable debris bed forms usually (O), sometimes (C), or very seldom (P). The auxiliary building is nearly always bypassed after containment failure. Oxidation and vaporization releases rarely (0), sometimes (C), or usually (P) occur.

l BMI-2104 CALCULATION (1) No Containment Failures A coolable debris bed is formed. No, but releases would The sprays are opera'ble after late power restoration. be minimal.

(2) Basemat Meltthroughs A coolable debris bed is not formed, No, but should be time sprays are operable af ter late power restoration preventing siellar to any basemat overpressure. meltthrough case.

(J) Late Overpressurization: Sprays are inoperable (0) or operable Calculation corresponds (C) after late power restoration. Containment f ailure occurs to central case with no due to steam and/or non-condensible gas buildup, sprays. Other calcula-tions are available which address ef fects of sprays late in accident.

(4) Late Hydrogen Burn e Containment fails due to a late hydrogen burn. No, but should be The sprays do not operate. similar to late over-pressurtration case with no sprays.

(5) Early Steam Spike / Hydrogen Burne The sprays do (C) or do not (P) Yes for case with no operate. The debris is not coolable. The auxiliary building sprays late. Other is usually bypassed. calculations available which address effects of late sprays.

(6) Early Steam Spike / Hydrogen Burn / Direct Heatings The sprays do No, calculations (C) or do not (P) operate. The debris is not coolable. The available involving auxiliary building is usually bypassed. direct heating.

(7) Early Hydrogen Burne The sprays do not operate. The debris No, but enough infor=

is not coolable. A detonation occurs in the ice-condenser mation may be available

. which fails the containment. The auxiliary bu11 ding is from case involving fail.

I usually bypassed. ure at vessel breach due to a hydrogen burn. Some additional information would be required for cases involving direct heating at vessel breach.

'

• Case 2 for central and pessimistic walkthroughs includes the ef fects of in-vessel steam explosions and ex-vessel direct heating. Case 1 does not include these effects. Direct l

i heating includes hydrogen burning if the atmosphere is flammable. See text for a discussion of limitations associated with the results for steam explosions and direct heating.

• • Small containment leaks ir.clude those producing leakage greater than containment design but not large enough to preclude gradual overpressurization from steam and/or noncondensibles.

Likewise, small steam generator tube leaks include those producinq primary-to-secondary leakage greater than design but not large enough to depressurize the primary system below the accumulator setpoint before vessel breach.

! 64 l

Table 3.7 LIKELIHOOD CF CCNTAINMENT FAILURE MODES SY PLANT DAMAGE STATE (OPTIMISTIC WALKTHROUGH)

PIANT DAMAGE STATE (FREQUENCY)

AIYBY AINBY AINIY ANYBY ABYNB S21YBY S2NNNY S2NNNN S2INIY CONTAINMENT REI?.ASE MODE' (5.4-7) (1.9-7) (1. 9- 7) n.6-7) (6.0-7) (2.2-6) (1.9-5) (1.1-5 ) (2.0-7)

1. No containment failure. .88 - -

.88 --

.98 --

.66 --

2. Basemat seltthrough. .098 - -

.098 --

.001 -. .074 --

3. Late overpressurization due to steam and non-condensible gases, --

1.00 -- - -- - -- - --

with ice not melted or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

4. Late overpressurization due to hydrogen burn, with ice not melted -- - -- - .- .- -- -- -.

or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

5. Same as 3., but no sprays prior to core-concrete interactions. .001 .001 1.00 .001 --

.001 1.00 .25 1.00

6. Same as 4., but no sprays prior to core-concrete interactions. -- - -- - -- - .- -- --

7. Late steam overpressurization, with ice melted or bypassed prior to --

001 -- - -- -- - -- ..

vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen burn, with ice melted or bypassed prior to vessel, -- -- -- - -- .. -_ - --

breach and with sprays through core-concrete interactions.

9. Same as 7., but no sprays prior to core-concrete interactions. -- --

001 - -- - -- - .001

10. Same as 8., but no sprays prior to core-concrete interactions. -- - -- - -- - .. --

11. Hydrogen burn before vessel breach, with ice and sprays. -- - -- - -- -- - .-

m 12. Early hydrogen burn with ice but no sprays. - - -- -. -- -. -- .- --

U1

13. Early hydrogen burn without ice but with sprays. -- - -- - -- .- -- .- --

14. Early hydrogen burn without ice or sprays. - - -- -. -- - -- -- .-

a

16. Hydrogen burn / steam spike at vessel breach, with ice and sprays. -- - -- - -- -- _. -. --

18. In-wessel steam explosion. -- - -- -- -- - -- - -.

20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. -- - -- - - -- .- -- --

21. Direct heating and steam spike, with ice but no sprays. - -. -- - - - -- -- --

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice. -- - -- - -- - -- - -.

I 25. Steam spike at vessel breach, with no ice or sprays. .. - -. - .. -- -- ..

l 26. Hydrogen burn / steam spike at vessel breach, with no ice or sprays. -- - .- -- -- - - - --

! 30. Containment failure before core melt, no ice but with sprays. -- - -- --

.90 - -. - --

31. Containment failure before core melt, no ice or sprays. -- - .. -

.10 - -. --

32. Containment bypass, point of release sutunerged. -- - - - .- -. -- - --

33. Containment bypass, point of release not submerged. -- -- -- - -- - -- -- --

34. Pre-existing leakage or containment isolation failure. .017 -- --

.017 --

.017 --

.013 -

4

35. Induced steam generator tube rupture. -- - -- - -- - -- --

0' "Numoers correspond to the listing of release modes in Section 3.1.7 of the main report. Some release modes are omitted because they have no f requency cont ribution.

i i

~_ - -- ~ . . = - _ . ,

i 1

Table 3.7 L KEL!HOCO CT CONTA!NMENT TAILCRE MCOES BY PLANT CAMAGE STATE (CPTIMIST!C WALKTHROCOM, CONTINUED)

PI. ANT DAMAOE STATE (TREQUENCY)

S2NYBY $3IYBYB S3IN!YS 53INBYB TNNNNN TNYBYR TKYBYB V CvNTAILMENT DELEASE. MOOE* (1.3-6) (4.1-5) (3.8-6) (3.0-7) ( 8. 9- 6) (2.6-6) (7.0-7) (1. 2- 6 )

1. No contairsent failure. .98 .96 -- -- .58 .73 .96 --

2. Basemat meltthrough. .001 .026 -- --

.055 .020 .026 -.

3. Late overpressarization due to steam and non-condensible gases, -- -- -- .97 -- -. .-

with ice r.ot melted or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

4. Late overpressurization due to hydrogen burn, with ice not melt ed -- -- -- -- -- -- --

or bypassed through vessel breach and with sprays operating through tre period of core-concrete interactions.

S. Same as 3., but no sprays prior to core-concrete interactions. .001 .001 97 .001 .22 .001 .001 --

6. Same as 4., but no sprays prior to core-concrete interactions.

7 Late steam overpressurization, with ice melted cr bypassed prior to -- -- --

.026 -- -- .. -.

vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen burn, with ice melted or bypassed prior to vessel -- -- -- -- -- -- --

breach and with sprays through core-concrete interactions.

9. Same as 7., but no sprays prior to core-concrete interactions. .026

10. Same as 8., but no sprays prior to core-concrete interactions.

11. Hydrogen burn before vessel breach, with ice and sprays. -- -- -- -- -- .. -. --

CN 12. Early hydrogen turn with ice but no sprays.

04 13. Early hydrogen burn without ice but with sprays.

14. Early hydrogen burn without ice or sprays.

16. Hydrogen burn / steam spike at vessel breach, with ice and sprays. -- -- -- -- -- -- -- -.

18. In-vessel steam explosion.

20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. -- -- -- .. .. -- -- --

21. Direct heating and stets spike, with ice but no sprays.

23. Rydrogen burn / steam spike at vessel breach, with sprays but no ice. -- -- -- .- .- -- --

25. Steam spike at vessel breach, with no ice or sprays.

26. Hydrogen burn / steam spike at vessel breach, with no ice er sprays. -- -- -- -- -- -- -- --

30. Containment failure before core melt, no ice but with sprays.

31. Containment failure before core melt, no ice or sprays.

32. Containment bypass, point of release submerged. -- -- .. -. -- -- --

1.00

33. Containment bypass, point of release not submerged. -- -- -- -- -- .. -- --

34. Pre-existing leakage or containment isolation failure. .017 .017 -- --

.011 .013 .017 --

35. Induced steam generator tube rupture. -- -- -- --

.14 .24 -- --

"Nambers correspond to the listing of release modes in Section 3.1.7 of the main report. Some release modes are omitted because they have no frequency cont ribution.

a 1 Table 3.8. LIKELIHOOD OF CCNTAINMEi'T FAILURE MODES BY PLANT DAMACE STATE (CENTRAL WALK *HROUGH WITHOUT DIRECT HEATING AND IN-VESSEL STEAM EXPLOSICN)

PLANT DAMAGE STATE (FREQUENCY)

AIYBY AINBY AINIY ANYBY ABYNB S21YBY S2NNNY S2NNNN S2INIY CONTAINMENT REI. EASE MODE' (5.4-7) (1.9-7) (1.9-7) (1. 6-7 ) (6.0-7) (2. 2 -6 ) (!.9-5) (1.1-5 ) (2.0-7)

1. No containment failure. .51 -- --

.51 --

.77 --

.005 --

2. Lasemat maltthrough. .036 - --

.036 --

.009 --

.005 --

3. Late overpressurization due to steam and non-condensible gases, .30 .81 -

.30 --

.071 --

.032 --

with ice not melted or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

4. Late overpressurization due to hydrogen burn, with ice not melted -- - - - -- - -- -- --

, or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

5. Same as 3., but no sprays prior to core-concrete interactions. .090 .090 .90 .090 --

.088 .88 .017 .98

6. Same as 4., but no sprays prior to core-concrete interactions. -- - -- - -- - --

.002 --

7. Late steam overpressurization, with ice melted or bypassed prior to .034 .092 --

.034 --

.008 --

.012 --

vessel breach, and with sprays through core-concrete interactions, i 8. Late hydrogen burn, with ice melted or bypassed prior to vessel -- - -- - -- - -- -- --

breach, and with sprays through core-concrete interactions.

I

9. Same as 7., but no sprays prior to core-concrete interactions. .010 .010 .10 .010 -

.010 .098 .005 .099

10. Same as 8., but no sprays prior to core-concrete interactions. -- - - - -- - --

.002 .015

11. Hydrogen burn before vessel breach, with ice and sprays. -- - - - -- - -- -- --

CB 12. Early hydrogen burn with ice but no sprays. - - -- -- -- -- --

.090 --

i 13. Early hydrogen burn without ice but with sprays. -- -- -- -- - -- - --

14. Early hydrogen burn without ice or sprays. -- - -- - -- -- --

.010 --

i 16. Hydrogen burn / steam spike at vessel breach, with ice and sprays. -- -- - - --

.014 -- - --

18. In-wessel steam explosion. - -- -- - -- - -- -- --

20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. -- - -- - --

.001 .015 .59 --

, 21. Direct beating and steam spike, with ice but no sprays. -- - -- - -- - -- - --

I I 23. Rydrogen burn / steam spike at vessel breach, with sprays but no ice. - - -- - --

.002 -- - --

4

25. Steam spike at vessel breach, with no ice or sprays. -- - - - -- - --

.002 --

26. Hydrogen burn / steam spike at vessel breach, with no ice or sprays. -- - -- - -- --

.002 .23 .003

30. Containment failure before core melt, no ice but with sprays. -- -- -- -

.90 -- -- - --

31. Containment failure before core melt, no ice or sprays. -- - -- -

.10 - -- -- --

32. Containment bypass, point of release submerged. -- -- -- - -- -- - -- --

33. Containment bypass, point of release not sutznerged. -- -- -- - -- - -- -- --

l 34. Pre-existing leakage or containment isolation failure. .021 - 002 .021 --

.029 .002 .001 .002

35. Induced steam generator tube rupture. -- - -- - -- - -- -- --

"Nur.bers correspond to the listing of release modes in Section 3.1.'J of the main report. Some release modes are omitted because they have no frequency contribution.

a Table 3.8. LIKELIHOOD CF CONTAINMENT FAILCRE MOCES BY PLANT DAMAGE STATE (CENTRAL WALKTHROUGH WITHOUT DIRECT HEATING AND STEAM EXPLOSICN, CONTINUED)

PLANT DAMAGE STATE (FREQUENCY)

S2NYBY 53IYBYS S3INIYB S3INSYB TNNNNN TNYBYR TNYBYB V CONTAINMENT RELEASE MODE * (1. 3- 6) (4.1-5) (3.8-6) (3.0-7) (8.9-6) (2.6-6) (7.0-7) (1. 2-6 )

1. No containment fa ilur e. 77 .68 -- -

.010 .52 .68 --

2. Basemat meltthrough. .009 .014 -- -

.003 .010 .014 -

3. Late overpressurization due to steam and non-condensible gases, .071 .070 -- .64 .014 .070 .092 --

with ice not melted or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

4. Late overpressur13ation due to hydrogen burn, with ice not melted -- - - - -- - -- --

or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

%-Geme as 3., but no sprays prior to core-concrete interactions. .088 .070 71 .071 .003 .058 .077 -

6. Same as 4., but no sprays prior to core-concrete interactions. -- - -- -

.001 - -- --

7. Late steam overpressurization, with ice melted or bypassed prior to .008 .062 --

.22 .019 .031 .041 --

vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen burn, with ice melted or bypassed prior to vessel -- -- -- -

breach and with sprays through core-concrete interactions.

9. Same as 7., but no sprays prior to core-concrete interactions. .010 .024 .24 .024 .009 .013 .018 --

10. Same as 8., but no sprays prior to core-concrete interactions. -- -

.001 -

.004 - -- -

11. Hydrogen burn before vessel breach, with ice and sprays. -- - -- - -- -- -- -

12. Early hydrogen burn with ice but no sprays. -- - - -

.042 - -- -

13. Early hydrogen burn without ice but with sprays. - -- -

14. Early hydrogen burn without ice or sprays. -- - -- -

.044 - -- -

16. Hydrogen burn / steam spite a*. vessel breach, with ice and sprays. .014 .010 -

.010 --

.008 .010 -

18. In-vessel steam explosion. -- - - - -- - --- --

20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. .002 .001 .011 001 .28 .001 .001 -

21. Direct heating and steam spike, with ice but no sprays. -- - -- - - - -- --

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice. .002 .037 --

.037 -- .029 .037 -

25. Steam spike at vessel breach, with no ice or sprays. -- - -- -

.004 - - --

26. Hydrogen burn / steam spike at vessel breach, with no ice or sprays. --

.004 .041 .004 .42 .003 .004 -

30. Containment failure before core melt, no ice but with sprays. - - -- - -- - -- -

31. Containment failure before core melt, no ice or sprafs. -- - -- - - - - -

32. Containment bypass, point of release submerged. -- - -- - -- - --

.75

33. Containment bypass, point of release not subnerged. -- - - - -- -- --

.25

34. Pre-existing leakage or containment isolation failure. 029 .026 002 --

.001 .020 .026 -

35. Induced steam generator tube rupture. -- - -- -

.14 .24 -- -

• Numbers correspond to the listing of release modes in Section 3.1.7 of the main report. Some release modes are omitted because they have no frequency contribution.

- - - . _ . -- _ ~- . - . _ - . -- . . . - . - . - - _ _ _ , - - - - - - - - - - - - - - - - - - - - - - - - - -

1 1

4 i

i Table 3.9. LIKELIHOOD OF CONTAINMENT FAILURE MODES BY PLANT DAMAGE STATE (CENTRAL WALRTHROUGH WITH DIRECT HEATING AND IN-VESSEL STEAM EXPLOSION)

PIANT DAMAGE STATE (FREQUENCY)

AIYBY AINBY AINIY ANYBY ABYNB $2IYBY S2NNNY S2NNNN S2TNIY CONTAINMENT RELEASE MODE' (5.4-7) (1. 9-7 ) (1.9-7) (1.6-7) (6.0-7) (2. 2- 6 ) (1.9-5) (1.1-5) (2.0-7)

1. No containment failure. .51 -- --

.51 -- 77 --

.005 --

2. Basemat meltthrough. .036 - -

.036 --

.009 --

.005 --

] 3. Late overpressurization due to steam and non-condensible gases, .30 .81 --

.30 --

.071 --

.032 --

1 with ice not melted or bypassed through vessel breach and with sprays cperating through the period of core-concrete interactions, s

4. Late overpressurisation due to hydrogen burn, with ice not melted -- - - -- -- - -- -- --

or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

5. Same as 3., but no sprays prior to core-concrete interactions. .090 .090 .90 .090 -

.088 .88 .017 .88

6. Same as 4., but no sprays prior to core-concrete interactions. -- -- -- -- -- -- --

.002 --

7. Late steam overpressurisation, with ice melted or bypassed prior to .034 .092 -

034 --

.008 --

.010 --

vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen burn, with ice melted or bypassed prior to vessel -- -- -- - -- - -- - --

]j breach and with sprays through core-concrete interactions.

1 9. Same as 7., but no sprays prior to core-concrete interactions. .010 .010 .10 .010 --

.010 .098 .004 .099

10. Same as 8., but no sprays prior to core-concrete interactions. -- - -- -- -- -- --

.002 .015

11. Hydrogen burn before vessel breach, with ice and sprays. -- - -- - -- - -- -- --

Cn 12. Early hydrogen burn with ice but no sprays. -- - - -- -- - -

090 --

@ 13. Early hydrogen burn without ice but with sprays. -- - -- -- -- - -- - --

14. Early hydrogen burn without ice or sprays. -- - -- -- -- - --

.010 --

1

16. Hydrogen burn / steam spike at vessel breach, with ice and sprays. -- - -- -- --

.014 -- - --

18. In-wessel steam explosion. -- - -- -- -- -- -- - --

j 20. Hydrogen burn / steam epike at vessel breach, with ice but no sprays. -- - -- - --

.001 .015 .59 --

21. Direct heating and steam spike, with ice but no sprays. - - -- - -- - --

.077 --

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice. - - -- -- --

.002 -- - --

25. Steam spike at vessel breach, with no ice or sprays. -- -- -- - -- - - - --

26. Hydrogen burn / steam spike at vessel breach, with no ice or sprays. -- - -- -- -- -

.002 .16 .003

30. Contairaut failure before core melt, no ice but with sprays. -- - -- --

.90 - -- - --

31. Containment failure before core melt, no ice or sprays. -- - -- -

.10 - -- - -

1

32. Containment bypass, point of release submerged. -- -- -- -- -- - -- -- --

-l 33. Contairment bypass, point of release not submerged. -- -- -- -- -- -- -- - --

34. Pre-existing leakage or containment isolation failure. .021 --

.002 .021 --

.029 .002 .001 002

35. Induced steam generator tube rupture. -- - -- - -- -- -- - -

• Numbers correspond to the listing of release modes in Section 3.1.7 of the main report. Some release modes are omitted because they have no frequency contribution.

,m e, _ - - ----n .- 2 Asu - "

Table 3.9. LIKELIHOOD CF CCNTAINMENT FAILURE MODES BY FIANT DAMAGE STATE (CENTRAL NALRTHROUCh NITH DIRECT HEATING AND STEAM EXPLOSICN, CONTINUED)

PLANT DAMACE STATE (FREQUENCY)

S2NYBY $3IYBYB S3INIYB S3INBYB TNNNNN TNYBYR TNYBYB V CCNTAINMENT RELEASE MCCE" (1.3-6) (4.1-5) (3.0-6) (3.0-7) (8.9-6) (2.6-6) (7.0-7) (1.2-6)

1. No contair: ment failure. .77 .68 -- -

.010 .52 .68 -

2. Basemat seltthrough. .009 .014 - -

.003 .010 .014 -

3. Late overpressurization due to steam and non-condensible gases, .071 .070 --

.64 .014 .070 .092 -

with ice not melted or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

4. Late overpressurization due to hydrogen burn, with ice not melted --

or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

5. Same as 3., but no sprays prior to core-concrete interactions. .088 .070 .71 .071 .008 .058 .077 -

6. Same as 4., but no sprays prior to core-concrete interactions. -- - - -

.001 - -- --

7. Late steam overpressurisation, with ice melted or bypassed prior to .008 .062 -

.22 .018 .031 .041 -

vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen burn, with ice melted or bypassed prior to vessel breach and with sprays through core-concrete interactions.

j 9. Same as 7., but no sprays prior to core-concrete interact.ons. .010 .024 .24 .024 .003 ,013 .018 -

10. Same as 8., but no spraysprior to core-concrete interactions. -- - .001 -

.004 - -- -

11. Hydrogen burn before vessel breach, with ice and sprays.

q 12. Early hydrogen burn with ice but no sprays. -- - - --

.042 - -- -

0 13. Early hydrogen burn without ice but with sprays. -- - - - -- - -- -

14. Early hydrogen burn without ice or sprays. -- - -- -

.044 - - -

16. Hydrogen burn / steam spike at vessel breach, with ice and sprays. .014 .010 -

.010 --

.008 .010 -

18. In-vessel steam esplosion.

) 20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. .002 .001 .011 .001 .28 .001 .001 --

21. Direct heating and steam spike, with ice but no sprays. - - -- -

.058 -- -- -

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice. .002 .037 - .037 -- .029 .037 -

25. Steam spike at vessel breach, with no ice or sprays. - - - - 004 - -- --

26. Hydrogen burn / steam spike at vessel breach, with no ice or sprays. -- 004 .041 .004 .37 .003 .004 -

36. Containment failure before core melt, no ice but with sprays. -- - -- - -- - -

31. Containment failure before core melt, no ice or sprays. - - - - - - -

32. Containment bypass, point cf release sutznerged. - - - - -- - --

.75

33. Containment bypass, point of release not submerged. - - - -- -- - --

.25

34. Pre-esisting leakage or containment isolation failure. .029 .026 .002 -

.001 .020 026 -

35. Induced steam generator tube rupture. -- - -- -

.14 .24 -- -

" Numbers correspond to the listing of release modes in Section 3.1.7 of the main report. Some release modes are omitted because they have no frequency contributien.

f

- -. . . . - - - - ~ . _ _ . . _ _ . .-. .- . . . , . - . - _ - - . . - - - - - - - - - - - - - - - - - -

Table 3.10. LIKELIHOCD OF CCNTAINMENT FAILURE MODES BY PLAFT DAMAGE STATE (PESSIMISTIC WALKTHROUCH WITHCUT DIRECT HEATINO AND IN-VESSEL STEAM EXPLOSICN)

PIANT DAMACE STATE (FREQUENCY)

AIYBY AINBY AINIY ANYBY ABYNB S2IYBY S2NNNY S2NNNN $21NIY CONTAINMENT REI. EASE MCCE* (5.4-7) (1.9-71 (1. 9- 7) (1.;-7) (6.0-7) (2.2-6) (1.9-5) (1.1 -5 ) (2.0-7)

1. No containment failure. -- - -- - --

.043 - - -

2. Basemat meltthrough. -- - -- - --

3. Late overpressurization due to steam and non-condensible gases, -- - -- - -- .018 -- - --

with ice not melted or bypassed through vessel breach and with sprays cperating through the period of core-concrete interactions.

4. Late overpressurization due to hydrogen barn, with ice not melted - -- -- -- --

.008 -- -- --

or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

5. Same as 3., but no sprays prior to core-concrete interactions. - - -- - --

.034 .093 -

.068

6. Same as 4., but no sprays prior to core-concrete interactions. -- - -- - -

.011 .030 -

.021

7. Late steam overpressurization, with ice melted or bypassed prior to -- - -- - --

.027 -- - --

vessel breach, and with sprays through core-concrete interactions.

i

8. Late hydrogen burn, with ice melted or bypassed prior to vessel - -- -- - --

.012 --

.003 --

breach, and with sprays through core-concrete interactions.

9. Same as 7., but no sprays prior to core-concrete interactions. - -- -- - -

.001 .001 - .002

10. Sase as 8., but no sprays prior to core-cencrete interactions. -- -

.001 - -- 066 .22 .004 .13

11. Hydrogen burn before vessel breach, with ice and sprays. .017 .017 -

.024 -- - -- - --

4 12. Early hydrogen burn with ice but no sprays. .32 .32 .34 .45 -- - --

.25 --

H 13. Early hydrogen burn without ice but with sprays. .033 .033 -

.026 - - -- -- -

14. Early hydrogen burn without ice or sprays. .62 .62 .66 .50 -- - --

.25 --

16. Hydrogen burn / steam spike at vessel breach, with ice and sprays. -- - -- - -- .027 -- - --

18. In-wessel steam explosion. - - -- -- - -- - -- --

20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. -- - -- - -- .24 .38 -

.27

21. Direct heating and steam spike, with ice but no sprays. - - - - -- - - - --

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice. - - - - --

.051 -- - --

25. Steam spike at vessel breach, with no ice or sprays. - - - - -- - --

.018 --

26. Hydrogen burs / steam spite at vessel breach, with no ice or sprays. -- - - - -- .46 .38 .47 .51

30. Containment failure before core melt, no ice but with sprays. - - -- -

.10 - -- - --

31. Containment failure before core melt, no ice or sprays. -- - -- -

.90 - -- -- --

32. Contairment bypass, point of release sutmerged. -- - --

33. Contairment bypass, point of release not submerged. -- -

34. Pre-existing leakage or containment isolation failure. - - -- - -

.003 -- - --

35. Induced steam generator tube rupture. -- - -- - -- - -- --

" Numbers correspond to the listing of release modes in Section 3.1.7 of the main report. Some release modes are omitted because they have no frequency cont ribution.

Tacle 3.10. LINELIHOCD CF CCNTAINMENT FAILORE MODES BY PLANT DAMACE STATE (PESSIMIST!C WAIJTHROUCN WITHOUT DIRECT HEAT!NG AND STEAM EXPLOSICN, CONTINUED)

PIANT CAMAGE STATE (FREQUENCY)

! S2NYBY 531YBYB S3!NIYB S3INBYB TNNNNN TNYBYR TNYBYB V

CCN*AINwENT RELLASE MODE' (1.3-6) (4 .1-5 ) (3.8-6) (3.0-7) (8.9-6) (2.6-61 (7.0-7) (1.2-6)

1. No containment failure. .043 .036 -- - -- .027 .036 -

2. Basemat seltthrough. -- - -- - - -- -- --

3. Late overpressurization due to steam and non-condensible gases, .025 .019 --

.035 --

.011 .015 -

with ice not melted or bypassed through vessel breach and with sprays operating through tne period of core-concrete interactions.

4. Late overpressurization dae to hydrogen burn. with ice not melted .011 .008 --

.011 -- .005 006 -

or bypassed through vessel breach ar!d with sprays operating through the period of core-concrete interactions.

5. Same as 3., but no sprays prior to core-concrete interactions. .047 .035 .069 .035 --

.021 .028 -

6. Same as 4., but no sprays prior to core-concrete interactions. .015 .011 .022 .011 --

.007 009 -

7. Late steam overpressurization, with ice melted or bypassed prior .025 .019 --

.035 --

.017 .022 --

to vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen burn, with ice melted or bypassed prior to vessel .011 .008 --

.011 .002 .007 .010 -

breach and with sprays through core-concrete interactions.

9. Same as 7., but no sprays prior to core-concrete interactions. .001 .001 .001 .001 .001 .001 .001 -

10. Same as 8., but no sprays prior to core-concrete interactions. .061 .045 .090 .045 .002 .041 .054 -

I

11. Nyorogen burn before vessel breach, with ice and sprays. - - -- -- - - .001 --

Q 12. Early hydrogen burn with ice but no sprays. - - - - .040 .009 .012 -

M 13. Early hydrogen burn without ice but with sprays. --

.013 --

.013 --

.009 .012 -

14. Early hydrogen burn without ice or sprays. -

.25 .26 .25 .39 .10 .24 --

16. Rydrogen burn / steam spike at vessel breach, with ice and sprays. .038 .028 - .C28 -- .017 .022 --

18. In-vessel steam explosion. -- - -- -- -- -- -- -

20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. .34 .25 .28 .25 --

.15 .20 -

21. Direct heating and steam spike, with ice but no sprays. -- - -- - -- - -- -

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice . 038 .028 -

.028 --

.025 .032 -

25. Steam spike at vessel breach, with no ice or sprays. -- -- -- - .017 - -- -

26. Hydrogen burn / steam spike at vessel breach, with no ice or sprays. .34 .25 .28 .25 .41 .23 .30 -

30. Containment failure before core melt, no ice but with sprays. - - - - -- - -- -

31. Containment failure before core melt, no ice or sprays. -- - -- - -- - - -

32. Contairmient bypass, point of release sutsierged. -- - - - -- - -- .50

33. Containment bypass, point of release not sutznerged. -- - - - -- - - .50

34. Pre-eatsting leakage or containment isolation failure. .004 .003 -- - - .002 .003 -

35. Induced steam generator tube rupture. - - -- -

.14 .24 -- -

{ "Nambers correspond to the listing of release modes in Section 3.1.7 of the main report. Some release modes are omitted because they have no frequency contribution.

- - _ ~ - . - - - - . . - - .. - --.--_ - - ~ _ - . . . _ _ _ _ _ ~ _ _ - . - _ - . _-.. _ . . . . - -- .. . . _ . - - - - - - - - - - - - - - - - - - - - - - - - - - - -

Table 3.11. LIKELIHOO3 OF CONTAINMENT FAILCRE MODES BY PLANT DAMAGE STATE (PESSIMISTIC WALKTiiRCUCH WITH DIRECT HEATING AND IN-VESSEL STEAM EXPLOSION)

PIANT DAMAGE STATE (FREQUENCY)

AIYBY AINSY AINIY ANYSY ABYNB S2IYBY S2NNNY S2NNNN S2INIY CCMTA1NMENT RELEASE MCDE' (5.4-7) (1.9-7) (1.9-7) (1.6-7) ( 6.0-7) (2.2-6) (1.9-5) (1.1-5) (2.0-7)

1. No contairment failure. -- - -- - --

.039 -- - ---

2. Basemat seltthrough. - - --- - -- - -- - --

1. Late overpressarization due to steam and non-condensible gases, -- - - - -

.016 -- - --

with ice not melted or bypassed throJGh Vessel breach and With sprays operating through the period of core-concrete interactions.

4. Late overpressarization due to hydrogen burn, with ice not melted -- - -- - -

.007 -- - --

cr bypassed through vessel breacn and with sprays operating through the period of core-concrete interactions.

5. Same as 3., but no sprays prior to core-concrete interactions. -- - - - --

.030 .084 -

.061

6. Same as 4., but no sprays prior to core-concrete interactions. -- - - - --

.007 .027 -

.019

7. Late steam overpressuriaation, with ice melted or bypassed prior - -- - - --

.024 -- - --

to vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen barn, with ice melted or bypassed prior to vessel -- -- -- -- --

.011 --

.004 --

breach and with sprays through core-concrete interactions.

9. Same as 7.~, but no sprays prior to core-concrete interactions. -- - - - --

.001 .001 -

.002

10. Same as 9., but no sprays prior to core-concrete interactions. -- -

.001 - --

.059 .11 .007 .12

11. Hydrogen burn before vessel breach, with ice and sprays. .015 .015 --

.021 -- - -- - --

q 12. Early hydrogen burn with ice but no sprays. .33 .33 .34 .45 -- -- --

.25 --

W 13. Early hydrogen burn without ice but with sprays. .030 .030 --

024 - - -- - --

14. Early hydrogen burn without ice or sprays. .63 .63 .66 .50 -- - -- .25 --

16. Rydrogen burn / steam spike at vessel breach, with ice and sprays. - - - -- -

.024 -- -- --

18. In-wessel steam esplosion. - - - -- --

.10 .10 .050 .10

20. Hydrogen burn / steam spike at vessel breach, with ice but no sprays. - - -- - --

.22 .34 -

.24

21. Cirect heating and steam spike, with ice but no sprays. -- -- -- - -- -- -- .21 - - -

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice. -- - -- - -- 046 -- -- --

25. Steam spike at vessel breach, with no ice or sprays. - - -- - -- - --

.001 --

26. Hydrogen burn / steam spike at vessel breach, with no ice or sprays. -- - - -- --

.41 .34 .22 .46

30. Containment failure before core melt, no ice but with sprays. - - -- -

.10 -- -- - --

31. Containment failure before core melt, no ice or sprays. -- - -- --

.90 - -- - --

32. Containment bypass, point of release subinerged. -- -- -- - -- - -- -- --

33. Contairment bypass, point of release not submerged. -- - -- - - - -- -- --

34. Pre-esisting leakage or containment isolation failure. -- - -- - --

.003 -- - --

35. Indaced steam generator tube rupture. -- -- -- - -- -- -- - --

'3iumbers correspond to the listing of release modes in Section 3.1.1 of the main report. Some release modes are omitteo because they have no frequency contribution.

. . - _ - _ - _ - _ . - - . . . - _ - . - _ __ . . . .- ;-.-__.._.____ .- - _ - _ . _ ~ - .. .

Table 3.11. LIKELIHOOD tF CONTAINMENT FAILURE MODES BY PLANT DAMAGE STATE (PESSIMISTIC NALKTHROOCH NITH DIRECT HEATING AND STEAM EXPLCSION, CONTIN 0ED)

PLANT DAMAGE STATE (FREQUENCY) 52NYBY S3IYBYB S3INIYB S3INSYS TNNNNN TNYBYR TNYBYB V CN TAINMENT RELIASE McCE* gg,3 6) (4.1-5) (3.8-6) (3.0-7) ( 8. 9- 6) (2.6-6) (7.0-7) (1.2-6)

I

1. No containment failure. .048 .032 - - - .024 .032 -

2. Basemat seltthrough. - - - -

3. Late overpressurization doe to steam and non-condensible gases, .023 .017 - .031 -

.010 .013 -

with ice not melted or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

4. Late overpressurization due to hydrogen burn, with ice not melted .010 .007 -- .010 - .004 .006 -

or bypassed through vessel breach and with sprays operating through the period of core-concrete interactions.

J l

5. Same as 3., but no sprays prior to core-concrete interactions. .042 .031 .062 .031 -

.019 .025 -

i 6. Same as 4., but no sprays prior to core-concrete interactions. .013 .010 .020 .010 -

.006 .008 -

7. Late steam overpressurtration, with ice melted or bypassed prior to .023 .017 - .031 -

.015 .020 -

vessel breach, and with sprays through core-concrete interactions.

8. Late hydrogen burn, with ice melted or bypassed prior to vessel .010 .007 - .010 .003 .007 .009 -

beach and with sprays through core-concrete interactions.

9. Same as 7., but no sprays prior to core-concrete interactions. .001 001 .001 .001 .001 .001 .001 -

' .055 .041 .081 .041 .004 037 .049 -

10. Same as 8., but no sprays prior to core-concrete interactions.

11. Hydrogen burn before vessel breach, with ice and sprays. -- - - - -- -

.001 -

'J 12. Early hydrogen burn with ice but no sprays. - - - - .040 .009 .012 -

e b

13. Early hydrogen burn without ice but with sprays. - .012 --

.012 -

.009 .011 -

14. Early hydrogen burn without ice or sprays. - .25 .26 .25 .39 .18 .23 -

16. arydrogen burn / steam spite at vessel breach, with ice and sprays. .034 .025 - 025 - .015 .020 -

18. In-wessel steam esplosion. .10 .074 .074 .074 .043 .056 .074 -

l l 20. Itydrogen burn / steam spike at vessel breach, with ice but no sprays. .30 .23 .25 .23 -- .14 .18 -

j 21. Direct heating and steam spike, with ice but no sprays. - - - -

.16 -- -- -

23. Hydrogen burn / steam spike at vessel breach, with sprays but no ice. .034 .025 -

.025 --

.023 .030 -

25. Steam spite at vessel breach, with no ice or sprays. - - - -

.004 - - -

26. Nydrogen burn / steam spite at vessel breach, with no ice or sprays. .31 .23 .25 .23 .22 .21 .30 -

30. Contairment failure before core melt, no ice but with sprays. -- - - - -- - --

31. Containment failure before core molt, no ice or sprays. -- - - - -- - --

32. Containment bypass, point of release submerged. -- - - - - - - .50

33. Containment bypass, point of release not submerged. - - - - -- - -- .50

34. Pre-esisting leakage or containment isolation f ailure. .003 .002 - - -- .002 .003 -

, 35. Induced steam generator tube rupture. - - -- -

.14 .24 -- --

i

" Numbers correspond to the listing of release modes in Section 3.1.7 of the sain report. Some release modes are omitted because they have no frequency contribution.

i i

1 i

k

TABLE 3.12 SENSITIVITY OF SAMPLE CASE RESULTS TO ALTERNATIVE ASSIGNBEENTS OF NUBE!RICAL VALUES OPTIMISTIC CENTRAL PESSINISTIC Unlikely .1 .01 .01 .1 .1 .01 .01 .1 .1 .01 .01 .1 Remote Poss. .001 .001 .0001 .01 .001 .001 .0001 .01 .001 .001 .0001 .01 l

i i

No Failure .95 .99 .99 .93 .46 .54 .54 .45 .006 .007 .007 .006 l Meltthrough .05 .006 .005 .06 .45 .45 .45 .45 .06 .06 .06 .06 4

Ut Late Overpressure .001 .001 .01 .09 .01 .01 .09 .03 .02 .02 .03 Late Leak .01 .01 .01 .01 Late H2 Bura .001 .001 .06 .05 .05 .06 Early Leak .03 .03 .03 .03 I

Steam Spike l

l Early H2 Burn .04 .04 .04 ,

.04 t

, SS + H2 88E8 77 77 77 77 l

l Iso. Failure .002 .002 .002 .002 .002 .002 .002 .002 l

l l

i l

l APPENDIX A SAMPLE EVENT TREE PROBLEM A sample event tree problem has been constructed to facilitate the discussion of how input for the event tree program is structured. Before discussing this sample problem, some of the capabilities of the program are discussed.

i The event tree program was written to do the bookkeeping of very large event trees that were necessary to properly address problems in the Severe Accident Risk Reduction Program (SARRP). The event trees developed in SARRP typically involved 50-60 " top questions" or " nodes". This in itself established a need for an automated bookkeeping procedure for the trees. In addition, without making very conservative assumptions (something we were to avoid in SARRP) nearly all of the branch point probabilities were dependent upon the occurrence or absence of prior events in the tree -- something very few other past event tree analyses have allowed for. Not only were most of the branch point probabilities dependent upon the path taken in the event tree up to that point, some of the probabilities needed to be calculated internally. That is, the user did not know a priori what the probability would be. This was most frequently the case for the probability of containment failure. In these instances, pressure loads were compared to estimates of a containment failure pressure distribution to obtain an estimate of containment failure probability. Keeping these points in mind, several key features of the program will be discussed.

As mentioned above, most of the questions involve user supplied branch point probabilities. However some questions utilize the capability of the program to internally calculate l the branch point probability. The way this is performed is l certain " parameters" which have had values specified at previous nodes are added together and compared to " reference parameters" which have a value specified at the current node.

Take for example the probability of containment failure due to an early hydrogen burn. This is dependent upon the baseline pressure in containment prior to the burn, the probability of a hydrogen burn occurring, the pressure rise due to the burn, and containment failure pressure. Typically the value of baseline pressure and pressure rise due to the burn would be supplied at prior nodes as " parameters". The probability of the burn occurring would be input as a branch point probability at an earlier node (most likely the same one that supplied the value of pressure rise due to the burn). Then at the node which addresses early hydrogen burn failures, the program would be instructed to add the values of baseline pressure and hydrogen-burn pressure rise and compare the sum to a distribution of containment failure pressure to obtain a branch point probability. The containment failure pressure (a " reference parameter") is specified through input for that node.

A-1

From the discussion above it becomes apparent that there are at least three different types of questions for the event tree. They are those which the user supplies only branch point probabilities, those for which the user supplies " parameter" values as well as branch point probabilities, and those which the user supplies a list of " parameters" to be added and

" reference parameters" to which the sum is to be compared. In fact, there are six different types of questions distinguished by the program. The remaining three types arise by considering the three types mentioned above'as being either independent of or dependent on prior events in the tree. A summary of the different question types is shown below in Table A.L.

Table A.1 Summary of the different question types for the event tree program Dependency upon prior events Independent Dependent User Supplies Branch Type 1 Type 2 Point Probabilities Only User Supplies Branch Type 3 Type 4 Point Probabilities and Parameter Values User Specifies a List Type 5 Type 6 of Parameters to be Summed and Compared to Reference Parameters to Obtain Branch Point Probabilities For those questions which are dependent upon prior events, different " dependency cases" are defined for which different sets of branch point probabilities are supplied. The

" dependency cases" are described by specifying branches which must have been or must not have been taken for a series of key questions. To specify a branch that must have been taken the question number and the branch number are supplied to the program in the set of data for that " dependency case". To specify that a branch must not have been taken, the negative of the branch number is supplied to indicate a "not" condition.

Using this strategy, many different " dependency cases" can be specified.

Marching through the event tree and finding the probability of each of the pathways would be useless without some means of condensing all of the different outcomes into some more tangible A-2

I fotn. Thus ths progcco was written to bin all of ths_different outcoass, in cono us@c cpacificd fcchion, for tho output. This binning is performed based upon certain key characteristics by which the user would like to be able to differentiate the output. For example, for the SARRP analysis we were' interested l in the fraction of the sequence initiating frequency which I would result in different modes of early and late containment failuce, with containment sprays operating or not, with different levels of secondary containment bypass, etc.

Based on these needs the program was set up to have charactecistically different " dimensions" for binning. These

" dimensions" are things such as those mentioned in the previous paragraph. Within each of these " dimensions", different bins are specified. For example, in the eacly containment failure category there may be several different levels of containment  ;

leakage as well as gross failures due to several different causes. In the case of containment sprays, commonly considered binc were different combinations of early and late spray failure and success. The binning information for the SARRP analyses was chosen to adequately describe scenarios such that fission product source terms could be assigned to the scenacios without requiring an ovetly conservative binning of.the outcomes.

The binning is completely user specified.. Input is prepared to specify the number and title of each of the dif f erent " dimensions" and the number and title of each of the bins within each of the dimensions. Then information is provided identifying combinations of key questions and branches which must oc must not have been taken for an outcome to fall within each of the respective bins in each " dimension". The output then consists of aggregate probabilities of occurrence of diftetent combinations of the various " dimensional bins".

With this background description, we will now proceed to describe the format of the input. A few key points are worth mentioning. While titles are supplied for each of the questions and branches and the questions are numbered in the input, the program does not use them. The program knows the questions and branches by number only (both assigned in order of occuttence). The titles of the questions and the btanches are provided in the input for the ease of the uset in preparation of the input deck.

Table A.2 provides a sample input deck constructed solely for the purpose of familiatizing the novice with the construct of the input. It deals with early failure of containment in a sub-atmosphetic PWR containment due to an early hydrogen burn.

The discussion of the input follows in Table A.3 with reference to line numbers in the sample input found in Table A.2.

l

A-3 , l

\

, 1

' Table A.2

h. g Senple Problen Input

,( Line No.

t ~ Input

1 1 SAMPLE - Early fallute due to H2 burning-in S3D at a PWR l 2 3 1.0 1.0 '1.0 8 1.OE-04 1.0E-07 l 13 Nwal Pinit Pinit Pinit Nguest Prntol Tolet 1

' 1 1s there pre-existing. leakage or isolation failute?

4 i 5 4 EE3 EE2 EEL EE0 6 1 1 2 3 4 7 0.000 0.000 0.013 0.987 8 0.000 0.002 0.028 0.970 9 0.000 0.002 0.051 0.947 i 10 2 Initial break size

' 11 4 A SL S2 S3

, 12 1 1 2 3 4

~

13 0.000 0.000 0.000 1.000

, .$ 14 0.000 0.000 0.000 1.000 j 15 0.000 0.000 0.000 1.000 1 16 3-Induced failure in primacy. system?

17 5 A St. S2 S3 none i 18 2 1 2 3 4 5

+

19 2 i 2 0 1 2 21 4 I

22 S3 j 23 0.000 0.000 1.000 0.000 0.000 24 0.250 0.000 0.250 0.000. 'O.500 l 25 0.000 0.000 0.000 0.000 1.000 l 26 Otherwise i 27 0.000 0.000 0.000 0.000 1.000 28 0.000 0.000 0.000 0.000 1.000 29 0.000 0.000 0.000 0.000 1.000

, 30 4 Sprays on early?

31 2 esp noESp 32 2 1 2 i 33 5

f. 34 1 1 35 1 j 36 EE3 i

37 0.000 1.000 j 38 0.000 1.000

, 39 0.000 1.000  ;

j 40 2 1 2 i [ 41 2 1  !

,,'42 EE2 A i 0.000 43 1.000 j 44 1.000 0.000

]

45 1.000 0.000 1

I A-4 f

Tablo A.2 (cont.)

Sorplo Problen Input Line No. e infat 46 2 1 2 47 2 2 48 EE2' ,S1 -

49 0.900' O.100 l 50 0.900 0.100 l 51 0.900 0.100 l 52 3 1 2 ~2 l

53 2' -1 -2 i 54 EE2 noA noS1 55 0.000 1.000 56 0.000 1.000 /

57 0.000 1.000 58 Otherwise 59 1.000 0.000

60 1.000 0.000 ,

61 1.000 0.000 62 5 Early baseline presure  ;

63 1 Pbase .,

64 4 1 65 5 66 3 1 1 4 67 -1 -2 1 68 noEE3 noEE2 esp 69 1.000 .

70 1.000 71 1.000 .

72 1 -

73 1 20.000 74 20.000 75 20.000 EPbase 76 3 1 1 4 77 -1 -2 2 78 noEE3 noEE2 noESp 79 1.000 80 1.000 81 '

. 1.000 80 1 83 1 40.000 84 40.000 ,

85 40.000 EPbase

, .~

> g .y

, l' s_

l

/

,l*

ti A-5

,l

).

f* -

gr ,

~

Tablo A.2 (cont.)

Sctplo Probica Input Line No.

Input 86 2 1 4 87 3 1 88 EE2 esp 89 1.000 90 1.000 91 1.000 92 1 93 1 15.0 94 15.0 95 15.0 EPbase 96 2 1 4 97 2 2 98 EE2 noESp

'19 1.000 100 1.000 101 1.000 l 102 1 103 1 20.0 104 20.0 105 20.0 EPbase 106 Otherwise 107 1.000 108 1.000 109 1.000 110 1

? 111 l 14.7 112 14.7 113 14.7 EPbase 114 6 Maximum break size in primary system 4 A S1 S2 S3 115 2 1 2 3 4 116 117 7 118 2 2 3 119 4 5 120 S3 none 121 0.000 0.000 0.000 1.000 122 0.000 0.000 0.000 1.000 123 C.000 0.000 0.000 1.000 124 2 2 3 125 3 5 126 S2 none 127 0.000 0.000 1.000 0.000 128 0.000 0.000 1.000 0.000 129 0.000 0.000 1.000 0.000 I

A-6

Table A.2 (cont.)

Sanple Problea Input Line No.

Input 130 2 2 3 131 2 5 132 S1 none l

133 0.000 1.000 0.000 0.000 j

134 0.000 1.000 0.000 0.000 135 0.000 1.000 0.000 0.000 136 1 3 137 4 138 S3 139 0.000 0.000 0.000 1.000 140 0.000 0.000 0.000 1.000 141 0.000 0.000 0.000 1.000 142 1 3 I 143 3 144 S2 145 0.000 0.000 1.000 0.000 146 0.000 0.000 1.000 0.000 147 0.000 0.000 1.000 0.000 148 1 3 149 2 150 S1 151 0.000 1.000 0.000 0.000 152 0.000 1.000 0.000 0.000

< 153 0.000 1.000 0.000 0.000 2

154 Otherwise 155 1.000 0.000 0.000 0.000 156 1.000 0.000 0.000 0.000 157 1.000 0.000 0.000 0.000 158 7 Hydrogen burn before vessel breach?

159 2 EHB noEHB 160 4 1 2 161 6 162 4 1 1 4 6 163 -1 -2 1 1 164 noEE3 noEE2 esp A 165 0.000 1.000 166 0.100 0.900 167 0.100 0.900 168 1 169 2 0.0 0.0 170 25.0 0.0 171 50.0 0.0 DP-EH A-7 i

-- y- m- --. y = m- g y -.e v - wy - - -

Tablo A.2 (cont.)

Sceple Problon Input Line No.

Input 172 4 1 1 4 6 173 -1 -2 1 2 174 noEE3 noEE2 esp S1 175 0.000 1.000 176 0.000 1.000 177 0.100 0.900 178 1 179 2 0.0 0.0 180 0.0 0.0 181 35.0 0.0 DP-EH 1 1 4 6 6 182 5 183 -1 -2 1 -1 -2 184 noEE3 noEE2 esp noA noS1 1 185 0.000 1.000 186 0.000 1.000 187 0.000 1.000 188 1 189 2 0.0 0.0 190 0.0 0.0 191 0.0 0.0 DP-EH 192 3 1 4 6 193 2 1 1 194 EE2 esp A 195 0.000 1.000 196 0.000 1.000 197 0.100 0.900 198 1 199 2 0.0 0.0 200 0.0 0.0 201 35.0 0.0 DP-EH 202 3 1 4 6 203 2 1 2 204 EE2 esp SL 205 0.000 1.000 206 0.000 1.000 207 0.100 0.900 208 1 209 2 0.0 0.0 210 0.0 0.0 211 25.0 0.0 DP-EH 212 Othetwise 213 0.000 1.000 214 0.000 1.000 215 0.000 1.000 216 1 217 2 0.000 0.000 218 0.000 0.000 219 0.000 0.000 DP-EH A-8

Table A.2 (cont.)

Sanple Problem Input Line No.

Input 220 8 Does containment fail due to an early hydrogen burn?

221 2 EH noEH 222 5 1 2 223 2 .1 2 224 EPbase DP-EH 225 'AND' 226 ' NORMAL' 2 134 2.5 227 134 15 228 ,

60 15 229 CFP MEAN SIGMA

> 230 ' SAMPLE Early failure due to H2 burning in S3D at a PWR 231 ' OPTIMISTIC' ' CENTRAL' ' PESSIMISTIC' I

232 2 'CF' ' Sprays' 233 5 5 'EE0' ' eel' 'EE2' 'EE3' 'EH' 234. 2 1 1 8 235 4 2 236 EE0 noEH 237 2 2 1 8 238 3 2 239 EE1 noEH 240 2 3 1 8 241 2 2 242 EE2 noEH 243 2 4 1 8 244 1 2 245 EE3 noEH 246 1 5 8 247 1 248 EH 249 2 2 ' esp' 'noESp' 250 1 1 4 251 1 252 esp 253 1 2 4 254 2 255 noESp i

i A-9 l

l

- . - .-= _. - .-. .

Table A.3 Dotalled Discussion of Senple Input 4

Line in Table A.2 Input Description 1 Dummy title for this set of input. Not actually used by program but made available for'the user for quick identification of the input 2-3 Number of walkthroughs of the tree, the initiating frequency of each of the walkthroughs (when set equal to'one, conditional probabilities are obtained), the number of questions in the tree, the print tolerance (output with aggregate probability less than the print tolerance is not found in the printed output), and the cutoff tolerance in the tree (if the probability of a ,

pathway is reduced below the cutoff tolerance, l calculations cease for-that portion of the tree). .The titles in line three are not used by the program but facilitate the user in preparing the input.

4 Question number one is concerned with the level of preexisting leakage or isolation f ailure, l 5 This first node in the tree has four branches. They are titled EE3, EE2, eel, and EEO. EE3 stands for leakage of sufficient size to be equivalent to a gross  ;

failure of containment. EE2 is-leakage smaller than  !

gross failure (in terms of amounts of fission products released) but still large enough to prevent containment failure due to long term and gradual overpressurization. EEL is leakage too small to prevent long term overpressurization but larger than design leakage of containme'nt. Design leakage or less is designated by EEO. These titles are not used by the program but are provided for ease of-preparing and reading the input.

6 The first' number-is all that is actually used by the

program and denotes this question as being of type 1 (see earlier discussion of question types). The other numbers are simply the branch numbers for this question which are supplied for ease of reading the input.

7-9 These three lines provide the branch point probabilities for this question for the three

- different walkthroughs. Each line supplies the probabilities for a different walkthrough. In this example (this will be true for'all the questions in this sample input), the first line. represents input

for an optimistic walkthrough, the third line for a pessimistic walkthrough, and the second line represents a central viewpoint.

A-10 i

e

_ _ _ _ _ . _ _ _ _ _ _ . . _ . , . - . . - - - - - - - . . , , .., _, , -_ -,, ,, , , , , , ,7_ _,,.,,.,,_y .

y

Tablo A.3 (cont.)

Detailed Discussion of Seaple Input Line in Table A.2 Input Description 10 Question number 2 addresses the size of the initial break in the primary system.

11-12 The question has four branches (A, S 1, S2, S3) and is of type 1. The branches are "large" (A),

" intermediate" (S1), "small" (S2), and "very small" (S3).

13-15 The branch point probabilities for the three different walkthroughs are given in these three lines.

16 Question 3 is concerned with the possibility of induced failures in the primary system in high pressure scenarios due to natural convection of hot gases throughout the system.

17-18 Question three has 5 branches (four similar to question 2 and one additional to account for no induced failure) and is of type 2. Type 2 means the branch point probabilities are dependent upon prior events such that " dependency cases" must be specified.

19 There will be two " dependency cases" specified.

20 For this first " dependency case," there is 1 question-branch pair specified which must either have or have not been taken previously in the tree for the following branch point probabilities to be used for this question. The question of dependency is question number 2.

21 The branch specification is 4. This means (since the branch specification is positive) that branch 4 must have been taken previously at question 2 for the following set of branch point probabilities to be used.

22 The title of the branch which must have been taken is S3. This title is not used by the program but is supplied f or ease of preparatior, and examination of the input.

23-25 The branch point probabilities for this dependency case for this question for the three different walkthroughs.

26 This is the second and last dependency case. In fact, it is a catchall case. That is, any paths which did

- not fit into any of the previous cases use this set of A-11

Table A.3 (cont.)

Datailed Discussion of Saaple Input Line in.

, Table A.2 Input Description 1

26 cont branch point probabilities. (Hence the "otherwise" i-descriptor.) This line is not actually used by the l program but is another clarification feature of the input structure.

1 27-29 The three sets of branch. point probabilities for the  !

[ "otherwise" catchall case.

30-32 Question 4 addresses early spray operation, it has 2

branches, and is of' type 2.

33 5 dependency cases are specified for this question.

1 34-36 The first dependency case has 1 question-branch pair specified, this is branch 1 of question 1 (the EE3 branch). ,

~

37-39 The. branch point probabilities of_for spray operation given preexisting leakage or isolation failure equivalent to gross rupture of containment.

40-42 The second dependency case has 2 question-branch pairs specified, these are questions.1 and 2 and branches'2

and 1 respectively, corresponding to preexisting leakage or isolation of' level 2 and a large initial i break in the primary system. Both question-branch I pairs must have been taken previous 1i for the set of l

branch point probabilities that follow to be used.

I j 43-45 The branch point probabilities for spray operation for

! the three different walkthroughs and the second dependency case.

l 46-51 The third dependency case specification (leakage level 2 and intermediate initial break in the primary system) and the corresponding branch point probabilities.

52-54 The fourth dependency case specifies a leakage level of 2 and an initial break in the primary system other than large or intermediate (signified by the negative branch number). Thus either a small or a very small initial break with a leakage level of 2 would fall into this case.

55-57 The branch point probabilities for the fourth dependency case.

A-12

Table A.3 (cont.)

Datailed Discussion of Sanple Input I Line in ,

i Table A.2 Input Description 58-61 The catchall branch point probabilities for question number 4.

62 Question 5 addresses the early baseline pressure in containment.

63-65 There is only one branch for this question since it is only in the tree to supply the value of the early baseline pressure " parameter". The question is of type 4 meaning it is dependent upon prior events and l

that a parameter value is specified. There are 5 dependency cases-specified.

66-71 The-first dependency case is specified as branch one-of question 4 and anything but branch-1 or 2 on question 1. Since this is a single branch node, the probabilities for all three walkthroughs are all unity.

72 The value of 1 parameter will be specified in this case. (The value of more than one parameter can be specified at a type 3 or 4 question.)

73-75 The parameter identification number is 1 and the value of-the parameter for the three-different walkthroughs is 20.0.76-113 Specification of the remaining four dependency cases, and the associated branch point probabilities, parameter identification numbers and-parameter values.

2 114 Question number 6 combines the information from question 2 and 3 to simplify the specification of dependency cases in later questions. No new information is gained by this node, but past information is summed up into a more convenient form.

115-117 Question 6 has 4 branches and is of type 2. There are

7 dependency cases specified.

118-157 Specification of the seven different dependency cases and the branch point probabilities. Note that since this question serves in a summary manner, the branch point probabilities are all unity or zero and are- i

invariant with the three different walkthroughs. l

\

158 Question 7 addresses the probability of a hydrogen  !

burn before vessel breach and the magnitude of such a l hydrogen burn. i i

A-13 l l

l

~

i h Table A.3 (cont.)

Detailed Discussion of Sample Input Line in Table A.2 Input Description 159-161 . Question seven consists of two branches and is of type

4. 6 dependency cases are specified.

162-167 The first dependency case has 4 question-branch pairs specified. They are "not" branch 1 or 2 of question 1 and branch.1 of both question 4 and 6. The three sets of branch point probabilities are also supplied.

168 The value of 1 parameter will be specified for this dependency case.

169-171 The parameter identification number is 2. (This parameter is the pressure' rise associated with the0.,

~

hydrogen burn.) The values of the parameter are i 25., and 50., for the three different walkthroughs if branch 1 is taken. Otherwise the values are all zero, l

since the other branch involves no combustion. (The value of the parameter for a given branch appears under that branch in order of the different walkthroughs.)

172-219 Specification of the remaining five dependency cases along with the parameter identification and values.

220 Question 8 addresses the probability of containment failure due to an early hydrogen burn.

221-222 Question eight has 2 branches (a necessity for this type of question) and is of type 5. This means it is of the type which compares a sum of specified parameters to some reference parameters and the branch point probability will be internally calculated by the program.

223-224 There are 2 parameters specified to manipulate. These are parameters 1 and 2. These are the values of early baseline pressure and early hydrogen burn pressure rise respectively. The titles in line 224 are provided entirely for clarification reasons and are +

not actually used in the program.

225 The parameters are to be "anded" together. That is they are to be summed. Other options here are ' MAX' and ' MIN' which specify the program to take the maximum or minimum of the specified parameter values respectively.

s A-14 4

4 i

_ - . . , .. , _ _ , , --,..m ,-,t .-, --w+---rw,ev--- s

Table A.3 (cont.)

D2 tailed Discussion of Sanple Input Line in Table A.2 Input Description 226-229 The outcome of the above mentioned manipulation of parameter values is to be compared to a normally distributed reference parameter, hence the ' NORMAL' descriptor in the input. (In this case the reference parameter is the containment failure pressure.) There are 2 reference parameters, each of'which has a value specified for the three different walkthroughs. For this case, these are the mean and the standard distribution of the normal distribution.

! Another possibility for this type of question is a threshold comparison in which case ' THRESH' would appear instead of ' NORMAL' and only one reference parameter would need to be specified, namely the value of the threshold. The titles supplied in line 229 are for clarification purposes and are not used by the program.

230 Problem title. (Will be printed in the output.)

231 Titles for the three different walkthroughs. (Will be printed in the output.)

232 The number of binning " dimensions" is specified to be

2. The titles of these " dimensions" are CF and Sprays.

233 The first dimension consists of 5 bins. 5 " pathways" for binning into these five bins are specified in the input. (The number of pathways must always equal or exceed the number of bins unless a vacuous bin has been defined.) The bin titles for this dimension are EEO, eel, EE2, EE3, and EH.

234-236 These lines describe the first " pathway". There are 2 question-branch pairs specified for this pathway and any outcomes of the tree matching this " pathway" criterion are to be place in bin number 1. The question branch pairs are branch 4 of question 1 and branch 2 of question 8.

237-248 Specification of the remaining four pathways. The first three of which utilize 2 question-branch pairs and the last only 1 pair.

249 Dimension two consists of 2 bins and has 2 " pathways" specified. The titles are esp and noESp.

A-15

Tablo A.3 (cont.)

Detailed Diocussion of Staple Input Line in Table A.2 Input Description 250-252 There is one question-branch pair specified for the first " pathway". Any outcomes of the tree which match this outcome will be placed in bin number one of this dimension.

253-255 The second and last " pathway" specification. Outcomes matching this criterion will be placed in bin number 2 of this dimension.

i A-16

APPENDIK B CONTAINMENT EVENT TREE FOR SEQUOYAH This appendix describes the containment event tree that was developed for investigation of the Sequoyah plant in the SARRP l

program. Table B.1 lists all of the questions in the order in which they appear in the event tree. Any dependencies upon prior questions in the tree are listed in the table for each question. Under each question is a description of each branch corresponding.to the branch descriptors found in the tree. A description of each of the questions and the branch point probabilities follow the table. The complete event tree input is provided in Appendix C.

i Table B.1 1

1 LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE

! AND THE DEPENDENCY UPON PRIOR EVENTS a

Prior Question Question Asked on Event Tree Dependencies

1. Is ac power available early? None E-Ac AC power is available upon initiation of the scenario.

EfAc AC power is not available upon initiation of the scenario.

l 2. Is there pre-existing leakage or isolation None failure?

EE3 Leakage present at the onset of the scenario large enough that it is equivalent to a gross rupture of containment.

EE2 Leakage present at the onset of the scenario large enough to preclude gradual overpressurization but too small to depressurize containment in a relatively short time.

EE1 . Leakage present at the onset of the scenario larger than design leakage but too small to preclude gradual overpressurization.

EEO Leakage present at the onset of-the scenario is equivalent to design ,

leakage or smaller.

i i

l l

J- B-1

.-- . - ~ . - . - - , , . _ , .- - - - . - . . - - - . , , _ . . _ _ _ . - . . . , . . . . - . _ _ . - - , . - - . , - . . - . . . _ - - ,

i Table B.1'(cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Question Aske. Event Tree Dependencies

3. What is the initial break location? None EERPV The location of loss of coolant from the primary system at the onset of the scenario is the reactor pressure. vessel.

EEHL The location of loss of coolant from the primary system at the onset of the scenario is the hot leg. i EECL The location of loss of coolant from-the primary system at the onset of the scenario is the cold leg.

EESGTR The location of loss of coolant-from the primary system at the onset of the scenario-is.the steam generator.

EEV Loss of coolant from.the primary system at the onset of the accident is via an

interfacing icw pressure system.

l EEPORV At the onset of the accident coolant is lost from the primary system via the.

PORV.

4. What is the initial break size? None EELg The loss of coolant from the primary system at the outset of the scenario occurs at a rate equivalent to a large-

- break LOCA.

EES1 The loss of coolant from the primary system at the outset of the scenario occurs at a rate equivalent to an St LOCA (2-6 inch diameter).

EES2 The loss of coolant from the primary system at the outset of the scenario occurs at a rate equivalent to an S2 LOCA (1-2 inch diameter).

EES3 The loss of coolant from the primary system at the outset of.the scenario-

! occurs at a rate equivalent to an S3 LOCA (<1 inch diameter).

I i

B-2 I

i

, ~ _ _ , - - . . . _ . _ . _

,_m., . . . , . . - _ _ , _ , , , , - . _ , _ - _ , , . , , _. , . , . .

Table B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question l Question Asked on Event Tree Dependencies

5. Is containment initially bypassed? 3 EEB Containment is bypassed at the onset of the scenario.

noEEB Containment is not bypassed at the l onset of the scenario.

l

6. Are the steam generators wet or dry? None SGWet The secondary side of the steam generators is wet at the onset of the scenario and continue to remain so l (some form of feedwater is available) at least until the time of vessel breach.

SGDry The secondary side of the steam generators is dry at the onset of the scenario or dry out before the time of vessel breach.

7. Are the air return fans operating initially?. 1 EEFan The air recirculation fans are operating at the beginning of the scenario.

EEaFan The air recirculation fans are not operating at the beginning of the scenario but are capable of doing so some time in the future if called on and ac power is available.

EEfFan The air recirculation fans have failed in a manner that they cannot operate at any future time in the scenario.

8. Does emergency core cooling operate in the 1 injection mode?

ECCInj Emergency core cooling operates in the injection mode at the onset of the scenario, noECCInj Emergency core cooling fails to operate in the injection mode at the onset of the scenario.

l u

) B-3 4

- - - .-- . , - . - . . _ ~ . _ _ , , . _ _ _ _ - , - ,

Table B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Cuestion Asked on Event Tree Dependencies

9. Do the sprays operate in the injection mode? 1 EESpInj The containment sprays opetate in the injection mode at the onset of the scenario.

EEaSpInj The containment sprays do not operate in the injection mode at the onset of the scenario but are available to operate later in the scenario.

EEfSpInj The containment sprays do not operate in the injection mode at the onset of the scenario and cannot operate at any time later in the scenario.

10. Do the sprays operate in the recirculation mode? 9 EESp The containment sprays operate in the l recirculation mode very early in the scenario.

EEaSp The containment sprays do not operate in the recirculation mode very early in the scenario but are available to do so later in the scenario.

EEfSp The containment sprays have failed and are not available for operation at any times later in the scenario.

11. To what degree is there breakthrough or bypass 2,3,6 of the auxiliary building?

EESB3 Any release from containment very early in the scenario completely bypasses the auxiliary building.

EESB2 Releases from the containment very early in the scenario either partially bypass the auxiliary building or take a path which minimizes their mitigative capability.

EESBl Releases from the containment very early in the scenario take tortuous paths through the auxiliary building, maximizing the mitigative effect on-the source term.

B-4

Table B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT. TREE AND THE DEPENDENCY UPON PRIOR EVENTS

, Prior Question j Ouestion Asked on Event Tree DeDendencies l 12. What is the location of any induced failure in 4,6 the primary system.

EfRPV The temperature induced failure in the reactor coolant system occurs in the reactor pressure vessel. .

.EfHL The temperature induced failure in the reactor coolant system occurs in the hot leg piping.

i EfCL The temperature induced failure in the '

! reactor coolant system occurs in the cold leg piping.

ESGTR The temperature induced failure in the reactor coolant system occurs in the

steam generator, noEfRCPB No temperature induced failure in the reactor coolant system occurs.

13. What is the effective size of any temperature- 12 induced primary system failure?

ELg The rate of loss of primary system coolant through the temperature induced failure is equivalent to a large LOCA.

ESL The rate of loss of primary system coolant through the temperature

, induced failure is equivalent to an St LOCA.

ES2 The rate of loss of primary system coolant through the temperature induced failure is equivalent to an S2 LOCA.

noEfRCPB No temperature induced failure in the reactor coolant system pressure boundary occurs.

14. What is the level of early containment bypass 3,4,12,13 during meltdown?

EB2 Containment-bypass early in the j scenario is of a level large'enough to

!. prevent gradual overpressurization of I

containment.

. EB1 Containment bypass early in the i

scenario.is not of a level large l enough to prevent gradual i overpressurization of containment.

l EBO There is no early containment bypass.

B-5

Table B.1-(cont.)

LIST OF QUESTIONS.FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Question Asked on Event Tree Dependencies

15. Do the igniters operate early in the scenario? 1

EIg. Igniters are on during the early stages of the scenario, noEIg Igniters are not on during the early stages of the scenario.

16. Is the lower compartment inert during core 7 degradation?

ELCIn The lower compartment is inert during the early stages of the scenario.

noELCIn The lower compartment is not inert during the early stages of the scenario.

i

, 17. What is the magnitude of the early baseline 7 l pressure in containment?

, EPBase The baseline pressure in containment during the early part of the scenario, i

j 18. Does a local detonation occur in the ice 7 condenser? If so, what are the effects?

EDt-EH A detonation occurs in the ice j condenser early in the scenario which

{ fails the containment.

i EDt-f1C A detonation occurs in the ice i condenser early in the scenario which

does not fail containment but prevents the ice condenser from performing its j

function.

noEDt No detonation occurs with sufficient magnitude to produce either j containment failure or failure of the

ice condenser function.

19. Is there a deflagration in the lower compartment 4,5,12,13, before vessel breach? Also, what is the 14,15,16 magnitude of the pressure rise from such a burn?

ELCDef There is a hydrogen deflagration in the lower compartment early in the scenario, noELCDef There is not a hydrogen deflagration in the lower compartment early in the scenario.

DPELCDef The pressure rise in containment associated with the lower compartment deflagration.

B-6 i

- , ,---,.n,r,.,,,--- n - - - - , , - -. , c., . - - , , .-..n--,,,-----.-,,,-e- ~ - , . - - , , < , - ~ - -

Table B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE i AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Question Asked on Event Tree Dependencies

20. Is there a deflagration in the. upper compartment 4,5,13,14, before vessel breach? Also, what is the 15,19 magnitude of the pressure rise from such a burn?

EUCDef There is a hydrogen deflagration in the upper compartment early in the scenario.

noEUCDef There is not a hydrogen deflagration l in the upper compartment early in the scenario.

DPEUCDef The pressure rise in containment attributable to the upper compartment I deflagration.

21. Does containment fail due to an early 17,19,20 4

deflagration? Also, what is the mean contain-

] ment failure pressure and the standard deviation?

EH Containment fails early in the scenario due to a deflagration in one

- or both of the upper and lower compartments, noEH Containment does not fail due to deflagrations in containment.

22. To what degree is there breakthrough or bypass 11,18,21 of the auxiliary building early in the scenario?

ESB3 Releases from contaiment in the early 1 portion of the scenario completely bypass the auxiliary building.

ESB2 Releases from containment in the early portion of the scenario pass through the auxiliary building but on a path l Which minimizet the mitigative effect on the source term.

{ ESB1 Releases from containment in the early portion of the scenario pass through j the auxiliary building on a path which j maximizes the mitigative effect on the j source term.

i 4

i B-7

- _ _ _ . -_ . - . . . _ ~ . _ _ . . , . . - _ _ . . _ _ _

- . . . . . - . = . . . . - . - - - - .-

l Table B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS

+

, Prior Question Question Asked on Event Tree. Dependencies

23. Has the ice melted out of or have bypass paths 3,4,8,13,18 developed in the ice condenser before vessel breach?

EnoIce The ice is completely melted before the time of vessel breach.

EIceBP Bypass paths are established in the ice condenser before.the time of vessel. breach, preventing the ice-condenser from serving its. function.

i' noEf1C The ice condenser serves its function during the period before vessel breach.

24. What is the status of the air return fans after 7,19,20 I

hydrogen burns?

! EFan The air recirculation fans operate during the early portion of the scenario.

EaFan The fans are not operating early-in the scenario but are capable of doing so at a later time if ac power is restored.

EfFan Fans are not operating early in the scenario due to an unrecoverable failure.

I

25. What is the status of the containment sprays 10,18,21 after hydrogen burns?

Esp The containment sprays are operating in the early portion of the scenario.

EaSp The containment sprays are not operating in the early portion of the scenario but are available to operate

, at some later time.

EfSp The containment. sprays a.e not operating in the early portion of the  ;

i scenario due to an unrecoverable

failure.

l i

B-8 l

i.

, , , - . , . - . , - - - .- . - - . . - - - - - . , , - - . . . - ~ - - - - - , . - - . - . - , - , . ,, . , -

Table B.1 (cont.) l l

LIST OF-QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Question Asked'on Event Tree Dependencies

26. What is the primary system pressure during core 4,12.23,24 degradation? Also, what is the pressure rise in containment due to primary system depressuri-zation?

hip The primary system is at high pressure

(>2000 psig) during the period of core degradation.

imp The primary system is at some intermediate pressure (g500-1000 psig) during the period of core degradation.

1 lop The primary system is at low pressure

(<500 psig) during the period of core degradation.

27. What is the mode of reactor vessel breach. 26 alpha An in-vessel steam explosion occurs which fails the upper-head of the 3

reactor pressure vessel and proceeds j to fail containment via a missle f projectile.

4 SE-UH An in-vessel steam explosion occurs which fails the upper head but the i containment remains intact.

SE-BH An in-vessel steam explosion occurs

i. which fails the bottom head of the reactor pressure vessel (containment remains intact).

PEj The bottom head fails due to attack by the hot debris and the pressurized primary system acts to drive the debris out as a jet.

Pour The bottom head fails.due to attack by the hot debris and the debris falls into the reactor cavity under the force of gravity.

28. Is the reactor cavity wet at vessel breach? 8,9 RCWet The reactor _ cavity is flooded at the i

time of vessel breach.

RCDry The reactor cavity is dry (or contains little water) at the time of vessel breach.

B-9 i

i I

_ _ _ _ _ - - _ . , _ - . _--.,_ _ _ - - . _ ~ , . _ _ __ .- ,. _ ,

Table B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Question Asked on Event Tree Dependencies

29. Does direct heating occur? Also, what is the 23,26,27,20 pressure rise in containment due to direct heating and/or a steam spike?

DH Direct heating of the containment atmosphere by debris ejected from the reactor pressure vessel occurs in I conjunction with a steam spike. l noDH The direct heating phenomenon does not I occur -- only a steam spike.

DPDHSS The pressure rise in containment due to direct heating and a steam spike.

30. Does a detonation occur in the ice condenser 18,23,24 at the time of vessel breach? Also, what are the effects?

IDt-IH A detonation occurs in the ice condenser at the time of vessel breach which fails containment.

IDtf1C A detonation occurs in the ice condenser at the time of vessel breach which does not fail containment but which defeats the ice condenser function.

noIDt No detonation of sufficient magnitude to cause containment failure or prevent the ice condenser function occurs at the time of vessel breach.

31. Hds the ice nelted out of or have bypass paths 3,4,8,13, developed through the ice condenser at the time 18,23,30 of vessel breach?

InoIce The ice is completely melted at the time of vessel breach.

IIceBP Bypass paths have developed in the ice condenser by the time of vessel breach preventing the function of the ice condenser, nolf1C The ice condenser serves its function at the time of vessel failure.

l B-10

Table B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Question Asked on Event Tree Dependencies

32. Does a hydrogen deflagration occur at the time 4,13,14 of vessel breach? Also, what is the pressure 15,19,20 rise due to the deflagration?

IDef A hydrogen deflagration occurs at the j time of vessel breach.

. noIDef No hydrogen deflagration occurs at the time of vessel breach.

DPIDef The pressure rise in containr.ent due to a hydrogen deflagration at vessel breach.

I

33. Does containment fail at vessel breach? Also, 17,26,29,32 what is the mean failure pressure and standard deviation?

CF@VB Containment failure occurs at the time

! of vessel breach.

j noCF@VB Containment remains intact after the processes at vessel breach.

34. What is the mode of intermediate containment 2,18,21,27 failure? 29,30,32,33 IA Containment failure at the time of vessel breach occurred due to an in-vessel steam explosion.

ID Containment failure at the time of vessel breach occurred due to. direct heating.

IHS Containment failure at the time of vessel breach occurred due to the combined effects of a hydrogen deflagration and a steam spike.

IS Containment failure at the time of t

vessel breach occurred due to a steam i spike.

I noICF Containment did not fail due to the j processes at the time of vessel breach.

a l

i B-11 i

I Table B.1 (cont.)

! LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS i

Prior Question Question Asked on Event Tree DeDendencies j 35. To what degree is there intermediate auxiliary 22,30,34 l building breakthrough or bypass? )

t ISB3 The releases from containment at the I time of vessel breach completely i

bypass the auxiliary building.

ISB2 The releases from containment at the time of vessel breach pass through the

< auxiliary building but on a path which minimizes the mitigative effect on the

source term.

I ISBl The releases from. containment at the

! time of vessel breach pass through the auxiliary building and on a. path which maximizes the mitigative effect on the

source term. ,

1 l 36. What is the status of the air return fans 24,32 i after vessel breach?

I IFan The air recirculation fans are

operating after the time of vessel breach, g

a IaFan The air recirculation fans are not i operating after the time of' vessel

! failure but are available to do so at j some later time.

j IfFan The air recirculation fans are not operating after the time of vessel breach due to some unrecoverable failure.

j 37. What is the status of the containment sprays 25,34 after vessel breach?

ISp The containment sprays are operating j after vessel breach.

IaSp The containment sprays are not

! operating after vessel breach but are available to do so at some later time. '

i IfSp The containment sprays are not operating after vessel breach due to some unrecoverable failure,

{

i i

B-12 I

}

- . . - - -.,-,,---.,,.n,.,,--_ n . , - . , , - , . - _ , , , , , ,--,-.e- --,,.mn,- . - . , . , - , , , , . . , . , - -

Table U.1 (cont.)

LIST OF QUESTIONS-FOR THE'SEQUOYAH EVENT TREE AND THE DEPENDENCY UPON PRIOR EVENTS Prior Question Question Asked on Event Tree Dependencies

38. Is ac power restored after vessel breach? 1 L-AC AC power is available after vessel breach.

LfAC AC power is not available after vessel breach.

39. Are the igniters on late? 1,15,38 LIg The igniters are on late in the scenario, noLIg The igniters are not on late in the scenario.

40. What is the status of the air return fans late? 36,38 LFan The air recirculation fans are operating late in the scenario.

LaFan The air recirculation fans are not operating late in the scenario but are available to operate at some later time.

noLFan The air recirculation fans are not operating late in the scenario due to some unrecoverable failure.

! 41. What is the status of the containment sprays

! late? 37,38 i LSp The containment sprays operate late in 1

the scenario.

Lasp The containment sprays do not operate i late in the scenario but are available j for operation at some later time, j Lfsp The containment sprays do not operate 4

4 late in the scenario do to an unrecoverable failure.

4

42. What is the late baseline pressure in

! containment? 31,40,41 I LPBase The magnitude of the baseline pressure in containment late in the scenario, j 43. Is a coolable debris bed formed and maintained 26,27,28

after vessel breach?

CDB A coolable debris bed is formed and maintained after vessel breach.

1 noCDB A coolable debris bed is not formed

and maintained after vessel breach and core-concrete interactions begin.

B-13 l i

Table B.1 (cCnt.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE

, AND THE DEPENDENCY UPON PRIOR EVENTS I

i . . Prior Question Question Asked on Event Tree. Dependencies

44. Is there a. late deflagration? Also, what is 2,19,20,21 the pressure rise associated with the 32,34,39,43 j deflagration?

j LHB A hydrogen deflagration occurs in containment late in the scenario.

noLHB .No hydrogen deflagration occurs in the containment late in the scenario.

I DPLDef The pressure rise in containment due j to a late hydrogen deflagration.

I

45. Does the containment fail due to a11 ate 2,21,34, deflagration? 42,44 LH Containment failure occurs due to a 1 late hydrogen deflagration in containment.

l noLH Containment does not fail due to any late hydrogen deflagrations.

46. What is the status of the containment sprays 41,45 very late?

LLSp Containment sprays are operating very late in the scenario.

LLaSp containment sprays are not operating very late in the scenario but'are available to operate at some later time.

LLfsp Containment sprays are not operating very late in the scenario due to an unrecoverable failure.

j 47. Does containment failure occur late due to 2,18,21,34,

! noncondensible gas and/or steam buildup? 43,46 LP Gross containment failure occurs late

! in the scenario due to gas

! accumulation in containment, noLP Containment does not fail late in the 3 scenario due to accumulation of gases i in containment.

4 I B-14 a

l

Tablo B.1 (cont.)

LIST OF QUESTIONS FOR THE SEQUOYAH EVENT TREE AND THE\ DEPENDENCY UPON PRIOR EVENTS Prior Question Question Asked on Event Tree Dependencies

j,

48. To what degree is there late auxiliary building '3s,45,47 breakthrough or bypass?

LSB3 The releases from containment late in cf l ,the scenario completely bypass the-l auxiliary building. -

LSB2 The releasec from containment late in the scenario pass through the secondary the auxiliary building but l on a path that minimizes the mitigative effects on the source term.

LSB1 The releasesstrom; containment late in the scenario p' ass through the secondary containment and/or the auxiliary building and on a path that ' '

N maximizes the mitigative effects on ,

the' source term. ,

~1 4

49. Does basemat meltthrough occur? 43,46 i MT Meltthrough of the basemat by the core  !

debris occurs, noMT Meltthrpugh of the basemat by the core ,

debris does not occur. i

, Each of the questions is now discussed in detail. A i

discussion.of the branch point probabilities and the parameter values is provided under each question heading.

s i

The outcomes j  !

and probabilities for some questions depend on the type of -l  !

accident entering the event tree, defined by the plant damage 7 states. For a description of these damage states, refer back to Section 2.4.2 of the main report.

QUESTION 1: Is ac power available early?

The availability of ac power is clearly important in 4

assessing the status of the containment safety features. The answer to this question is defined by the initiating sequence. . h.

43-QUESTION 2: Is there pre-existing leakage or isolation failure? '

The first failure mode of containment we consider is the r I existence of a leakage path from containment prior to the core melt or, similarly, the failure of containment isolation to be effected during the sequence. Such leakage' paths may-lead to a k 6 -

I 1,

.i ,

'!  ?

B-15 4 s

..._,-_n,_-_--.. -_,.---_'.

't i

direct release of fission pecducts. In.eddition, if the leak is, big enough, it.may provide a relief path, precluding:

overpressurization of containment.

We' considered four levels of containment leakage to be.of interest:

(1) Leakage within limits imposed by technical specifications (i.e., no. failure of isolation or

) significant leakage path):

(2) Leakage greater than that allowed by technical specifications but insufficient to preclude gradual,

, ( long-term overpressurization:

(3) Leakage sufficient to preclude gradual '

overpressurization, but insufficient to depressurize containment in a relatively short period of time: and (4) Leakage sufficient to depressurize containment i relatively quickly.

.A The Reactor Safety Study (RSS, Ref. l probability of isolationfailureof2x10gestimateda . The major i contributor was inadvertent opening of a containment purge 2

valve with coincident failure of the radiation interlock. The t Reactor Safety Study Methodology Applications Program (RSSMAP, i Ref. 2) increased the probability to 7x10-3 to reflect concerns about a recently discovered large leak at another plant. On the other hand, the Industry Degraded Core (IDCOR) program estimated coincident opening of double purge valves to be no greater than 10-6 We have taken the probability of isolation failure to.be negligible (optimistic) or the RSS value (central and pessimistic).

Weinstein (NRR, Ref. 4) studied the frequency of observed leaks at nuclear power plants based on licensee event reports.

We analyzed Weinstein's data assuming a binomial distribution for leak occurrence. The analysis yielded 5th, 50th, and 95th percentiles of 0.017, 0.033, and 0.058 for small leaks -

(equivalent to level 2 above), and 4x10-5, 5x10-4 and 2x10-3,for large leaks (i.e., level 3), respectively.

\

We-have taken the combination of large preexisting leaks and isolation failure for the branch point probabilities on this question and the analysis of the Weinstein data for small leaks. This produced the following results.

3 .

i B-16 "J

'l i e

, \

i

..n.. . - , . . - . , -.. - - . . - , . , , , - , , . . . ... , . , , _ . . , ,

Branch Point Probabilitice "

Optimistic Central Pessimistic level (1) 0.983 0.994 0.938 level (2) 0.017 0.033 0.058 level (3) negligible 0.0025 .004 level (4) negligible negligible negligible QUESTION 3: What is the initial break location?

The location of theinb;ialRCS break determines in part the pathway take'n by. fission products escaping from the RCS, and consequently the,Dagnit'ude of the source term. Examination of the piping and insttumentation diagrams for a similar l reactor showed approximately equal numbr.rs of connections in the range of 1-2" on the hot and cold legs. We have estimated the proba- bility of a pipe break in the S2 size range to be approximately equal for the hot and cold legs. S3 LOCAs (1/2

- 1" diameter or equivalent) have been taken to be pump seal failures and are hence always on the cold leg. Nearly all St size connections (2-6" diameter or equivalent) are on the cold leg. Thus given a break equivalent to St, the location was ..

taken to be the cold leg with unit probability. Large LOCAs were assumed to occur in either the cold leg or the hot leg with equal probability.

QUESIION l' What is the initial break size?

?

~

The sine of the initial RCS break determines.the pressure in thc-priqary at the time of core melt and the relative timing of the'dispharge of the cold leg accumulators, which in' turn affect both the behavior of the core debtis during and follow-ing the reactor vessel breach and the discharge of fission products to the containment atmosphere. The initial break size in the RCS is defined by the initiating sequence.

QUESTION 5: Is centainment initially bypassed 0 The answer t'o this question is defined by the break loca-tion. If the,, initial break location is in the steam generator' or in an interfacing system, the containment is bypassed.

Otherwise the containment is not bypassed.

QUESTION 6: Are the steam generators wet or dry?

The presence of water in the ,steamm generators can have several effects on the accident progression: (1) if there is water in the steam generators, the likelihood of a temperature B-17 l

induced tubo rupture is substentially reduced; (2) if the ecqusnco involvos a tubs rupturo, water in the steam generators will serve to mitigate the resulting release; and (3) water in the steam generators will produce a large, cold surface area upon which fission products transported around the primary system may plate out, thereby reducing the source term. The question here is whether the secondary side of the steam generators is wet or dry during core degradation. This is defined by the initiating sequence. If auxiliary feedwater is available the steam generators are wet. Otherwise, they are dry.

QUESTION 7: Are the air return fans operating initially?

The set point for actuation of the air return fans is 3 psig within containment (FSAR, Ref 5). From analyses performed for Sequoyah (IDCOR, Ref. 3: SASA, Ref. 6), it was determined that this threshold would always be reached when there was any blowdown to containment. Thus the only possibility of the air return fans not operating would be due to loss of ac power. If ac power is available initially in the accident (question 1),

the fans are assumed to operate with unit probability.

Otherwise they are assumed not to operate but remain available for later operation should power be restored.

QUESTION 8: Does emergency core cooling operate in the injection mode?

The operation of emergency core cooling in the injection mode, followed by subsequent failures that lead to core melt, implies that the ice in the ice condenser may have melted long before the time of vessel breach. Whether or not the injection systems operate also affects the amount of water in the reactor cavity at the time of vessel breach. If ac power is not available, emergency coolant injection is not permitted.

Otherwise, whether or not it succeeds is dictated by the plant damage state.

QUESTION 9: Do the sprays opotate in the injection mode?

The set point for actuation of the containment spray system is 3 psig within containment (FSAR, Ref. 5). From analyses performed for Sequoyah (IDCOR, Ref. 3: SASA, Ref. 6), it was determined that this threshold would always be reached when there was any blowdown to containment. Thus the only possibility of the sprays not operating would be due to loss of ac power or mechanical fTilure. If ac power is available initially in the accident (question 1) the sprays are assumed to operate unless specified as failed by the plant damage state. If ac power is not available and the sprays are not specified as failed in the initiating sequence, they remain available for operation later in the scenario if ac power is restored.

B-18

QUESTION 10: Do tha-aprays operate in the recirculation Eods?

If the sprays operate in the injection mode (question 9).

and the initiating sequence does'not specify them to-fail in the recirculation mode, continued operation is assumed.

Similarly, if the sprays were initially available but not operating (question 9) and the initiating sequence did not include failure in the recirculation mode, they were considered available in the recirculation mode. Otherwise, the sprays were considered inoperable in the recirculation mode.

QUESTION 11: To what degree.is there breakthrough or bypass of l the auxiliary building?

If there was preexisting leakage or isolation failure, but no containment bypass, it was considered "likely" in all three walkthroughs that the release from containment would follow an

. extensive and tortuous path through the auxiliary building. A

! partial bypass of the auxiliary building was considered "unlikely" and a total bypass was considered " impossible".

If containment bypass existed in the form of a steam generator tube rupture but the secondary side of the steam  ;

generator was filled with water, the same degree auxiliary building bypass was assumed as described in the previous paragraph. (It was assumed that the water in the steam generator would have a similar effect on the fission product release as the auxiliary building.) However, if the steam generator secondary were dry, the equivalent of a total bypass of the auxiliary building was assumed.

If the scenario involved an interfacing systems LOCA, it was considered "likely" in all three walkthroughs that the release from containment would follow an extensive and tortuous path through the auxiliary building. A partial bypass of'the auxiliary building was considered "unlikely" and a total bypass j was considered impossible.

Later questions in the tree dealing with bypass are dependent upon the answer to this earlier question. Any later

level of bypass cannot be less than any previously established

! level of bypass. Thus all remaining scenarios were assigned to the branch corresponding to maximum auxiliary building retention. i QUESTION 12: What is the location, if any, of temperature-induced failure in the primary system?

Temperature-induced failures in the reactor coolant system.

(RCS) during core degradation have not been considered in most previous risk assessments. However, recent analyses (NRR, i

B-19 1

Ref. 7) have indicated that gas temperatures as high as 2000*F may result during accidents involving a total loss of core cooling with the primary system remaining at high pressure, such as a station blackout. Similar temperatures would also be expected for very small LOCAs with failure of core coolant injection. Based on these observations, the CLWG and others have raised the possibility of induced failure in some part of the RCS.

IDCOR (Ref. 5) assumed that the induced LOCA would be a failure of the reactor coolant pump seals. Others consider a major rupture of a hot leg to be possible. We have also considered that the induced failure may be a steam generator tube rupture (SGTR); the Zion-Indian Point Study (Ref. 8) raised the possibility of induced SGTRs. There have also been reports of tube failures even without the high temperatures and pressures expected in the type of sequences of interest. On the other hand, IDCOR (Ref. 3) considered SGTRs not to be risk significant.

1 Induced failure may or may not be optimistic. If a large '

enough failure were to occur, later direct heating scenarios would be precluded, and the threat of a hydrogen burn at the time of vessel breach would also be reduced. On the other hand, there would be less retention of fission products in the RCS. Furthermore, a large induced failure in sequences involving success of emergency core cooling in the injection phase but failure in recirculation could result in depleting much of the ice in the ice condenser prior to vessel breach.

We therefore considered five outcomes for this event: (1) no induced failure. (2) induced failure in the reactor vessel, (3) failure in a hot leg, (4) cold leg failure, and (5) induced steam generator tube rupture. We considered induced failure to be possible only for sequences in which the RCS would remain at high pressure, i.e., very small LOCAs and transients with cycling relief valves.

The information available to serve as the basis for quantifying the relative likelihoods of these outcomes was quite limited. Moreover, it was not possible for us to pre-determine the degree to which induced failure might increase or decrease the consequences for each of the plant damage states. We therefore based our estimates on the values developed for the sensitivity study performed for the risk rebaselining and risk reduction effort under SARRP. These values were formulated by a group of containment experts who met to consider the most important issues. The composite values are those listed below. We used the same values for the optimistic, central, and pessimistic walkthroughs.

l l B-20 i

Probability of Induced LOCA Large Small Steam Generator No Induced Hot Leg Cold Leg Tube Rupture Failure (1) Steam generators dry,-seal cool-ing available 0.2 0.0 0.24 0.56 (2). Steam generators

dry, seal cool-l ing not avail-l able 0.12 0.40 0.14 0.34 (3) Steam generators wet, seal cool-ing available 0.26 0.0 0.0 0.74 (4) Steam generators i wet, seal cool-ing not avail-able 0.14 0.46 0.0 0.40 QUESTION 13

What is the effective size of any temperature-induced failure?

If an induced failure of the RCS were to occur, we considered its size to be dependent upon the induced failure location. The induced failure size was the same for all three l'

walkthroughs. If the induced failure occurred in the reactor vessel itself or in the hot leg, we assumed the failure size-to be equivalent to a large LOCA. If the induced failure occurred in the cold leg or the steam generator tubes, we considered it equivalent to an S 2 break size. Otherwise, no induced failure occurred.

, QUESTION 14: What is the level of early containment bypass j during meltdown?

The answer to this question is predetermined by the previous questions dealing with location and size of both the initial and any induced RCS failure. If the larger of the failures is a steam generator tube rupture or an interfacing systems LOCA~and the size is equivalent to an Si or larger LOCA, a large, early containment bypass exists. If the larger

, of the failures is a steam generator tube rupture or interfacing systems LOCA and the size is equivalent to an S2 1

or smaller, a small, early containment bypass is assumed. '

i Other cases result in no early containment bypass. )

t I

j B-21 1

- , - - - , . - . . . - . , - , - , --. -.~ --, , - . . - - - , - - - - - - - - - , , - , - - . . . -_--n-

QUESTION 15: DD tha igniterc cparate early in the scenario?

Igniter operation is dependent upon the availability of ac power and operator actuation. If ac power is not available early in the scenario, the igniters are assumed to be off but available for operation if power is restored at some later time. For the cases in which ac power is available, we followed guidelines for assessing a human error of failing to follow a procedure (1278, Ref. 9). We took the optimistic, central, and pessimistic estimates of the probability that the operator would fail to turn on the igniters to be 0.0, 0.003, and 0.03, respectively.

QUESTION 16: Is the lower compartment inert during core degradation?

Reviews of previous analyses (IDCOR, Ref. 3: SASA, Refs. 6

& 10), indicate that during core degradation the lower compartment will be inert due to high steam concentrations if the air return fans are not operating. However, if the fans are operating the lower compartment will not be inert. Thus, the branch taken at this node depends singularly upon the operation of the air return fans.

QUESTION 17: What is the magnitude of the early baseline pressure in containment?

This question involves no selection of probabilities since it corresponds to a one-branch node. The purpose of the question is to supply to the program a value of the baseline pressure in containment to be used later in calculating contaiment failure mode probabilities. A review of past analyses for Sequoyah (IDCOR, Ref. 3: SASA, Refs. 6 & 10: BMI, Ref. 24) revealed that the only factor dominant in the value for baseline pressure early in a scenario is whether or not the air return fans operate. If the fans operate, the original inventory of noncondensible gases is kept mixed throughout containment. Without the operation of the air return fans, the noncondensibles are compressed into the upper compartment and the baseline pressure is increased slightly. A small range of baseline pressures was identified from these past analyses. We took the values in the case of fans operating to be 18, 19, and 20 psig for the optimistic, central, and pessimistic walkthroughs respectively. For the case in which fans did not operate, we took the values to be 19, 21 and 23 psig for the three respective walkthroughs.

QUESTION 18: Does a local detonation occur in the ice condenser?

There are three possibilities for this question: (1) a detonation occurs in the ice condenser which fails containment, B-22

(2) a detonation cccurs in ths ice condsnsor which dose not fail containment but disrupts the internal geometry of the ice condenser such that its steam condensation function is defeated; and (3) either a detonation occurs which is of insufficient magnitude to cause (1) or (2), or no detonation occurs.

For a detonation to occur, the concentration of hydrogen must reach approximately 14%. In past analyses of Sequoyah (SASA, Refs. 6 & 10), this level of concentration was only observed when the air return fans were not operational. Thus, if the fans were operating, we took node 3 (no or non-damaging detonation) to be "certain".

If fans were not operating, the issue is more clouded. Not only must a concentration of 14% hydrogen be attained, but there must be an ignition source as well. Possible sources of l ignition include a flame propagating in from the upper plenum (with the fans off the lower compartment will be inert:.hence i

no propagation from below is credible), forming an accelerated flame which transitions to a detonation: or one of the intermediate deck doors slamming shut could produce a spark of sufficient magnitude to initiate a detonation.

In view of these questions surrounding the issue of detonation intiation and in the absence of any calculations of containment loading due to local detonations in the ice condenser, the probability of failing containment due to a local detonation was taken as " impossible", "unlikely", and

" indeterminate" in the optimistic, central, and pessimistic walkthroughs, respectively. The probability of defeating the function of the ice condenser due to a local detonation was taken as " impossible" in the optimistic case, "unlikely" in the central case, and " indeterminate" in the pessimistic case.

QUESTION 19: Is there a deflagration in the lower compartment before vessel breach? If so, what is the magnitude of the pressure rise from such a burn?

There are two aspects to this question: (1) whether or not a deflagration will occur, and (2) if one were to occur, what the resultant loading on containment would be. The answers depend upon several factors, including the RCS break size, whether the containment is bypassed, igniter operation, and lower compartment inerting. Past analyses of Sequoyah (IDCOR, Ref. 3: SASA, Refs. 6 & 10; BMI, Ref. 24) were reviewed for answers to these questions.

Characteristically different results are observed if the size of the RCS break is equivalent to an Si or larger LOCA from those obtained if it is an S 2 OE S3 LOCA. In the case B-23 l

of ths lergar break sizac, little hydrogon in hold up in the primary ayatem, and cost is coloaccd to containment (unloac containment bypass has occurred). This allows higher probabilities of reaching combustible mixtures in containment and the possibility of.large hydrogen burns. In the case of the smaller break sizes, much of the hydrogen generated during the core degradation process remains bottled up in the RCS and is not released until vessel breach. As was implied earlier, if containment bypass has occurred, any hydrogen released from RCS also bypasses containment and no threat to containment exists due to hydrogen burns (note however that containment function has already been defeated due to the existing bypass).

Igniter operation is crucial to some models of how the hydrogen will burn in containment (IDCOR, Ref. 3). If igniters do not operate, ignition of a combustible containment atmosphere is less predictable. With igniters the maximum achievable containment loading due to hydrogen deflagrations is generally considered less than if combustion is dependent upon some other (possibly random) ignition source. Combustion also requires that the lower compartment not be inert.

We considered these factors and examined prior analyses, and found five characteristically different scenarios to exist. IDCOR. SASA, and BMI have shown amounts of zirconium oxidation ranging from about 25% up to nearly 100% before the time of vessel breach. Accordingly, we took the amount of zirconium oxidation before vessel breach to be 35%, 50%, and 100% for the optimistic, central, and pessimistic walkthroughs, respectively. These cases, along with the probabilities and values of containment loading (as extracted from the previous analyses), are listed below.

1) St or larger, no containment bypass, igniters on, and lower compartment not inert.

We assumed that a lower compartment deflagration was "certain" in all three walkthroughs. The range of loadings on containment (due to the lower compartment deflagration alone) of 1, 5, and 25 psid for the optimistic, central, and pessimistic walkthroughs, respectively, is representative of previously reported results.

2) Si or larger, no containment bypass, igniters off, and lower compartment not inert.

We took ignition in the lower compartment to be

" impossible" (insufficient hydrogen is released to meet the flame temperature criterion for ignition assumed for the optimistic case) in the optimistic case and "likely" in the central and pessimistic cases. We took the representative B-24 l

range of loads on contain ent given ignition (only possible in the centrsl and pessimistic cases) to be 25 and 37 paid for the central and pessimistic cases, respectively.

3) S2 OE S3, no containment bypass, igniters on,.and lower compartment not inert.

We considered ignition in the lower compartment to be "certain", " indeterminate", and "certain" for the

optimistic, central, and pessimistic walkthroughs l respectively. The reason that ignition would be assured in i the optimistic case is due to the consideration of localized combustion in the vicinity of the igniters as the model for combustion. In the central and pessimistic cases a compartmentally global combustion criterion is assumed i and the variation in amounts of zirconium oxidation leads to marginally combustible concentrationc of hydrogen in the central case, while combustible levels are definitely l reached in the pessimistic case. The resulting loads due to the deflagration in the lower compartment only of 1, 4, and 6 psid for the optimistic, central, and pessimistic walkthroughs, respectively, are representative of previously reported results. The pressure rises are much lower in the small break cases (than in the previously

] described large break cases) since muc h of the hydrogen 1 remains bottled up in the RCS until vessel breach.

] 4) S2 OE S3, no containment bypass, igniters off, and lower compartment not inert.

, We took ignition in the lower compartment to be j " impossible" in the optimistic case (insufficient hydrogen is released to meet the flame temperature criterion for

ignition which was assumed for the optimistic case). In I i the central and pessimistic cases it is marginal whether enough hydrogen is released prior to vessel breach to reach a combustible levels without igniters. Since it is more pessimistic to delay combustion until vessel breach, when additional large amounts of hydrogen are released, we took ignition as "unlikely" and " remotely possible" for the central and pessimistic walkthroughs, respectively. The representativt pressure loadings on containment (no ignition occurs in the optimistic case) were 6 and 8 paid for the central and pessimistic walkthroughs, respectively.

4

5) Other cases I

The remaining cases involve either the lower l compartment being inert or the release of hydrogen bypassing containment. Thus, ignition is not possible and no pressure loadings occur due to combustion.

1

! B-25 1

l t

1 QUESTION 20: Is there a deflagration'in the upper. compartment before vessel breach? Also, what is-the magnitude of the pressure rise fron'such a burn?

. This question is very'similar to the previous one and the reader is referred to the~ discussion preceding the individual case descriptions for details. The probabilities and j '- containment loadings are_ described below on a case-by-case

basis.

3 1) Air return fans do not operate, and no lower compartment

j deflagrations. occur.

i j In this case there is-no motive force to move hydrogen j from the lower compartment to the upper compartment since 'l the ice condenser condenses-the steam out of the. gas mixture flowing from the lower to upper compartment. Thus j

i

ignition was considered impossible in the upper compartment and pressure loadings were not required.

2) Si or larger, no containment bypass, and the igniters on.

In the optimistic and. central cases, insufficient hydrogen is available to render the upper compartment combustible. This is not'true in.the pessimistic case.

• Therefore, we considered ignition to be " impossible" in the optimistic and central cases and "certain" in the i pessimistic case. The pressure loading was not required for the optimistic and central cases and 25 psid was used 4

in the pessimistic case as being representative of results from a prior analysis (SASA, Ref. 11).

I

3) Si or larger, no containment bypass, and the igniters off.

2 This is the same as the previous case, except that the 4

igniters are off. This makes ignition'in the pessimistic case less clear, but it would probably-occur at higher j hydrogen concentrations if it-did occur. Ignition whs

, still " impossible" in the optimistic and central cases but was only considered "likely" in the pessimistic' case rather than assured. The pressure loading was not requirec in the optimistic and central cases and 32 psid was used in the pessimistic case (slightly higher than with no igniters due i to ignition occurring at higher concentrations of hydrogen) i as a' representative value.

I

! 4) S2 or S3 J

b In the case of small break LOCAs, insufficient i hydrogen is released before vessel breach to render the 1 upper compartment combustible. Thus, we took ignition'tr

) be " impossible", and values of containment loading were .mt

! required.

i r

B-26 i

.. ~ . - - . . .- .- - - _ _ . _ _ _ _

l l

l

5) Other cases All-the remaining cases involve containment bypass so that no hydrogen-is released to containment, and there is no chance of establishing a combustible containment atmosphere. Ignition was precluded for these cases, and no pressure loadings are produced.

QUESTION'21: Does containment fail due to an early deflagration?

At this node the containment event tree program calculates the probability of failing containment early due to a hydrogen l

burn by summing the values of baseline pressure. pressure rise l

~

due to a lower compartment deflagration, and pressure rise due to an upper compartment deflagration, and comparing the sum to a normal distribution for containment capacity. The data-supplied for the capacity consist of a mean failure pressure i

for containment and a standard deviation. We drew upon published values of containment failure pressure (IDCOR, Ref. 3: AMES, Ref. 12; SER, Ref. 13). We took the'mean failure i pressure to be 75, 65 and 51 psia for the optimistic, cen'ral-

! and pessimistic walkthroughs, respectively. A standard j deviation in containment failure pressure was supplied for the i analysis from which we took the optimistic mean failure i pressure. For the central and pessimistic cases the difference between the pessimistic mean failure pressure and the proof i pressure was taken to represent four standard deviations (a l probability of failure less than .0001). Thus the values of l standard deviation used were 7.5 psi for the optimistic case i and 6 psi for the central and pessimistic cases.

QUESTION 22: To what degree is the auxiliary building bypassed

, early in the scenario?

The answer to this question is dependent upon previous levels of auxiliary building bypass. The degree of bypass'can j never decrease as the accident progresses. It can, however, j increase due to some additional mode of containment failure.

l This question addresses the possibility of auxiliary building i bypass being induced by early containment failure'due to a hydrogen deflagration or detonation. If containment fails due to an early hydrogen deflagration, the probability of total bypass is considered to tHe " indeterminate," "likely," and j "almost certain" for the optimistic, central, and pessimistic i walkthroughs, respectively. If total bypass of the auxiliary

} building does not occur, we assumed a partial bypass (or

• i breakthrough) when containment fails due to an early hydrogen deflagration.

i i l

I 1

l B-27 l l l

l l C_ . _ - . . _ _ _ _ _ _ . _ _ . . . _ _ . _ . _ . - _ . . . _ - . _ _ . _ _ _ . _ . _ ,_ __._ _ _

In the caso of hydrogen detonations which fail containment, we made slightly different acsumptions. In the optimistic case, we consider a total bypass to be " impossible," and a partial bypass "unlikely." In the central case, a total bypass is considered " indeterminate" with the remainder being split evenly between partial and no bypass (unless, of course, some prior level of bypass had been established, in which case the outcome is limited to something equal to or worse than the original level of bypass). In the pessimistic case we considered complete bypass of the secondary containment to be "almost certain," with the remainder going to a partial bypass.

1 No other early containment failure modes are feasible for Sequoyah. Thus, all other cases are placed in the "no-bypass" category.

QUESTION 23: Is the ice melted out of or have bypass paths developed through the ice condenser before vessel breach?

One way of developing bypass paths through the ice condenser is via a local detonation which disrupts the internal geometry of the ice condenser. This has been discussed previously under question 18. If this occurs we assessed that an early bypass path would be formed.

In sequences in which emergency core cooling is initially successful in the injection mode, significant ice melt can occur before the core is even uncovered. From past analyses, (IDCOR, Ref. 3; SASA, Refs. 6& 10), from 60 to 90% of the ice had melted by the time of vessel breach for break sizes in the primary system equivalent to S2 and St LOCAs. Typically, the flow through the ice condenser is considered to be axisymetric. However, substantial concentration and density gradients are apt to be present in the ice condenser due to the condensation of steam. In fact, these gradients may lead to asymetric flow in the ice condenser. This will in turn lead to asymetric ice melting and the possibility for developing channels through the ice condenser which in effect bypass the ice condenser and defeat its steam condensation function (SASA, Ref. 25). No test data are available for ice condenser i l

performance under conditions other than those experienced during design basis accidents.

We considered the possibility of defeating the function of the ice condenser if emergency core cooling functioned in the injection mode as a function of primary system break size. For an 31 break size as much as 90% of the original ice inventory has been calculated to be melted by the time of vessel failure (GASA, Ref. 6); thus, we considered the defeat of the ice condenser function in this scenario to be " impossible,"

B-28

" indeterminate," and "likely" in the~ optimistic, central, and pessimistic walkthroughs, respectively. For an S2 break size, fractions of'the ice melted by the time of vessel breach are calculated to be approximately 60%. In this case, we considered the probability of defeating the ice condenser function by preferential melting to be " impossible," " remotely possible," and "unlikely" in the optimistic, central, and pessimistic walkthroughs.

Recent MARCH calculations performed by Battelle have shown that a large fraction of ice would be melted before vessel breach for sequences in which the initial RCS breach is a cycling PORV, followed by a large induced LOCA. In this case, we estimate bypass to be " impossible" (optimistic),

" indeterminate" (central), or "unlikely" (pessimistic). We considered defeat of the ice condenser function by melting to

be " impossible" (optimistic and central) or "unlikely" j (pessimistic).

For all other cases the probability of defeating the

function of the ice condenser was considered " impossible" in both the optimistic and central cases and "unlikely" in the

pessimistic c'se.

QUESTION 24: What is the status of the air return fans after hydrogen burns?

We considered the possibility that a global burn in i containment could render the air return fans inoperable. The l air return fans take suction from the upper compartment and i discharge into the dead-ended annulus region. The fans are of the axial flow type and are fitted to ductwork which serves as the collection and distribution system. In the ductwork are j gravity actuated dampers to prevent backflow through the fans when they are not operating, averting bypass of the ice j condenser. It is conceivable that the overpressure from a 1

global burn could defeat the air return fans by collapsing j ductwork, bending fan blades or sticking dampers shut.

4 However, no definitive work exists in this area and there do i exist two completely independent systems, so that fan failure j is improbable. Since the system takes suction from the upper 4

compartment, we felt that damage was more likely to result from a deflagration in the upper compartment.

1 For a deflagration in the upper compartment, we took the

failure of the air return fans to be " impossible" (optimistic),

j " remotely possible" (central), and "unlikely" (pessimistic).

j' For the case of a lower compartment deflagration, failure of the air return fans was considered " impossible" (optimistic and l central) or " remotely possible" (pessimistic).

! l l

i, B-29 i

QUESTION 25: What is the status of containment sprays after hydrogen burns?

We assumed that the only method for failing the containment sprays due to hydrogen combustion was if containment failure resulted from the combustion. This could result from either tearing of piping or flashing of sump water. RSSMAP (Ref. 2) took spray failure as certain at the time of containment failure. However, we believed this to be very conservative.

In this analysis, we took spray failure as a result of containment failure due to a hydrogen burn to be "unlikely" for the optimistic and central walkthroughs and "likely" for the pessimistic walkthrough. We believed that a detonation in the ice condenser which failed containment would be a more localized event and less likely to produce containment spray failure. We took spray failure in these cases to be

" impossible" for both the optimistic and central walkthroughs and " remotely possible" in the pessimistic case.

QUESTION 26: What is the primary system pressure during core degradation?

This question not only specifies the level of pressure in the primary system at vessel breach but also supplies the amount of pressure increase in containment due to vessel depressurization. The level of primary system pressure is solely determined by the size of the break in the primary system at the time of vessel breach. An S3 size break results in high pressure (no accumulator dump) at vessel breach. An S2 size break results in intermediate pressure (near the accumulator set point) at the time of vessel breach, and any larger size results in low primary system pressure (about 100 psia) at the time of vessel breach.

The increase in pressure in containment due to vessel breach depends on three conditions: the primary system pressure at vessel breach, the status of the ice condenser, and the status of the air return fans. If the ice condenser has been defeated (either by ice melt or bypass) the steam released from the primary system will pressurize the containment.

Otherwise, the steam will be condensed in the ice condenser.

If the ice condenser functions and the air return fans are operating at the time of vessel breach, noncondensible gases in the lower compartment will be purged into the upper compartment and some incremental pressurization will result. The effects of the noncondensibles were ignored for cases in which the ice condenser was not effective in condensing the steam from the RCS.

Table B.2 shows the pressure increases which were used in the tree for the different situations described above. The pressure increments were developed from MARCH calculations (Ref.6) and MAAP calculations (Ref.3).

B-30

.--. - -. - . ~~ ,. . - _ - -

Table B.2 PRESSURE INCREASES DUE TO VESSEL DEPRESSURIZATION (psid)

- Primary System Pressure High Intermediate- Low l

l Ice Defeated 37 19 0 Effective IC Fans on 5 5 0 Effective IC

Fans Off 0 0 0 i

l QUESTION 27: What is the mode of reactor vessel' breach?

In the RSS (Ref. 1), it was assumed that the reactor vessel would fail during core meltdown either as a result of a large i steam explosion or by meltthrough of the vessel skirt, causing the hemispherical bottom head to detach. The ZIP study (Ref.

8) pointed out that other failure modes might be at least as-likely as these. The Zion Probabilistic Safety Study (ZPSS, Ref. 14) and IDCOR (Ref. 3) argued that the most probable point 4 l of failure would be a local failure at a core instrumentation  !

j tube, and that the core debris would emerge as a jet in high pressure sequences or would dribble out in low pressure j sequences.

J We postulated five different possibilities for the mode of  ;

} vessel breach: (1) a steam explosion which fails the upper '

i head and generates a missile that fails containment; (2) a

! steam explosion which fails the upper head only (not the I containment); (3) a steam explosion which fails the bottom j head; (4) a pressurized ejection of the debris; and (5) a i

{ gravity-driven pouring of the melt from the pressure vessel.

The first three all result from an in-vessel steam explosion; i

the last two are possibilities if the vessel does not fail from (

a steam explosion.

The assessment of the likelihood, and even'the possibility, i of a steam explosion within the reactor vessel capable of

, generating a missile of sufficient force to. result directly in "

I failure of containment (referred to as the a-mode of containment failure), is an example of a phenomenon about which there is a very wide range of opinions, indicating the large

]i uncertainties that exist. To reflect these uncertainties, we t

] treated this issue as a sensitivity case by first calculating all three walkthroughs of the containment event tree assuming j that there is no potential for vessel failure, then repeating the calculations including the potential for a damaging steam explosion.

i t

B-31 i

l'

The probability of the a-mode was estimated in the RSS to be 0.01. More recent estimates (for example, ZPSS and IDCOR) have disputed this estimate as being unrealistic, or as being physically impossible at high pressures. SAUNA (Ref. 15) reported that triggering an explosion at high pressures might be more difficult, but that the probability of an adequate trigger being available was unknown. They cite experimental evidence that, while it may be more difficult to trigger a high pressure steam explosion, this effect is counteracted by the fact that the conversion ratio tends to be higher. The high-pressure limitation has therefore been neglected for this study.

Berman et al. (IVSE, Ref. 16) conducted a Monte Carlo study -

of the probabilities of steam explosions causing failure of the reactor vessel, and, for those cases in which upper head failure occurred, of the subsequent probability of a large missile with sufficient velocity to fail containment. Several geometric and physical parameters used in the study wore uncertain and were given uniform probability distributions which were sampled by the Monte Carlo procedure. To assess the sensitivity of the results to the assumed distributions, three different distributions were assumed for each parameter. These were labeled " low," " middle," and "high." For example, the three distributions for conversion ratio from thermal to mechanical energy were 0-1.7%, 1.7-3.3% and 3.3-5.0%.

For our central estimate, we used the results of Berman et al., based on " middle" distributions for all the uncertain parameters. Out of 10,000 samples, it was determined that 2126 resulted in a failure of the bottom head, while only one led to a failure of the upper head. That one also created a miss,ile with velocity judged sufficient to threaten containment. The probability of one in 10,000 for a-mode containment failure is consistent with ZPSS and other industry-sponsored risk assessments. The majority of the members of the Steam Explosion Review Group (SERG, Ref. 17) selected "best-estimate" values of 10-4 within a factor of three either way. We, therefore, used a probability of 0.0001 for the central case.

The event tree includes the potential for a steam explosion that could fail the reactor vessel upper head without failing containment. However, since the analyses by Berman, et al.,

showed that the slug energy required to fail the upper head would also be sufficient to produce a missile with high enough velocity to threaten containment, we took the probability of an upper head failure that did not fail containment to be zero.

The probability of a bottom-head failure due to a steam explosion was taken as 0.21. The remainder of the split fraction was a signed to the pressurized ejection mode (for higher or intermediate-pressure core uelts) or to low-pressure meltthrough (for low-pressure core melts).

B-32

We selected a pessimistic value for occurrence of the a failure mode of 0.1. This value is indicated by a number of sources. One member of the SERG selected 0.1 as his upper value, while another selected 0.04. The RSS also cited an upper value of 0.1 in conjunction with its best estimate of 0.01. A value of 0.1 was also used in the Gittus report (Ref.

18). Berman et al. also obtained a value of approximately 0.04 if the conversion ratio is assigned its high distribution while the others are kept at their middle distributions. In addition, a recent experiment by Berman et al. estimated the -

conversion ratio to be much higher than the 3.3-5.0% "high" distribution quoted above. While the results of this experiment have not yet been validated, it is a further indication that this value may not be unreasonably pessimistic.

We estimated the likelihoods for the other failure modes based on the Monte Carlo analyses discussed above. The estimates were made based on an average of the results obtained by Berman et al. When one parameter was assigned its "high" distribution and the others were kept at their " middle" distributions. Since five parameters were varied in this fashion, the result corresponds to the average of five cases, each involving 10,000 samples. For the pessimistic case, a value of 0.59 results if the conversion ratio is assigned its "high" distribution and the others are kept at their " middle" distributions. We selected a slightly lower value, 0.47, which represents the average of the five cases obtained when any one parameter is assigned its "high" distribution and the others are kept at their " middle" distributions. The remainder was split as noted for the central case.

For the optimistic estimate, we used the lowest of the results obtained by Berman et al. when one parameter was assigned its " low" distribution and the others were kept at their " middle" distributions. This resulted in no failures.

The majority of the members of the SERG (Ref. 17) produced "best estimates" of the probability of a-mode containment failure on the order of 0.0001, which is the central estimate for this study. The minimum estimate ("vanishingly small") was consistent with the optimistic estimate. The mean upper limit estimate, 0.02, is low with respect to the pessimistic estimate. The results are summarized as follows:

Optimistic Central Pessimistic Estimate Estimate Estimate Upper head failure (a-mode) 0.0 0.0001 0.1 Upper head failure 0.0 0.0 0.0 (no a-mode)

B-33

Bottom head failure 0.0 0.21 0.47 Pressurized ejection High or intermediate-pressure core melt 1.0 0.79 0.43 Low-pressure core melt 0.0 0.0 0.0 Low-pressure meltthrough High or intermediate-pressure core melt 0.0 0.0 0.0 Low-pressure core melt 1.0 0.79 0.43 QUESTION 28: Is the reactor cavity wet at vessel breach?

Calculations using drawings from the FSAR indicate that the l water in the lower compartment will begin overflowing into the l reactor cavity when 51,000 ft3 has accumulated. The l refueling water storage tank holds between 31,000 and 40,000 ft3 of water, the primary system holds 12,000 ft3, the accumulators hold up to 5400 ft3, the upper head injection system holds between 1350 and 1800 ft3, and the ice amounts to 41,000 to 48,000 ft3 of water. Since some ice melt occurs before vessel breach in all core melt scenarios, any sequence in which the refueling water storage tank is injected before vessel breach will have a wet reactor cavity at the time of vessel breach. The only sequences in which the reactor cavity will be dry at the time of vessel breach are those in which the refueling water storage tank contents are not injected (requiring the failure of both emergency core cooling and sprays in the injection mode), for example, a station blackout sequence. Thus, the reactor cavity was considered to be wet with certainty if emergency core cooling or the sprays operated in the injection mode. Otherwise, the cavity was considered dry at the time of vessel breach.

QUESTION 29: Does direct heating occur? Also, what is the pressure rise in containment due to direct heating and/or a steam spike?

The core debris ejected from the reactor vessel after vessel breach can be quenched either by water or by direct heat transfer to the atmosphere. Both can produce a pressure spike, the former by generating steam (hence, the term " steam spike")

and the latter by adding thermal and/or chemical energy directly to the atmosphere (hence, the term " direct heating').

The two phenomena are linked because the occurrence of one subtracts from the amount of core debris energy available for the other.

Because direct heating is basically an unexplored issue l which can have a large effect on our results, we considered two B-34 1

l

cases in our analyses. In the first case, we excluded direct heating and included only steam spikes. In the second case, we allowed for both to occur concurrently.

Both the CLWG (Ref. 19) and IDCOR (Ref. 3) agree that at least a fraction of the core would be involved in a steam spike if water is present to interact with the core debris after it exits the reactor vessel.

In Sequoyah, the lower compartment and reactor cavity rapidly fill up with water before the time of vessel breach in all sequences with the exception of those noted above. The extent of water in the lower compartment is such that the exit from the reactor cavity is likely to be under approximately 10 ft of water at the time of vessel breach. This fact, along with the geometry of the reactor cavity, was considered as sufficient to preclude direct heating for all scenarios except those with a dry reactor cavity (e.g., station blackout).

Instead, the debris is quenched and the result is a steam spike only.

Direct heating may occur if the reactot vessel fails at high pressure, causing a highly aerosolized jet of debris, and if the geometry of the reactor cavity area is such that there is a pathway for the aerosolized debris to disperse into the containment atmosphere. We postulated that direct heating could also occur as a result of a steam explosion that failed the upper head, providing a direct pathway to the upper containment: however, this possibility did not play a significant role in our final results.

IDCOR discounted the occurrence of direct heating by arguing that the pathways were too tor'tuous for the core debris to be dispersed rapidly into the containment atmosphere. In particular, IDCOR argued that, for Sequoyah, the configuration of the cavity precludes the dispersal of core debris into the lower compartment of containment. CLWG, on the other hand, was sharply divided on the issue. One subgroup (Group A) maintained that less than 2% of the core debris could participate in direct heating, while fully 80% might participate in a steam spike. The other subgroup (Group B) had as much as 50% of the core debris participating in both thermal and chemical heating of the atmosphere, with the remainder of the core debris contributing to a steam spike.

Experiments have been conducted at SNL (HIPS, Ref. 20) using 80 kg of thermitic melt discharged under pressure into 1/10 scale cavities with and without water. Discharge into dry cavities yielded a highly aerosolized plume of ejecta, a large fraction of which was thoroughly dispersed into the atmosphere. We estimated that the high driving forces and B-35

l velocities involved in high and intermediate pressure meltdowns would be adequate to disperse the ejecta not only throughout the steam generator enclosures, but thoroughout the reactor containment as well.

When melt was ejected into a water-filled cavity, the water was immediately blown out of the cavity as a slug, which then appeared to become finely divided. Hot debris followed, and, t

at least prior to cavity failure, appeared to be intimately mixed with the cloud of water droplets. White clouds (apparently water fog) could be observed throughout the ejected material. We estimated for this study that an intimate mixture of hot debris and water droplets or steam would be likely to be ejected from water-filled cavities, and considered a steam spike to be assured. Direct heating was considered f " impossible" for low-pressure sequences, for steam explosions that failed the upper or bottom head, and for all other 1

sequences in which the reactor cavity is flooded, since at Sequoyah the exits from the cavity would be deeply covered with

, water. We reasoned that the core debris would be completely quenched or, if not, that the water would cool the heated j atmosphere. For high pressure sequences with a dry cavity, i

direct heating was considered "certain" for sensitivity case 2 (with magnitudes adjusted for the ejection pressure), and

" impossible" for sensitivity case 1.

We adjusted the pressures calculated for direct heating and steam spike for Surry based on the ratios of rated thermal power and containment volume for the two plants. For those sequences in which the ice condenser had been defeated, the values were then used for resulting pressure increases.

IDCOR calculated a steam spike of 15 psi (1.0 bar) for the S2 D sequence at Zion, assuming that 50% of the core debris

, was quenched. CLWG (Ref. 21) reported that the spike from core debris quenching in water could be nil or as high as 26 psi j (1.9 bar), the latter corresponding to quenching 100% of core

debris. The CLWG figures were obtained without consideration I of containment cooling; however, BMI-2104 calculations indicated that the effect of sprays on the steam spike would be small if one assumed, for the pessimistic case, that the debris was highly fragmented. Thus, we took the size of the steam spike to be 10 psi (optimistic), 15 psi (central) and 26 psi (pessimistic) for Surry, and used the same values for i Sequoyah.

l The high pressure, minimal water sequences were used as bases for estimatec for sequences of intermediate pressure.

CLWG did not give guidance for sequences in which the ejected i melt would not become highly aerosolied. We would expect a lower direct heating pulse in such sequences. However, we did 4

B-36 i

.m . ,,, - - , . . _ _ , , . , _ . . . - _ . - _ - - . . . . ~ . .- , _ . - - - - _ - - - . _ . . . __

not assume a corresponding increase in the size of the steam spike, because of the coarser fragmentation expected. The direct heating contribution was assumed to be one-half the difference between the direct heating plus steam spike pulse and the pulse due to steam spike alone. For low pressure sequences, no direct heating contribution was expected. For sequences that might have limited water in the cavity, some guidance was available from a calculation performed for CLWG with limited cavity water; the predicted steam spike for that calculation was used for the pessimistic estimate, one-half the calculated value for the central estimate, and zero for the optimistic estimate. Reasoning that the steam spike pulses expected in lower pressure minimal water sequences would be less likely to threaten containment, it did not seem fruitful l to perform extensive recalculations.

l The pressures we derived for Sequoyah from the Surry values are tabulated in Table B.3.

Table B.3 PRESSURE INCREASES DUE TO DIRECT HEATING AND STEAM SPIKE (FOR SCENARIOS INVOLVING LOSS OF ICE CONDENSER FUNCTION)

Primary System Pressure High Intermediate Low Direct Heating Optimistic 23 0 0 Central 90 30 0 Pessimistic 213 80 0 Steam Spike Optimistic 23 23 23 Central 31 31 31 Pessimistic 54 54 54 QUESTION 30: Does a detonation occur in the ice condenser at vessel breach?

Just as in a similar question earlier, there are three possibilities for this question. A detonation can occur in the ice condenser which either fails the containment or defeats the function of the ice condenser, or no detonation damage occurs to the ice condenser. The first two possibilities are only considerations if there is still ice remaining in the ice condenser and if the air return fans are not operating. These requirements must be met in order to achieve detonable j concentrations in the ice condenser. The probability of a 1 detonation in the ice condenser failing containment was taken I as " impossible," "unlikely," and " indeterminate" in the optimistic, central, and pessimistic walkthroughs. In the B-37

absence of any prior damage to the ice condenser, a failure of the ice condenser function due to a detonation within the ice condenser was considered " impossible" in the optimistic case, "unlikely" in the central case, and " indeterminate" in the pessimistic case.

QUESTION 31: Has the ice completely melted or have bypass paths developed after vessel breach?

The answer to this question is dependent upon answers to prior questions related to the ice condenser. Specifically, if an earliar answer to a similar question involved degraded function of the ice condenser, that condition can only be maintained or worsened. If a detonation occurred in the ice condenser at the time of vessel breach which defeated its function, we assumed a bypass of the ice condenser.

Examination of past analyses (IDCOR, Ref. 3: SASA, Ref. 6) l showed that typically more than half of the ice had melted by the time of vessel breach for sequences which invulved loss of emergency core cooling injection. These same analyses indicated as much as 60-90% had melted by the time of vessel breach for sequences which involved initial break sizes of either Si or S2 and initial success of emergency core cooling in the injection mode followed by failure in the recirculation mode. Based on this, the two types of scenarios were treated seperately. For those sequences involving success of emergency core cooling injection and a break size in the primary system equivalent to an Si LOCA, the probability of complete ice melt was considered " remotely possible,"

"unlikely," and " indeterminate" and that of developing bypass paths in the ice condenser considered "unlikely,"

" indeterminate," and " indeterminate" for tiie optimistic, central, and pessimistic walkthroughs, respectively.

For those sequences involving success of emergency coolant injection and a break size equivalent to an S2 LOCA, the probability of complete ice melt was considered " impossible,"

" remotely possible," and "unlikely" and that of developing bypass paths in the ice condenser was considered " remotely possible," "unlikely," and " indeterminate" for the optimistic, central, and pessimistic walkthroughs, respectively.

In those scenarios which involved failure of core cooling in the injection mode we considered complete ice melt to be

" impossible" for all three walkthroughs. We considered the probability of developing bypass paths through the ice condenser " impossible", "unlikely" and " indeterminate" for the optimistic, central, and pessimistic cases respectively.

B-38

Rapid melting of the ice is expected for sequences in which the cavity is dry up to the time of vessel breach, with discharge of the cold leg accumulators and upper head injection system at the time of vessel breach. Ice melting is especially rapid if the core debris is coolable, as we would expect under these conditions (refer to Question 43). In this case, the decay heat is entirely transferred to the evaporation of water, which places a heavy load upon the ice condenser (although less ice would have been melted before vessel breach than if l emergency coolant injection had succeeded). We considered bypass of the ice to be " impossible" (optimistic),

" indeterminate" (central), or "likely" (pessimistic). We considered complete melt to be " impossible" (in the optimistic and central walkthroughs) or "unlikely" (pessimistic.

QUESTION 32: Does a hydrogen deflagration occur at the time of vessel breach? Also, what is the pressure rise in containment if a deflagration does occur?

This question is very similar to two earlier questions dealing with burns in the upper and lower compartments. For burns at vessel breach, the distinction between lower and upper compartment burns need not be made insofar as loading on containment is concerned. Either the primary system is at low pressure long before vessel breach occurs, or it is at some elevated pressure. If it is at low pressure, much of the hydrogen generated in-vessel has been released prior to vessel breach, leaving little to be released at the time of vessel breach. However, if the primary system is at high pressure, large amounts of hydrogen are released at the time of vessel breach, leading to global burns.

Examination of prior analyses (IDCOR, Ref. 3: SASA, Ref. 6) revealed that global burns at vessel breach only occurred for S2 break sizes or smaller. The magnitude of the pressure spike estimated to result from. combustion was affected by several factors, including whether or not any prior burning had taken place, the amount of zirconium which had reacted, whether the cavity was wet or dry, and the type of combustion models assumed. In the optimistic case, we used the IDCOR assessment of combustion and zirconium oxidation, which tended to minimize any pressure loads. For the pessimistic walkthrough, we estimated 100% zirconium oxidation by the time of vessel breach, and used global combustion models. For the central case, we used a zirconium oxidation level of approximately 50%

along with global combustion models. Whether the cavity is wet or dry plays an important role. If the cavity is wet (and it can only be dry or flooded at the time of vessel breach) much if not all of the steam released from the primary system at vessel breach is condensed and the driving force to transport any hydrogen released to the upper compartment is absent.

B-39

However, if the cavity is dry, the steam and hydrogen released from the vessel are pushed up through the ica condenser and much more hydrogen is deposited in the upper compartment. This happens very rapidly and leads to large pressure loadings.

The actual values used for pressure rises due to burns are tabulated below.

Table B.4 PRESSURE RISES ASSOCIATED WITH HYDROGEN COMBUSTION AT VESSEL BREACH Pressure Rise (psid)

Optimistic Central Pessimistic Prior Burning S2 and S3 LOCAs 1 20 45 Other, with fans available 5 12 30 No Prior Burning With Fans 5 30 100 Without Fans 5 60 100 QUESTION 33: Does containment fail at vessel breach?

This question is treated in a manner similar to the previous question for early containment failure. The pressure rises due to vessel depressurization, direct heating / steam spike, and hydrogen combustion are added to the early baseline pressure and compared to a normally distributed containment failure pressure. The normal distributions are those described for question 21.

QUESTION 34: What is the mode of intermediate containment failure?

This question serves to summarize the mode of intermediate containment failure to facilitate later questions in the tree.

There are four possible modes of intermediate containment failure, denoted in the event tree as branches IA, ID, IHS, and IS. IA stands for an alpha mode failure. If an in-vessel steam explosion results in direct failure of containment this branch is taken. All containment failures which are not alpha modes, but involve direct heating, are classified as ID and this branch is taken. IHS refers to containment failures resulting from a coincident steam spike and hydrogen burn at the time of vessel breach (excluding any IA or ID failures).

IS refers to containment failure due to a steam spike alone.

As mentioned previously, this node in the tree summarizes B-40

l information previously deduced in the tree; hence, the probabilities of the different branches are always either zero  ;

or one and do not vary with the three different walkthroughs.

QUESTION 35: To what degree is there intermediate bypass of the auxiliary building?

This node is nearly identical to question 22. The only difference arises due to the possibility of an alpha mode failure of containment at the time of vessel breach. The postulated location for alpha mode failures is always in the i dome of containment. Since no portion of the auxiliary l building surrounds the containment dome and since the shield l

building is relatively weak, we assumed complete bypass to be certain when alpha mode failures were involved. The probability of bypass was treated the same as in question 22 for a failure due to a local detonation in the ice condenser.

All other failures (ID, IHS, IS) were treated the same as failures due to global hydrogen burns in question 22.

QUESTION 36: What is the status of the air return fans after vessel breach?

This is similar to the previous question 24. Burns at the time of vessel breach are either global burns or upper compartment burns only, and upper compartment burns were given the most credit in the previous question for failing the fans.

Thus the same probabilities are used here for fan failure in the case of a deflagration during the period of vessel breach as were used previously for fan failure due to an upper compartment deflagration.

QUESTION 37: What is the status of the sprays after vessel breach?

We considered two mechanisms by which the containment sprays might fail. Either containment failure could rip and tear the piping for the spray headers, or debris (e.g.,

preexisting, pipe insulation, corium debris) in the recirculation sump could cause cavitation failure of the spray system pumpc. At Sequoyah the spray piping runs outside containment from the auxiliary building up to the dome where it then penetrates the containment and connects to the spray headers in the dome. In RSSMAP (Ref. 2) containment failure assured spray failure, an assumption we evaluated as overly conservative.

For the case of containment failure, the work done for Task Action Plan A-43 (A-43, Ref. 22) showed that pumps of the type used for the spray system could operate for long periods of time at a net positive suction head less than that for which B-41

i they were designed. We therefore estimated spray failure following containment failure to be "unlikely" in the optimistic and central walkthroughs. We estimated spray failure following containment failure to be "likely", following the RSS and RSSMAP studies, for the pessimistic walkthrough.

If no containment failure occurred, we noted that although the sump water at Three-Mile Island was laden with debris, the pumps operated normally. We also noted that some pumps in

, industrial service operate for years with debris laden fluid.

Furthermore, the intake screens at nuclear power plants are

! designed to make blockage by debris difficult. We estimated that spray failure without containment failure would be

" remotely possible" (optimistic), "unlikely" (central), or

" indeterminate" (pessimistic).

, QUESTION 38: Is ac power restored after vessel breach?

i RSSMAP (Ref. 2) did not consider the effect of late power j restoration for core melt scenarios. The largest impact for an i ice-condenser containment is apt to be the prevention of long term overpressure for those scenarios in which coolable debris beds are formed. This is because ac power is needed to power i the containment sprays which are the only means of long term heat removal. In light of this and the small volume of an ice condenser containment, we used the probability of power restoration at four hours (0.75) from the Electric Power Research Institute report NP-2301 (EPRI, Ref. 23). Power restoration at times much later than this would be incon-i sequential since the containment would have already failed due e to steam overpressure.

QUESTION 39: Are the igniters on late?

This question is very similar to the earlier question 15.

l If the igniters were on early, they remained on late. If ac l power were unavailable early, but restored late, the same

! probabilities were used as in question 15 based on operator i failure to follow a procedure (1278, Ref. 9). If ac power were 2

available early, but the igniters had not been turned on, we

! considered it " indeterminate," "unlikely," and " impossible"

! that the operators would realize their oversight and turn on the igniters late for the optimistic, central, and pessimistic walkthroughs, respectively.

QUESTION 40: Do the air return fans operate late?

This question addresses the initiation of air return fan operation following the restoration of ac power. The probabilities are either zero or one and do not change for the different walkthroughs. Previous questions regarding fan I

i B-42 i

l.,____---__-_-..-_ , - - _ - . _ _ _ - . - - . - - . _ . _ _ . - . _ _ . . _ - . . _ - . - - - _ , _ , _ - _ . - _ _ . _ . _ _

k operation have specified the fans to be operating, avhilable, or failed. The.available category refers to the fans being operable but not running due to the absence.of ac power. Thus, if the prior status of the fans were "available" and ac power were restored late, we took the probability of fans operating late to be unity. However, if the fan's were previously f ailed or ac power were not available late, the tans were prevented from operating. h; QUESTION 41: Do containment sprays ope' rate late?

The logic for this question is identica1'to the previous question but applies to the containment sprays instead of the air return fans.

3 T

! -QUESTION 42: What is the late baseline pressure in co.od/IAment?

l -

, This is a one-branch node and is in the tree to supply the 1 baseline pressure to the event tree p:ogram in order that the \

I probability of containment failure dye to late hydrogen burns can be calculated. Thus no probabilities need to be discussed. '

only the values of pressure. '

The value of baseline pressure is dependent on the i availability of containment heat removal.' This means,it should 'q be dependent upon the condition of the ice condenser and the '

i status of the air return fans and containment sprays. If the \1 i

ice condenser or the sprays are operational and the air return '

fans are operating, the pressure in containment will be held to <

{ a minimum. If tha sprays are operating but the fans are not, i the pressure will be slightly higher since all of the

noncondensible gases will have been purged into the' upper i compartment by the steam being generated in the lower 4

compartment. If no form of containment hea't removal is available the steam partial pressure will continue to, rise j ,

until containment fails in some manner. 'Since the baseline  ;

i pressure was only used in the late hydrogen burn question, .

i pressures greater than that corresponding to' steam inerting o I

containmentwereactuallymoreoptimisticandhenceanupperf. ,

bound pressure corresponding to steam inerting;of the l

, containment was imposed. One other case arose; that is an operational ice condenser with failed fans and sprays. The ice 3

condenser could still function to condense some steam, but

' without the air return fans its performance would be impaired; thus, the pressures used were those for complete loss of 4

containment heat removal but reduced slightly to account for r  !

some reduction due to the ice condenser. The act.ual values we ,

J used were-derived from previous analyses of accidents in i

! ice-condenser plants (IDCOR, Ref. 3; SASA, Ref. 6) and are

{ tabulated below. <

1 i

)

i i

tl i

! s $

t \

l \ B-43 ' s

/ i l . . , 1 i , (

j ~

f t

. -.- . - -. - - =

J 3

s.

."' Il j N Table B.5 I p, I, ,

VALUES USED FOR LATE BASELINE PRESSURE l

y. Optimistic Central Pessimistic CHR & Fans, 18 19 20 Sprays /no F ns 19 21. 23 no CHR ( 20 30 40 i

Ice /no Sprays 20 25 35

/no Fans 4

( . QUESTION 43: Is a coolable debris bed formed and maintained after vessel breach?

a .

) Models for debris bed formation and quench from both the

, industry and regulatory sponsored studies were considered in j/ addressing this question. We considered the chance of i attaining a coolable debris bed to be dependent upon two things

(1) whether or not the cavity was dry at the time of vessel breach, and (2) the mode of debris ejection from the j vessel. If the cavity were dry at the time of vessel breach,

! we considered the mode of debris ejection to be irrelevant 4

'since the' debris would collect in the cavity and only water i intrusion from above could account for coolable debris beds

being formed. In order for the debris to be' cooled, we imposed t the additional restriction that the RCS not be at low pressure, so that water from the accumulators would be available to cool i ,

the core following vessel breach. The IDCOR model is based on '

j .the premise that the water intrucion from above will slowly

. quench the debris, while NRC-sponsored research indicates that

! such' intrusion is not likely to occur. For this scenario, we

considered the probability of attaining a coolable debris configuration to be "likely " "unlikely," and'" remotely Possible" for the optimidpic, central, and pessimistic' cases.

-l In the case of pressurized ejection into a wet cavity, the debris is likely to be very well dispersed and consequently quenched by the water. Thus formation of a coolable

, configuration was considered "almost certain," "likely," and j ',

" indeterminate" for the optimistic, central, and pessimistic walkthroughs, respectively. If the debris ejection came in the

! form of a gravity-driven pour, we considered the probability of i forming a coolable debris bed less likely than for the l ,

pressurized ejection case due to the absence of reliable i,' dispersal mechanisms. We considered the formation of a coolable debris bed to be "likely," " indeterminate," and j " remotely possible" for the optimistic, central, and I, pessimistic walkthroughs respectively.

b \ ,

i i

I B-44 s I

l

/-

l , .i

! t ,

.__ _-.--__~.,__L.__-_-.-_ -

if c' ,* , ,

QUESTION 44: Is there a late deflagration? Also, what is the

/ associated pressure rise in containment?

l This question addresses not only the probability of

. occurrence of a<1 ate hydrogen burn, but also the extent of such a burn. That is, as well as probabilities of a burn occurring, the magnitude of the resultir.g pressure rise is also supplied for a later calculation of probability of containment failure.

We assumed different models for the late combustion as well.

In the pessimistic and central cases the pressure rises we used are based on past analyses (SASA, Refs. 6 & 10) which used a global combustion criterion. The pressure rises for the optimistic case are based upon the assumption of continuous l localized burning of the combustible gases in the vicinity of the debris in the reactor cavity (IDCOR, Ref. 3).

/

The answers to this question depend upon the status of the igniters and whether there was any prior burning or containment failure. In the case of no late igniters and no prior failures

, ignition was considered "likely." In the cases with igniters and no prior failures, ignition was considered "certain." The only case left then is that of some prior failure. Ignition was' precluded in this case based upon nearly all the oxygen having been purged from containment, rendering the containment inert.

The pressure rises we used are tabulated below for the different cases of interest.

Table B.6

. PRESSURE RISES FOR A LATE HYDROGEN BURN (WITH NO PRIOR CONTAINMENT FAILURES)

Optimistic Central Pessimistic No igniters 5 35 185 Igniters /

No prior burns 5 25 55 Igniters /

Prior burns 5 18 25 QUESTION 45: Does containment fail due to a late deflagration?

This question is very similar to the two earlier questions of containment failure. The late baseline pressure and the pressure rise due to a hydrogen burn are summed and compared to the normally distributed containment failure pressure. The specification of the failure pressure is exactly the same as was described in the prior questions for the cases in which B-45

- - - - . - - . . - .. ~ - - _ _ . - . -.

there were no prior failures. However, if there were a prior failure, the failure pressure was raised artificially high to i prevent the program from calculating some probability of failure in the pessimistic case due to the high baseline pressure alone. This type of failure (late overpressure) is considered later in the tree.

, QUESTION 46: Do containment sprays operate very late?

This question is very similar to earlier questions

regarding containment-failure-induced spray failure. If i

containment fails due to a hydrogen burn, containment spray failure is considered to be "unlikely" in the optimistic and

central cases and "likely" in the pessimistic case.

l QUESTION 47: Does containment failure occur late due to j noncondensible gas and/or steam accumulation?

i

.The answer to this question depends upon the status of the debris (coolable or non-coolable) and the availability of

, containment heat removal. (It is also contingent upon no i

earlier failures of containment.) If containment heat removal is available but the debris is initially not coolable, it has i

been suggested that the debris might stop penetrating the concrete basemat after decay heat levels had decreased sufficiently, thus terminating the associated non-condensible gas production. This would result in a steady state configuration without failing containment. We considered the chance of this occurring to be "certain" in the optimistic case, "unlikely" for the central case, and " remotely possible" for the pessimistic case. Of course if the debris is coolable  !

and containment heat removal is available, the containment will not fail since the steam partial pressure is kept low and no concrete attack takes place to produce noncondensible gases.

On the other hand, if the debris is coolable and if no 4

' containment heat removal capability exists the containment is assured to fail due to overpressurization by steam.

QUESTION 48: To what degree is there late auxiliary building

, breakthrough or bypass?

s I

The logic for this question is the same as the previous questions dealing with auxiliary building bypass failures associated with containment failures due to overpressurization L and/or combustion. '

QUESTION 49: Does basemat meltthrough occur?

[

The only possibility for preventing basemat meltthrough is -

if coolable debris beds are formed and containment heat removal is available. In this case basemat meltthrough was considered impossible. Otherwise, basemat meltthrough was assured.

, B-46 f

REFERENCES FOR APPENDIX B .

1. (RSS). " Reactor Safety Study--An Assessment of Accident Risks in U.S. Commercial Nuclear Power Plants," U.S.

Nuclear Regulatory Commission Report WASH-1400 (NUREG75/U14), 1975.

2. (RSSMAP). " Reactor Safety Study Methodology Applications-Program," U.S. Nuclear Regulatory Commission Report t NUREG/CR-l' ,9, Volumes 1-4, Sandia National Laboratories. .

1981-1982.

3.- (IDCOR). "IDCOR Task 23.1: Integrated Containment Analysis," Technology for Energy Corporation, 1984.

l 4. (NRR). Weinstein, M. B. " Primary Containment Leakage Integrity: Availability and Review of Failure Experience,"

Nuclear Safety 21: 1980.

5. (PSAR). Tennessee Valley Authority, " Final Safety Analysis Report for the Sequoyah Nuclear Power Plant," 1974. '

! 6. (SASA). Camp, A. L., et al., " MARCH-HECTR Analysis of I

Selected Accidents in an Ice-Condenser Containment," U.S.  !

Nuclear Regulatory Commission Report NUREG/CR-3912 Sandia l National Laboratories, 1984. 1 L

i 7. (NRR). Lyons W. G., Presentation at NRC/IDCOR Meeting on Containment Loads and Fission Product Behavior, U.S. i i

Nuclear Regulatory Commission, May 15-17, 1984.  !

l i

8. (ZIP). Murfin, W. B., et al., " Report of the Zion / Indian

[

Point Study," U.S. Nuclear Regulatory Commission Report i NUREG/s' 1-1410, Sandia National Laboratories, 1980.

f

9. (1278). Swain A. D. and H. E. Guttmann, " Handbook of Human Reliability Analysis with Emphasis on Nuclear Power Plant Applications " U.S. Nuclear Regulatory Commission Report NUREG/CR-1278, Sandia National Laboratories, August '

1983.

10. (SASA). Haskin, F. E. and V. L. Behr, "HECTR Results for Ice-Condenser Containment Standard Problem," Proceedings of

~ the Second Containment Integrity Workshop, Crystal City, VA, June 1984.

4

' 11. (SASA). Haskin, F. E., et al., " Analysis of Hypothetical.

Severe Core Damage Accidents for the Zion Pressurized Water Reactor," U.S. Nuclear Regulatory Commission Report NUREG/CR-1989 Sandia National Laboratories, 1982.

B-47

12. (AMES). Greinann, et al., " Final Report, Containment Analysis Techniques, a State-of-the-Art Summary," U.S.

Nuclear Regulatory Commission Report NUREG/CR-3653, Ames Laboratory and Sandia National Laboratories, March 1984.

13. (SER). " Safety Evaluation Report Related to the Operation of Sequoyah Nuclear Plant, Units 1 and 2." U.S. Nuclear Regulatory Commission Report NUREG-0011, Supplement No. 6 November 1982.

14. (ZPPS). " Zion Probabilistic Safety Study," Commonwealth Edison Company, 1981.

15. (SAUNA). Rivard, J. B., et al., " Identification of Severe Accident Uncertainties," U.S. Nuclear Regulatory Commission Report NUREG/CR-3440, Sandia National Laboratories, 1984.

16. (IVSE). Berman, M., et al., "An Uncertainty Study of PWR Steam Explosions," U.S. Nuclear Regulatory Commission Report NUREG/CR-3369, Sandia National Laboratories, 1984.

17. (SERG). Steam Explosion Review Group, "A Review of the ,

Current Understanding of the Potential for Containment Failure Arising from In-Vessel Steam Explosions," U.S.

Nuclear Regulatory Commission Report NUREG-1116, February 1985.

18. (Gittus). (To be proveded).

19. (CLWG). " Consensus Summaries for Standard Problems 1 through 6," U.S. Nuclear Regulatory Commission, May-June 1984.

20. (HIPS). High Pressure Ejection Test Series, Sandia National Laboratories, 1985.

21. (CLWG). Bergeron, K. D., and Williams, D. C.,

" Calculations for PNR Standard Problems 1 and 2 with Direct Heating," Submitted to Containment Loads Working Group.

22. (A-43). Serkiz, A. W, " Containment Emergency Sump Performance, Technical Findings Related to Unresolved

Safety Issue A-43 " U.S. Nuclear Regulatory Commission Report NUREG-0897, 1983.

23. (EPRI). McClymont, A. S. and B. W. Poehlman, " Loss of OffSite Power at Nuclear Power Plants: Data and Analysis,"

Eler *.ric Power Research Institute Report NP-2301 (Interim Report), March, 1982.

B-48 l

\

24. Gieseke , J. A., et al., "Radionuclide Release Under Specific Accident Conditions," BMI-2104, Battelle Columbus Laboratories, July 1984.

25. (SASA). Dingman, S. E. and A. L. Camp, " Pressure-Tem-perature Response in an Ice-Condenser Containment for Selected Accidents," SAND 85-1824C, Sandia National Laboratories, in press.

B-49

Appendix C COMPUTER INPUT FOR SEQUOYAH CONTAINMENT EVENT TREE The computer input used in the analysis of the Sequoyah containment event tree is listed in this appendix. The format of the input is identical to that described in Appendix A of this report. The input in this appendix is specifically for plant damage-state S2IYYB, which is a small LOCA with successful emergency core

, cooling in the injection phase but failure in recirculation, and with the containment spray system operating to provide containment heat removal. For this set of walkthroughs, direct heating and in-vessel steam explosion are considered possible. The questions for which the input must be changed as a function of plant damage state are annotated in the input listing. The branching options are designated in abbreviated form; the detailed definitions are provided in Appendix B.

i I

2 C-1 9

SEQUOYAH - S2IYYB - DH and Alpha Included 3 1.0 1.0 1.0 49 1.0E-04 1.0E-07 Nwal Pinit Pinit Pinit Nguest Prntol Toler 1 Is ac power available early? This is specified directly by the plant damage state.

2 E-AC EfAC 1 1 2 1.000 0.000 1.000 0.000 1.000 0.000 2 Is there pre-existing leakage or isolation failure?

4 EE3 EE2 EE1 EE0 1 1 2 3 4 0.0000 0.0000 0.0170 0.9830 0.0000 0.0025 0.0330 0.9645 0.0000 0.0040 0.0580 0.9380 3 What is the initial break location? This is a direct function of the accident type, as denoted by the first designation in the damage state.

6 EERPV EEHL EECL EESGTR EEV EEPORV 1 1 2 3 4 5 6 0.000 0.500 0.500 0.000 0.000 0.000 0.000 0.500 0.500 0.000 0.000 0.000 0.000 0.500 0.500 0.000 0.000 0.000 4 What is the initial break size? This is a direct function of the accident type, as denoted by the first designation in the damage state.

4 EELg EES1 EES2 EES3 1 1 2 3 4 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 5 Is the containment initially bypassed? The outcome reflects bypass only for plant damage state V.

2 EEB noEEB 2 1 2 2

4 3 3 3 3

-1 -2 -3 -6 noEERPV noEEHL noEECL noEEPORV 1.000 0.000 1.000 0.000 1.000 0.000 Otherwise 0.000 1.000 0.000 1.000 0.000 1.000 6 Are the steam generators wet or dry? The outcome for this question is determined directly by the plant damage state.

2 SGWet SGDry 1 1 2 1.000 0.000 1.000 0.000 1.000 0.000 7 Is there initial operation of the air return fans?

3 EEFan EEaFan EEfFan 2 1 2 3 2

1 1 C-2 i

2 EfAC 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 8 Does ECC operate in the injection mode? This is indicated by the second category in the plant damage state.

2 ECCInj noECCInj 2 1 2 2

1 1 1

E-AC 1.000 0.000 1.000 0.000 1.000 0.000 Otherwise 0.000 1.000 0.000 1.000 0.000 1.000 9 Do the sprays operate in the injection mode? This is determined by the plant damage state.

3 EESpInj EEaSpInj EEfSpInj 2 1 2 3 2

1 1 2

EfAC 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 10 Do the sprays operate in the recirculation mode? This is determined by the plant damage state.

3 EESp EEaSp EEfSp 2 1 2 3 3

1 9 1

EESpInj 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1 9 2

EEaSpInj 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise C-3

0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 11 To what degree is the auxiliary building bypassed?

3 EESB3 EESB2 EESBl 2 1 2 3 5

3 2 3 3

-4 -4 -5 noEE0 noEESGTR noEEV 0.000 0.100 0.900 0.000 0.100 0.900 0.000 0.100 0.900 2 3 6 4 1 EESGTR SGWet 0.000 0.100 0.900 0.000 0.100 0.900 0.000 0.100 0.900 2 3 6 4 2 EESGTR SGDry 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1 3 5

EEV 0.000 0.100 0.900 0.000 0.100 0.900 0.000 0.100 0.900 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 12 What is the location, if any, of induced failure in the primary system?

This is determined by the plant damage state.

5 EfRPV EfHL EfCL ESGTR noEfRCPB 2 1 2 3 4 5 5

3 1 4 6 1 4 2 E-AC EES3 SGDry 0.000 0.200 0.000 0.240 0.560 0.000 0.200 0.000 0.240 0.560 0.000 0.200 0.000 0.240 0.560 3 1 4 6 2 4 2 EfAC EES3 SGDry 0.000 0.120 0.400 0.140 0.340 0.000 0.120 0.400 0.140 0.340 0.000 0.120 0.400 0.140 0.340 3 1 4 6 1 4 1 E-AC EES3 SGWet 0.000 0.260 0.000 0.000 0.740 0.000 0.260 0.000 0.000 0.740 C-4

0.000 0.260 0.000 0.000 0.740 3 1 4 6 2 4 1 EfAC EES3 SGWet 0.000 0.140 0.460 0.000 0.400 0.000 0.140 0.460 0.000 0.400 0.000 0.140 0.460 0.000 0.400 Otherwise 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 13 What is the area of temperature-induced primary system failure?

4 ELg ES1 ES2 noEfRCPB 2 1 2 3 4 5

1 12 1

EfRPV 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1 12 2

EfHL 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1 12 3

EfCL 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000

. 0.000 0.000 1.000 0.000 1 12 4

ESGTR 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 Otherwise 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 1.000 14 Does early containment bypass occur during meltdown?

3 EB2 EB1 noEB 2 1 2 3 5

7 3 3 3 3 4 4 12

-1 -2 -3 -6 -1 -2 5 noEERPV noEEHL noEECL noEEPORV noEELg noEES1 noEfRCPB 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 7 3 3 3 3 4 4 12

-1 -2 -3 -6 -3 -4 5 noEERPV noEEHL noEECL noEEPORV noEES2 noEES3 noEfRCPB 1.000 0.000 0.000 C-5 l

t f

1.000 0.000 0.000 1.000 0.000 0.000 2 12 13 4 3 ESGTR ES2 0.000 1.000 0.000

' 0.000 O 000 1.000 0.000 1.000 0.000 3 12 13 13 4 -3 -4 ESGTR noes 2 EfRCPB 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 .l 15 Do igniters operate early?

2 EIg noEIg

2 1 2 2

1 1 1 l E-AC l 1.000 0.000 0.997 0.003 0.970 0.030 Otherwise 0.000 1.000 0.000 1.000 0.000 1.000 16 Is the lower compartment inert during core degradation?

2 ELCIn noELCIn 2 1 2 2

j 1 7 4

1 EEFan O.000 1.000 0.000 1.000

, 0.000 1.000 i Otherwise i 1.000 0.000 1.000 0.000 i; 1.000 0.000

! 17 What is the early baseline pressure in containment?

! 1 EPBase 4

1

2 1 7 1

EEFan

< 1.000 1

1.000 1.000 1

C-6 4

d

• .-w, , -.e ,,r ..-e ~ ,- -

c,w . --- - - ,--s- , --, -.---- -- ,- ..= - r-~i-r,2i---e, .- - - -.

• 1 18.0 19.0 20.0 Otherwise 1.000 1.000 1.100 1

l 1 19.0 l 21.0 23.0 18 Does a local detonation occur in the ice condenser?

3 EDt-EH EDt-fIC noEDt 2 1 2 3 2

1 7

-1 noEEFan 0.000 0.000 1.000 0.100 0.100 0.800 0.500 0.500 0.000 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 19 Is there a deflagration in the lower compartment before vessel breach?

2 ELCDef noELCDef 4 1 2 9

5 4 4 5 15 16

-3 -4 2 1 2 noEES2 noEES3 noEEB EIg noELCIn 1.000 0.000 1.000 0.000 1.000 0.000 1

2 1.0 0.0 5.0 0.0 25.0 0.0 5 12 13 14 15 16

-5 -3 3 1 2 EfRCPB noes 2 noEB EIg noELCIn 1.000 0.000 1.000 0.000 1.000 0.000 1

2 1.0 0.0 5.0 0.0 25.0 0.0 5 4 4 5 15 16

-3 -4 2 2 2 noEES2 noEES3 noEEB noEIg noELCIn 0.000 1.000 0.900 0.100 0.900 0.100 1

2 0.0 0.0 C-7

25.0 0.0 37.0 0.0 5 12 13 14 15 16

-5 -3 3 2 2 EfRCPB. noes 2 noEB noEIg noELCIn 0.000 1.000 0.900 0.100 0.900 0.100 1

• 2 0.0 0.0 25.0 0.0 37.0 0.0 ,

5 4 4 5 15 16  !

i

-1 -2 2 1 2 noEELg noEES1 noEEB EIg noELCIn 1.000 0.000 0.500 0.500 1.000 0.000 1 1

2 1.0 0.0 4.0 0.0 6.0 0.0 6 12 13 13 14 15 16

-5 -1 -2 3 1 2 EfRCPB noELg noESl noEB EIg noELCIn 1.000 0.000 0.500 0.500 1.000 0.000 1

2 1.0 0.0 4.0 0.0 6.0 0.0 5 4 4 5 15 16

-1 -2 2 2 2 noEELg noEES1 noEEB noEIg noELCIn 0.000 1.000 0.100 0.900 0.001 0.999 1

2 1.0 0.0 6.0 0.0  ;

8.0 0.0 6 12 13 13 14 15 16

-5 -1 -2 3 2 2-EfRCPB noELg noes 1 noEB noEIg noELCIn 0.000 1.000 0.100 0.900 0.001 .0.999 1

2 1.0 0.0 6.0 0.0 8.0 0.0 Otherwise 0.000 1.000 )

0.000 1.000 0.000 1.000 1 l C-8

2 0.0 0.0 0.0 0.0 0.0 0.0 20 Is there a deflagration in the upper compartment before vessel breach?

2 EUCDef noEUCDef 4 1 2 8

2 7 19

-1 2 noEFan noELCDef 0.000 1.000 0.000 1.000 0.000 1.000 1

3 0.0 0.0 0.0 0.0 0.0 0.0 4 4 4 5 15

-3 -4 2 1 noEES2 noEES3 noEEB EIg 0.000 1.000 0.000 1.000 1.000 0.000 1

3 0.0 0.0 0.0 0.0 25.0 0.0 4 13 13 14 15

-3 -4 3 1 noes 2 EfRCPB noEB EIg 0.000 1.000 0.000 1.000 1.000 0.000 1

3 0.0 0.0 0.0 0.0 25.0 0.0 6 4 4 13 13 14 15

-1 -2 -1 -2 3 1 noEELg noEESl noELg noes 1 noEB Eig 0.000 1.000 0.000 1.000 0.000 1.000 1

3 0.0 0.0 0.0 0.0 0.0 0.0 4 4 4 5 15

-3 -4 2 2 noEES2 noEES3 noEEB noEIg 0.000 1.000 0.000 1.000 0.900 0.100 1

3 0.0 0.0 0.0 0.0 32.0 0.0 C-9

4 13 13 14 15

-3 -4 3 2 noes 2 EfRCPB noEB noEIg

, 0.000 1.000 0.000 1.000 0.900 0.100 1

3 0.0 0.0 0.0 0.0 32.0 0.0 6 4 4 13 13 14 15 4 -1 -2 -1 -2 3 2 noEELg noEES1 noELg noESl noEB noEIg 0.000 1.000 0.000 1.000 0.000 1.000 1

3 0.0 0.0 0.0 0.0 0.0 0.0 Otherwise (Bypass) 0.000 1.000 0.000 1.000 0.000 1.000 1

3 0.0 0.0 0.0 0.0 0.0 0.0 21 Does containment fail due to an early deflagration?

2 EH noEH l 5 1 2 3 1 2 3 EPBase DPELCDef DPEUCDef

'AND'

' NORMAL' 2 75.0 7.5 l

65.0 6.0 51.0 6.0 CFP MEAN SIGMA 22 To what degree is the auxiliary building bypassed as a result of early containment failure?

3 ESB3 ESB2- ESB1 2 1 2 3 6

4 1 11 1

EESB3 1.000 0.000 0.000 1.000 0.000 0.000 '

1.000 0.000 0.000 2 11 21

-1 1 noEESB3 EH 0.500 0.500 0.000 0.900 0.100 0.000 O.999 0.001 0.000 3 11 18 21 2 -1 2 ,

C-10 l

l

EESB2 noEDt-EH noEH 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 2 11 18 l 3 1

! EESB1 EDt-EH l 0.000 0.100 0.900 O.500 0.250 0.250 0.999 0.001 0.000 2 11 18 2 1 EESB2 EDt-EH 0.000 1.000 0.000 0.500 0.500 0.000 0.999 0.001 0.000 Otherwise

, 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 23 Ice melted or bypass paths developed before vessel breach?

3 EnoIce EIceBP noEfIC 2 1 2 3 10 1 18 2

EDt-fIC 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 2 13 3 1 6 ELg EEPORV 0.000 0.000 1.000 0.000 0.500 0.500 0.100 0.800 0.100 2 13 3 2 6 ES1 EEPORV 0.000 0.000 1.000 0.000 0.500 0.500 0.100 0.800 0.100 2 4 8 1 1 EELg ECCInj 0.000 0.000 1.000 0.000 0.001 0.999 0.000 0.100 0.900 2 13 8 1 1 ELg ECCInj 0.000 0.000 1.000

! 0.000 0.500 0.500 0.100 0.800 0.100 3 13 8 4

-1 1 2 l noELg ECCInj EES1 l

C-11

0.000 0.000 1.000 0.000 0.001 0.999 0.000 0.100 0.900 2 13 8 2 1 ES1 ECCInj 0.000 0.000 1.000 0.000 0.500 0.500 0.100 0.800 0.100 4 4 8 13 13 3 1 -1 -2 EES2 ECCInj noELg noes 1 0.000 0.000 1.000 0.000 0.001 0.999 0.000 0.100 0.900 2 13 8 3 1 ES2 ECCInj 0.000 0.000 1.000 0.000 0.001 0.999 0.000 0.100 0.900 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.001 0.999 24 What is the status of the air return fans after hydrogen burns 3 EFan EaFan EfFan 2 1 2 3 7

3 7 19 20 1 1 2 EEFan ELCDet noEUCDef 1.000 0.0 s0 0.000 1.000 0.0:30 0.000 0.999 0.000 0.001 2 7 20 1 1 EEFan EUCDef 1.000 0.000 0.000 0.999 0.000 0.001 0.900 0.000 0.100 3 7 19 20 2 1 2 EEaFan ELCDef noF:iCDef 0.000 1.000 C.000 0.000 1.000 0.000 0.000 0.999 0.001 2 7 20 2 1 EEaFan EUCDef 0.000 1.000 0.000 0.000 0.999 0.001 0.000 0.900 0.100 3 7 19 20 1 2 2 EEFan noELCDefnoEUCDef 1.000 0.000 0.000 C-12

1.000 0.000 0.000 1.000 0.000 0.000 3 7 19 20 2 2 2 EEaFan noELCDefnoEUCDef 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 0.000 0.000 1.000 i 0.000 0.000 1.000 0.000 0.000 1.000 25 What is the status of containment sprays after hydrogen burns?

3 esp EaSp EfSp 2 1 2 3 7

2 10 18 1 1 EESp EDt-EH 1.000 0.000 0.000 1.000 0.000 0.000 0.999 0.000 0.001 2 10 21 1 1 EESp EH 0.900 0.000 0.100 0.900 0.000 0.100 0.100 0.000 0.900 3 10 18 21 1 -1 2 EESp noEDt-EH noEH 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 10 18 2 1 EEaSp EDt-EH 0.000 1.000 0.000 0.000 1.000 0.000 0.000 0.999 0.001 2 10 21 2 1 EEaSp EH 0.000 0.900 0.100 0.000 0.900 0.100 0.000 0.100 0.900 3 10 18 21 2 -1 2 EEaSp noEDt-EH noEH 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 1 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 26 What is the primary system pressure during core degradation?

C-13

.. . . . . _ . _ -- -_. _ . - . ~ . - - ~

b t

3 hip imp lop 4 1 2 3 10 4 4 12 23 24 4 5 3 -1 EES3 noEfRCPB noEfIC noEFan 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1

4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 4 4 12 23 24 4 5 3 1

EES3 noEfRCPB noEfIC EFan

, 1.000 0.000 0.000 l 1.000 0.000 0.000 1.000 0.000 0.000 i 1 4 5.0 5.0 0.0 5.0 5.0 0.0 5.0 5.0 0.0 i 3 4 12 23 i

4 5 -3

EES3 noEfRCPB EfIC 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1

4 37.0 0.0 0.0 37.0 0.0 0.0 37.0 0.0 0.0 4 4 12 23 24 3 5 3 -1 EES2 noEfRCPB noEfIC noEFan 0.000 1.000 0.000 i

0.000 1.000 0.000 O.000 1.000 0.000 1

l 4 0.0 0.0 0.0

~

0.0 0.0 0.0 0.0 0.0 0.0 4 4 12 23 24 3 5 3 1 EES2 noEfRCPB noEfIC EFan 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 1

4 5.0 5.0 0.0 5.0 5.0 0.0 5.0 5.0 0.0 3 4 12 23 3 5 -3

! EES2 noEfRCPB EfIC i 0.000 1.000 0.000 l

C-14 i

_ . _ _ _ _ _ _ _ . , - _ -_ . . , . . - . _ , . _ _ # . . _ _ . . , . , . , . . . . . . _ _ _ _ . . , , - . , , _ . , _ - . _.m_..__... . ~ .

6 0.000 1.000 0.000 0.000 1.000 0.000 1

4 0.0 19.0 0.0 0.0 19.0 0.0 0.0 19.0 0.0 3 13 23 24 3 3 -1 ES2 noEfIC noEFan 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 1

4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 7 3 13 23 24 3 3 1 ES2 noEfIC EFan 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 1

4 5.0 5.0 0.0 5.0 5.0 0.0 5.0 5.0 0.0 2 13 23 3 -3 ES2 EfIC 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 1

4 0.0 19.0 0.0 0.0 19.0 0.0 0.0 19.0 0.0 Otherwise (S1 or larger) 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 1

4 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 0.0 27 What is the mode of vessel breach?

5 alpha SE-UH SE-BH PEj Pour 2 1 2 3 4 5 4 3 1 26 '

1 l hip O.0000 0.0000 0.0000 1.0000 0.0000 0.0001 0.0000 0.2100 0.7899 0.0000

, 0.1000 0.0000 0.4700 0.4300 0.0000 i 1 26 l i 2 imp i C-15

0.0000 0.0000 0.0000 1.0000 0.0000 0.0001 0.0000 0.2100 0.7899 0.0000 0.1000 0.0000 0.4700 0.4300 0.0000 Otherwise (lop) 0.0000 0.0000 0.0000 0.0000 1.0000 0.0001 0.0000 0.2100 0.0000 0.7899 0.1000 0.0000 0.4700 0.0000 0.4300 28 Is the reactor cavity wet at vessel breach?

2 RCWet RCDry 2 1 2 2

2 8 9 2 -1 noECCInjnoEESpInj 0.000 1.000 0.000 1.000 0.000 1.000 Otherwise 1.000 0.000 1.000 0.000 1.000 0.000 29 What is the magnitude of containment pressure due to a steam spike at vessel breach and, if it occurs, direct heating?

2 DH noDH 4 1 2 6

4 23 26 27 28

~3 1 4 2 EfIC hip PEj RCDry 1.000 0.000 1.000 0.000 1.000 0.000 1

5 23.0 0.0 90.0 0.0 213.0 0.0 3 23 27 28

-3 3 2 EfIC SE-BH RCDry 0.000 1.000 0.000 1.000 0.000 1.000 1

5 23.0 23.0 31.0 31.0 54.0 54.0 4 23 26 27 28

-3 2 4 2 EfIC imp PEj RCDry 1.000 0.000 1.000 0.000 1.000 0.000 1

5 0.0 0.0 30.0 0.0 80.0 0.0 1 27 C-16

s r

'z 2

SE-UH 1.000 0.000 1.000 0.000 1.000 0.000 1

5 1.0 0.0 l 31.0 0.0 1 100.0 0.0 2 23 28

-3 1 EfIC RCWet 0.000 1.000 0.000 1.000 0.000 1.000 1

5 23.0 23.0 31.0 31.0 54.0 54.0 Otherwise 0.000 1.000 0.000 1.000 0.000 1.000 1

5 0.0 0.0 0.0 0.0 0.0 0.0 30 Does a detonation occur in the ice condenser?

3 IDt-IH IDtfIC noIDt 2 1 2 3 4 5 1 18 1

EDt-EH 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 3 18 23 24 2 -1 -1 EDtfIC E!ce noEFan 0.000 1.000 0.000 0.100 0.900 0.000 0.500 0.500 0.000 3 18 23 24 2 -1 1 EDtfIC EIce EFan 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 3 18 23 24 3 -1 -1 noEDt EIce noEFan 0.000 0.000 1.000 0.100 0.100 0.800 0.500 0.500 0.000 Otherwise 0.000 0.000 1.000 I C-17 1

I

. - - .= . = . . .. _= -. .. . . . . -

1 i

! 0.000 0.000 -1.000 0.000 0.000 1.000 31 Has ice melted or.have. bypass paths developed after vessel breach?

3 InoIce IIceBP noIfIC i 2 1 2 3 18 1 23 1

EnoIce

1.000 0.000 0.000

1.000 0.000 0.000 1.000 0.000 0.000 1 30

, 2 IDtfIC 0.000 1.000 0.000-0.000 1.000 0.000 0.000 1.000 0.000 5 28 3 23 13 13

' 2 6 3 -1 -2 RCDry EEPORV noEfIC noELg noes 1 0.000 0.000 1.000 0.000 0.500 0.500 0.100 0.800 0.100 5 28 3 23 13 13 2 6 2 -1 -2

! RCDry EEPORV EIceBP noELg noes 1

0.000 1.000 0.000 i 0.001 0.999 0.000 O.100 0.900 0.000 i 5 28 3 23 13 13 I 1 6 3 -1 -2 I RCWet EEPORV noEfIC noELg noes 1

0.000 0.001 0.999 l 0.001 0.100 0.899 0.500

~

0.100 0.400

5 28 3 23 13 13 1 6 2 -1 -2 RCWet EEPORV EIceBP noELg noes 1

0.000 1.000 0.000 1 0.001 0.999 0.000 1 0.100 0.900 0.000 3 4 8 23 1 1 3 j EELg ECCInj noEfIC j 0.000 0.001 0.999

0.001 0.100 0.899 j 0.100 0.500 0.400 l

3 13 8 23

1 1 3 i ELg ECCInj noEfIC

0.001 0.100 0.899 0.100 0.500 0.400

! 0.500 0.500 0.000 l 4 4 8 13 23

! 2 1 -1 3 l EESl ECCInj. noELg noEfIC 1

i C-18 i

I._ _ . _ _ _ _ _ ,,_ _. . . _ - - . ...-_. _ .._. - ._, _ . , _ -. . , . . ,

0.000 0.001 0.999 0.001 0.100 0.899 0.100 0.500 0.400 3 8 13 23 1 2 3 ECCInj ES1 noEfIC 0.001 0.100 0.899 l

0.100 0.500 0.400 0.500 0.500 0.000 5 4 8 23 13 13 3 1 3 -1 -2 I EES2 ECCInj noEfIC noELg noESl 0.000 0.001 0.999 0.001 0.100 0.899 0.100 0.500 0.400 3 8 13 23 1 3 3 ECCInj ES2 noEfIC 0.000 0.001 0.999 0.001 0.100 0.899

! 0.100 0.500 0.400

) 4 4 8 13 23 2 1 -1 2 EES1 ECCInj noELg EIceBP 0.001 0.999 0.000 0.100 0.900 0.000 0.500 0.500 0.000 3 8 13 23 1 2 2 ECCInj ESl EIceBP 0.001 0.999 0.000 0.100 0.900 0.000 0.500 0.500 0.000 5 4 8 23 13 13 3 1 2 -1 -2 EES2 ECCInj EIceBP noELg noS1 0.000 1.000 0.000 0.001 0.999 0.000 0.100 0.900 0.000

3 8 13 23

1 3 2 1 ECCInj ES2 EIceBP 0.000 1.000 0.000 0.001 0.999 0.000 0.100 0.900 0.000 4

1 23 2

EIceBP 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 0.000 0.000 1.000 0.000 0.100 0.900 0.000 0.500 0.500 32 Does a hydrogen burn occur at vessel breach?

2 IDef noIDef i

C-19 .

i I

T

.g - - . ww4- >v ,y -m -p-p w-n>- - ---iw -m--. - w w - --- - --

4 1 2 9

5 14 15 19 20 24 3 1 2 2 1 noEB EIg noELCDef noEUCDef EFan 1.000 0.000 1.000 0.000 1.000 0.000 1

6 5.0 0.0 30.0 0.0 100.0 0.0 5 14 15 19 20 24 3 2 2 2 1 noEB noEIg noELDef noEUCDef Efan 0.900 0.100 0.900 0.100 0.900 0.100 1

6 5.0 0.0 30.0 0.0 100.0 0.0 5 14 15 19 20 24 3 1 1 2 1 noEB EIg ELCDef noEUCDef EFan 1.000 0.000 1.000 0.000 1.000 0.000 1

6 5.0 0.0 12.0 0.0 30.0 0.0 5 14 15 19 20 24 3 2 1 2 1 noEB noEIg ELCDef noEUCDef Efan 0.900 0.100 0.900 0.100 0.900 0.100 1

6 5.0 0.0 12.0 0.0 30.0 0.0 5 14 15 19 20 24 3 1 2 2 -1 noEB EIg noELCDef noEUCDef noEFan 1.000 0.000 1.000 0.000 1.000 0.000 1

6 5.0 0.0 60.0 0.0 100.0 0.0 5 14 15 19 20 24 3 2 2 2 -1 noEB noEIg noELCDef noEUCDef noEfan 0.900 0.100 0.900 0.100 C-20

0.900 0.100 1

6 5.0 0.0 60.0 0.0 100.0 0.0 7 4 4 13 13 13 14 15

-1 -2 -1 -2 -3 3 1 noEELg noEES1 noELg noEIm noes 1 noEB EIg 1.000 0.000 1.000 0.000 1.000 0.000 1

6 1.0 0.0 20.0 0.0 45.0 0.0 7 4 4 13 13 13 14 15

-1 -2 -1 -2 -3 3 2 noEELg noEES1 noELg noEIm noESl noEB noEIg 0.900 0.100 0.900 0.100 0.900 0.100 1

6 1.0 0.0 20.0 0.0 45.0 0.0 Otherwise 0.000 1.000 0.000 1.000 0.000 1.000 1

6 0.0 0.0 0.0 0.0 0.0 0.0 33 Does containment fail at vessel breach?

2 CF@VB noCFGVB 5 1 2 4 1 4 5 6 EPBase DPVB D?DHSS DPIDef

'AND'

' NORMAL' 2 75.0 7.5 65.0 6.0 51.0 6.0 CFP MEAN SIGMA 34 What is the mode of intermediate containment failure?

5 IA ID IHS IS noICF 2 1 2 3 4 5 6

1 27 1

alpha 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 6 2 18 21 27 29 33

-1 -1 2 -1 1 1 noEE3 noEDt-EH noEH noalpha DH CFGVB 0.000 1.000 0.000 0.000 0.000 C-21

0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 7 2 18 21 27 29 32 33

-1 -1 2 -1 2 2 1 noEE3 noEDt-EH noEH noalpha noDH IDef CFGVB 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 6 2 18 21 27 29 30

-1 -1 2 -1 2 1 noEE3 noEDt-EH noEH noalpha noDH IDt-IH 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 7 2 18 21 27 29 32 33

-1 -1 2 -1 2 2 1 noEE3 noEDt-EH noEH noalpha noDH noIDef CFGVB 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 Otherwise 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 0.000 0.000 0.000 0.000 1.000 35 To what degree is there bypass of the auxiliary building after intermediate containment failure?

3 ISB3 ISB2 ISB1 2 1 2 3 8

1 22 1

ESB3 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1 34 1

IA 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 4 22 30 34 34 2 -1 -1 -5 ESB2 noIDt-IH noIA ICF 0.500 0.500 0.000 0.900 0.100 0.000 0.999 0.001 0.000 4 22 30 34 34 3 -1 -1 -5 ESB1 noIDt-IH noIA ICF 0.500 0.500 0.000 0.900 0.100 0.000 0.999 0.001 0.000 3 22 30 34 2 -1 5 ESB2 noIDt-IH noICF 0.000 1.000 0.000 C-22

0.000 1.000 0.000 0.000 1.000 0.000 2 22 30 3 1 ESB1 IDt-IH 0.000 0.100 0.900 0.500 0.250 0.250 0.999 0.001 0.000 2 22 30 2 1 ESB2 IDt-IH 0.000 1.000 0.000 0.500 0.500 0.000 0.999 0.001 0.000 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 36 Are air return fans on after vessel breach?

3 IFan IaFan IfFan 2 1 2 3 5

2 24 32 1 1 EFan IDef 1.000 0.000 0.000 0.999 0.000 0.001 0.900 0.000 0.100 2 24 32 2 1 EaFan IDef 0.000 1.000 0.000 0.000 0.999 0.001 0.000 0.900 0.100 2 24 32 1 2 EFan noIDef 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 24 32 2 2 EaFan noIDef 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 37 Are containment sprays on after vessel breach?

3 ISp Iasp Ifsp 2 1 2 3 5

2 25 34 1 5 esp noICP C-23

0.999 0.000 0.001' O.900 0.000 0.100 0.500 '0.000 0.500 3- 25 34 '34 1 -1 -5

-esp noAlpha .ICF 0.900 0.000 0.100 0.900 0.000 0.100 0.100 0.000 0.900 2 25 34 2 5 i EaSp noICF 4 0.000 0.999 0.001 0.000 0.900 0.100 0.000 0.500 0.500 l 3 25 34 :34

. 2 -1 -5 i i EaSp noAlpha .ICF 0.000 0.900 0.100 0.000 0.900 0.100 0.000 0.100 0.900

Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000

3) Is ac power restored after vessel breach? This is a function of the time credited for power recovery in the sequences comprising the plant-damage states.

2 L-AC LfAC 2 1 2

. 2 1 1 1

E-AC 1.000 0.000 1.000 0.000 1.000 0.000 Otherwise 0.750 0.250 0.750 0.250 0.750 0.250 39 Are the igniters on late?

2 LIg noLIg i 2 1 2 1 4 1 15 I 1 EIg 1.000 0.000 1.000 0.000 1.000 0.000 3 1 15 38 1 2 1 E-AC noEIg L-AC 0.500 0.500 0.100 0.900 l,- 0.000 1.000 l

C-24

l l

3 1 15 38 2 2 1 EfAC noEIg L-AC 1.000 0.000 0.999 0.001

! 0.900 0.100 Otherwise 0.000 1.000 0.000 1.000 0.000 1.000 40 Do the air return fans operate late?

3 LFan LaFan noLFan 2 1 2 3 4

1 36 1

IFan 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 36 38 2 1 IaFan L-AC 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 36 38 2 2 IaFan LfAC 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 41 Do contlinment sprays operate late?

3 LSp LaSp noLSp 2 1 2 3 4

1 37 1

ISp 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 37 38 2 1 IaSp L-AC  !

1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 37 38 '

2 2 ,

IaSp LfAC l 0.000 1.000 0.000 l 0.000 1.000 0.000 C-25

0.000 1.000 0.000 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 42 What is the late baseline pressure in containment? The pressures in this question must be adjusted to accomodate damage states that include success of containment sprays but unavailability of containment heat removal.

1 LPBase 4 1 5

2 31 40 3 1 noIfIC LFan 1.000 1.000 1.000 ,

1 7 18.0 19.0 20.0 3 31 40 41 3 -1 -1 noIfIC noLFan noLSp 1.000 1.000 1.000 1

7 20.0 25.0 35.0 3 31 40 41

-3 1 1 IfIC LFan LSp 1.000 1.000 1.000 1

7 18.0 19.0 20.0 2 40 41

-1 1 noLFan LSp 1.000 1.000 1.000 1

7 19.0 21.0 23.0 Otherwise (no CHR of any kind) 1.000 1.000 1.000 1

C-26

7 20.0 30.0 40.0 43 Is a coolable debris bed formed and maintained after vessel breach?

2 CDB noCDB 2 1 2 3

2 28 26 2 -3 RCDry notLoP 0.900 0.100 0.100 0.900 0.001 0.999 2 27 28 5 1 Pour RCWet _

0.900 0.100 0.500 0.500 0.001 0.999 otherwise 0.999 0.001 0.900 0.100 0.500 0.500 44 Is there a late hydrogen burn?

2 LHB noLHB 4 1 2 4

6 2 18 21 34 39 43

-1 -3 2 5 2 2 noEE3 noEDt-EH noEH noICF noLIg noCDB 0.900 0.100 0.900 0.100 0.900 0.100 1

8 5.0 0.0 35.0 0.0 185.0 0.0 8 2 18 19 20 21 32 34 39

-1 -1 2 2 2 2 5 1 noEE3 noEDt-EHnoELCDefnoEUCDef noEH noIDef noICF LIg 1.000 0.000 1.000 0.000 1.000 0.000 1

8 5.0 0.0 25.0 0.0 55.0 0.0 5 2 18 21 34 39

-1 -1 2 5 1 noEE3 noEDt-EH noEH noICF LIg 1.000 0.000 1.000 0.000 1.000 0.000 1

8 5.0 0.0 18.0 0.0 25.0 0.0 C-27

Otherwise

! 0.000 1.000 0.000 1.000 0.000 1.000 1

8 0.0 0.0 0.0 0.0 0.0 0.0 45 Does containment fail due to a. late burn?

2 LH noLH

, 6 1 2

2 4 2 18 21 34

-1 -1 2 5 noEE3 noEDt-EH noEH -noICF 2 7 8 LPBase DPLHB

'AND' l

' NORMAL' 2 75.0 7.5 65.0 6.0 51.0 6.0 CFP MEAN SIGMA 2 Otherwise

2 7 8 LPBase DPLHB

'AND'

' NORMAL' 2 200.0 1.0 200.0 1.0 200.0 1.0 l CFP MEAN SIGMA 46 Do containment sprays operate very late?

3 LLSp LLaSp LLfSp 2 1 2 3 5

2 41 45 1 2 LSp noLH 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 41 45 1 2 2 i LaSp noLH i

0.000 1.000 0.000

! 0.000 1.000 0.000 2

0.000 1.000 0.000 2 41 45 1 1 LSp LH 0.900 0.000 0.100 0.900 0.000 0.100 0.100 0.000 0.900 2 41 45 2 1 a LaSp LH

., 0.000 0.900 0.100 1 0.000 0.900 0.100 C-28 l

0.000 0.100 0.900 Otherwise 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 47 Does containment failure occur late due to noncondensible gas generation or steam?

2 LP noLP 2 1 2 4

7 2 2 18 21 34 43 46

-1 -2 -1 2 5 2 1 noEE3 noEE2 noEDt-EH noEH noICF noCDB LLSp 0.000 1.000 0.900 0.100 0.999 0.001 2 43 46 1 1 CDB LLSp 0.000 1.000 0.000 1.000 0.000 1.000 6 2 2 18 21 34 46

-1 -2 -1 2 5 -1 noEE3 noEE2 noEDt-EH .noEH noICF noLLSp 1.000 0.000 1.000 0.000 1.000 0.000 Otherwise 0.000 1.000 0.000 1.000 0.000 1.000 48 To what degree is there bypass of the auxiliary building after late containment failure?

3 LSB3 LSB2 LSB1 2 1 2 3 5

1 35 1

ISB3 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 0.000 2 35 45

-1 1 noISB3 LH 0.500 0.500 0.000 0.900 0.100 0.000 0.999 0.001 0.000 2 35 47

-1 1 noISB3 LP 0.500 0.500 0.000 0.900 0.100 0.000 0.999 0.001 0.000 3 35 45 47 2 2 2 C-29

ISB2 noLH noLP 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 Otherwise 0.000 1.000 0.000 0.000 1.000 0.000 0.000 1.000 0.000 49 Does basemat meltthrough occur?

2 MT noMT 2 1 2 2

2 43 46 1 1 CDB LLSp  ;

0.000 1.000 '

O.000 1.000 '

0.000 1.000 Otherwise 1.000 0.000 1.000 0.000 1.000 0.000

'SEQUOYAH - S2IYBY - DH and Alpha (Binned by Containment Release Mode) '

' OPTIMISTIC' ' CENTRAL' ' PESSIMISTIC' 5 'CFM' 'CI' 'SGTR' 'IceBVB' 'IceBCCI' 10 12 'NoCF' 'MT' 'LP' 'LH' 'IS' 'IHS' 'ID' 'IA' 'EH'

'CI' 8 1 2 2 18 21 34 45 47 49

-1 -2 -1 2 5 2 2 2 noEE3 noEE2 noEDt-EH noEH noICF noLH noLP noMT 8 2 2 2 18 21 34 45 47 49

-1 -2 -1 2 5 2 2 1 noEE3 noEE2 noEDt-EH noEH noICF noLH noLP MT 4 3 18 21 34 47

-1 2 5 1 noEDt-EH noEH noICF LP 4 4 18 21 34 45

-1 2 5 1 noEDt-EH noEH noICF LH 3 5 18 21 34

-1 2 4 noEDt-EH noEH IS 3 6 18 21 34

-1 2 3 noEDt-EH noEH IHS 3 7 18 21 34

-1 2 2 noEDt-EH noEH ID 3 8 18 21 34

-1 2 1 noEDt-EH noEH IA 1 9 18 1

noEDt-EH 2 9 18 21

-1 1 noEDt-EH EH C-30

1 10 2 1

EE3 1 10 2 2

EE2 3 4 'Large' 'Small' 'None' 1 1 2 1

EE3 1 1 2 2

EE2 1 2 2 3

EE1 1 3 2 4

EE0 2 2 'SGTR' 'NoSGTR' 1 1 12 4

ESGTR 1 2 12

-4 noESGTR 2 2 'EfIC' 'noEfIC' 1 1 23

-3 EfIC 1 2 23 3

noEfIC 2 2 'IfIC' 'noIfIC' 1 1 31

-3 IfIC 1 2 31 3

noIfIC l

l l

l I

c-31

Appendix D REPRESENTATIVE COMPUTER OUTPUT FOR SEQUOYAH CONTAINMENT EVENT TREE The computer output for the containment event tree for three plant damage states (S2IYBY, S3INIYB, and TNNNNN) is listed in this appendix. The results listed are presented at a detail that permits examination of some of the additional parameters associated with the outcomes, including the degree to which there is a failure of containment isolation, whether or not there is an induced steam generator tube rupture, and the status of the ice condenser at the time of vessel breach and at the inception of core-concrete interactions.

1 I

I 1

D-1

SEQUOYAH - S2IYBY - DH and Alpha (Binned by Containment Release Mode)

RESULTS FOR THE OPTIMISTIC WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 1.6949E-02 NoCF Small NoSGTR noEfIC noIfIC 9.8104E-04 NoCF None NoSGTR noEfIC IfIC 9.8005E-01 NoCF None NoSGTR noEfIC noIfIC 9.8104E-04 MT None NoSGTR noEfIC noIfIC 9.8202E-04 LP None NoSGTR noEfIC noIfIC ,

.9999471 TOTAL FREQUENCY OF BINS PRINTED l

TOTAL PROBABILITY IN ALL BINS = 1.000000 l TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 5.1949E-08 l TOTAL PROBABILITY FOR THIS WALKTHROUGH = 1.000000 D-2

SEQUOYAH - S2IYBY - DH and Alpha (Binned by Containment Release Mode)

RESULTS FOR THE CENTRAL WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 2.6504E-03 NoCF Small NoSGTR noEfIC IfIC 2.3601E-02 NoCF Small NoSGTR noEfIC noIfIC 7.7500E-02 NoCF None NoSGTR noEfIC IfIC 6.8984E-01 NoCF None NoSGTR noEfIC noIfIC 2.6156E-04 MT Small NoSGTR noEfIC noIfIC 8.5997E-04 MT None NoSGTR noEfIC IfIC 7.6632E-03 MT None NoSGTR noEfIC noIfIC 5.8846E-04 LP Small NoSGTR noEfIC IfIC 5.2695E-03 LP Small NoSGTR noEfIC noIfIC 1.7289E-02 LP None NoSGTR noEfIC IfIC 1.5414E-01 LP None NoSGTR noEfIC noIfIC 4.9382E-04 IHS Small NoSGTR noEfIC noIfIC 9.6019E-04 IHS None NoSGTR EfIC IfIC 1.6250E-03 IHS None NoSGTR noEfIC IfIC 1.4480E-02 IHS None NoSGTR noEfIC noIfIC 2.4967E-04 CI Large NoSGTR noEfIC IfIC 2.2421E-03 CI Large NoSGTR noEfIC noIfIC

.9997084 TOTAL FREQUENCY OF BINS PRINTED TOTAL PROBABILITY IN ALL BINS = .9999999 TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 4.7975E-05 TOTAL PROBABILITY FOR THIS WALKTHROUGH = 1.000048 I

l D-3

~ __. . .. . . _ _ _. ._. _ _ _ _ __. , _.._____--_.

l-

, SEQUOYAH - S2IYBY - DH and Alpha (Binned by Containment Release Mode) 1 RESULTS FOR THE PESSIMISTIC WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 1.4456E-03 NoCF Small NoSGTR noEfIC IfIC 9.6395E-04 NoCF Smal. NoSGTR noEfIC noIfIC 2.3385E-02 NoCF None NoSGTR noEfIC IfIC 1.5590E-02 NoCF None NoSGTR noEfIC noIfIC i 1.4786E-03 LP Small NoSGTR noEfIC IfIC 2.7213E-03 LP Small NoSGTR noEfIC noIfIC 2.3964E-02 LP None NoSGTR noEfIC IfIC l 4.4021E-02 LP None NoSGTR noEfIC noIfIC

4.0485E-03 LH Small NoSGTR noEfIC IfIC i

9.6552E-04 LH Small NoSGTR noEfIC noIfIC 6.5533E-02 LH None NoSGTR noEfIC IfIC j 1.5643E-02 LH None NoSGTR noEfIC noIfIC

- 2.5275E-04 IS None NoSGTR EfIC IfIC 5.2001E-03 IHS Small NoSGTR EfIC IfIC

, 2.1211E-02 IHS Small NoSGTR noEfIC IfIC a

1.4126E-02 IHS Small NoSGTR noEfIC. noIfIC 8.4163E-02 IHS None NoSGTR EfIC IfIC 3.4310E-01 IHS None NoSGTR noEfIC IfIC 2.2849E-01 IHS None NoSGTR noEfIC noIfIC 1 5.7865E-04 IA Small NoSGTR EfIC IfIC 3.1320E-03 IA Small NoSGTR noEfIC IfIC 2.0868E-03 IA Small NoSGTR noEfIC noIfIC j 9.3797E-03 IA s.one NoSGTR EfIC IfIC

! 5.0667E-02 IA None NoSGTR noEfIC IfIC i 3.3751E-02 IA None NoSGTR noEfIC noIfIC 3.9245E-04 CI Large NoSGTR EfIC IfIC

]' 2.1477E-03 CI Large NoSGTR noEf1C IfIC 1.4331E-03 CI Large NoSGTR noEfIC noIfIC

.9998689 TOTAL FREQUENCY OF BINS PRINTED TOTAL PROBABILITY IN ALL BINS - 1.0000C0

' TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 7.6221E-05

, TOTAL PROBABILITY FOR THIS WALKTHROUGH = 1.000076 i

• • • • • • • PROCESSING COMPLETE FOR 3 WALKTHROUGHS *******

l i

l D-4 L

l

- . - - . _ , - - - . . . . - . - . . - ,-- --,, , - . - . . , - - , - ~ - , .. . . , . - . . . -,.

SEQUOYAH - S3Ih*IYB - DH and Alpha (Binned by Containment Release Mode)

RESULTS FOR THE OPTIMISTIC WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 4.4638E-04 LP Small NoSGTR noEfIC IfIC 1.6554E-02 LP Small NoSGTR noEfIC noIfIC 2.5814E-02 LP None NoSGTR noEfIC IfIC 9.5719E-01 LP None NoSGTR noEfIC noIfIC

.9999999 TOTAL FREQUENCY OF BINS PRINTED TCTAL PROBABILITY IN ALL BINS = .9999999 TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 4.4200E-08 TOTAL PROBABILITY FOR THIS WALKTHROUGH = 1.000000 l

l D-5

SEQUOYAH - S3INIYB - DH and Alpha (Binned by Containment Release Mode)

RESULTS FOR THE CENTRAL WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 2.9565E-03 LP Small NoSGTR EfIC If1C 4.9573E-03 LP Small NoSGTR noEfIC IfIC 2.3319E-02 LP Small NoSGTR noEfIC noIfIC 8.6437E-02 LP None NoSGTR EfIC IfIC 1.4501E-01 LP None NoSGTR noEfIC IfIC 6.8169E-01 LP None NoSGTR noEfIC noIfIC 2.4050E-04 LH None NoSGTR EfIC IfIC

] 3.7932E-04 LH None NoSGTR noEfIC IfIC 1.3230E-03 IHS Small NoSGTR EfIC IfIC 3.6698E-04-IHS Small NoSGTR ncEfIC noIfIC 3.8693E-02 IHS None NoSGTR EfIC IfIC 1.1925E-03 IHS None NoSGTR noEfIC IfIC 1.0738E-02 IHS None NoSGTR noEfIC noIf1C 3.2364E-04 CI Large NoSGTR EfIC IfIC 3.7812E-04 CI Large NoSGTR noEfIC IfIC 1.7930E-03 CI Large NoSGTR noEfIC noIfIC

.9998046 TOTAL FREQUENCY OF BINS PRINTED TOTAL PROBABILITY IN ALL BINS = .9999999 l TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 3.0655E-05 TOTAL PROBABILITY FOR THIS WALKTHROUGH = 1.000031 D-6

y SEQUOYAH - S3INIYB - DH and Alpha (Binned by Containment Release Mode)

RESULTS FOR THE PESSIMISTIC WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 3.6376E-03 LP Small NoSGTR noEfIC noIfIC

1.2570E-03 LP None NoSGTR noEfIC IfIC l 5.8837E-02 LP None NoSGTR noEfIC noIfIC 4.7030E-03 LH Small NoSGTR noEfIC IfIC 1.1403E-03 LH Small NoSGTR noEfIC noIfIC 7.6077E-02 LH None NoSGTR noEfIC IfIC 1.8449E-02 LH None NoSGTR noEfIC noIfIC 3.77SSE-04 IS None NoSGTR EfIC IfIC 1.4524E-02 IHS Small NoSGTR noEfIC IfIC 1.4506E-02 IHS Small NoSGTR noEfIC noIfIC 7.6819E-04 IHS None NoSGTR EfIC IfIC 2.3491E-01 IHS None NoSGTR noEfIC IfIC 2.3460E-01 IHS None NoSGTR noEfIC noIfIC 2.1447E-03 IA Small NoSGTR noEfIC IfIC 2.1426E-03 IA Small NoSGTR noEfIC nolfIC 1.2674E-04 IA None NoSGTR EfIC IfIC 3.4693E-02 IA None NoSGTR noEfIC IfIC 3.4654E-02 IA None NoSGTR noEfIC noIfIC 1.3532E-02 EH Small NoSGTR EfIC IfIC 1.5006E-03 EH Small NoSGTR noEfIC IfIC 2.1891E-01 EH None NoSGTR EfIC IfIC 2.4322E-02 EH None NoSGTR noEfIC IfIC 9.3406E-04 CI Large NoSGTR EfIC IfIC 1.5788E-03 CI Large NoSGTR noEfIC IfIC 1.4762E-03 CI Large NoSGTR noEfIC noIfIC

.9998077 TOTAL FREQUENCY OF BINS PRINTED TOTAL PROBABILITY IN ALL BINS = .9999999 TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 4.0582E-05 TOTAL PROBABILITY FOR THIS WALKTHROUGH = 1.000041-

• • • • • • • PROCESSING COMPLETE FOR 3 WALKTHROUGHS *******

i l

D-7

SEQUOYAH - TNNNNN - DH and Alpha (Binned by containment Release Mode)

RESULTS FOR THE OPTIMISTIC WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 1.6049E-03 NoCF Small SGTR noEfIC noIfIC 1.0010E-02 NoCF Small NoSGTR noEfIC noIfIC 9.2801E-02 NoCF None SGTR noEfIC noIfIC 5.7881E-01'NoCF None NoSGTR noEfIC noIfIC 1.7832E-04 MT Small SGTR noEfIC noIfIC 9.4407E-04 MT Small NoSGTR noEfIC noIfIC 1.0311E-02 MT None SGTR noEfIC noIfIC 5.4590E-02 MT None NoSGTR noEfIC noIfIC 5.9633E-04 LP Small SGTR noEfIC noIfIC 3.6640E-03 LP Small NoSGTR noEfIC noIfIC 3.4508E-02 LP None SGTR noEfIC noIfIC 2.1198E-01 LP None NoSGTR noEfIC noIfIC

.9999975 TOTAL FREQUENCY OF BINS PRINTED TOTAL PROBABILITY IN ALL BINS = 1.000000 TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 2.7253E-06 TOTAL PROBABILITY FOR THIS WALKTHROUGH = 1.000003 I

D-8

SEQUOYAH - TNNNNN - DH and Alpha (Binned by Containment Release Mode)

RESULTS FOR THE CENTRAL WALKTHROUGH i- FREQUENCY CFM CI SGTR IceBVB IceBCCI 1.2457E-04 NoCF Small SGTR noEfIC IfIC 1.1583E-04 NoCF Small NoSGTR EfIC IfIC 1.2584E-04 NoCF Small NoSGTR noEfIC noIfIC 3.1783E-04 NoCF None SGTR EfIC IfIC 3.6446E-03 NoCF None SGTR noEfIC IfIC 2.9157E-03 NoCF None SGTR noEfIC noIfIC

3.3992E-03 NoCF None NoSGTR EfIC IfIC

, 2.5160E-03 NoCF None NoSGTR noEfIC IfIC l 3.6882E-03 NoCF None NoSGTR noEfIC noIfIC 1

1.1212E-04 MT Small SGTR noEfIC IfIC 2.8601E-04 MT None SGTR EfIC IfIC 3.2799E-03 MT ~ None SGTR noEfIC IfIC 2.6240E-03 MT None SGTR noEfIC noIfIC 1.5034E-04 MT None NoSGTR EfIC IfIC 1.7988E-03 MT None NoSGTR noEfIC IfIC 1.4578E-03 MT None NoSGTR noEfIC noIfIC 1.2148E-04 LP Small SGTR EfIC IfIC 1.4149E-03 LP Small SGTR noEfIC IfIC 1.2269E-03 LP Small SGTR noEfIC noIfIC 1.1239E-04 LP Small NoSGTR EfIC IfIC 7.7676E-04 LP Small NoSGTR noEfIC IfIC 7.1164E-04 LP Small NoSGTR noEfIC noIfIC 3.6105E-03 LP None SGTR EfIC IfIC 4.1409E-02 LP None SGTR noEfIC IfIC 3.5887E-02 LP None SGTR noEfIC noIfIC 3.4439E-03 LP None NoSGTR EfIC IfIC 2.2960E-02 LP None NoSGTR noEfIC IfIC 2.0935E-02 LP None NoSGTR noEf7C noIfIC 1.9258E-04 LH Small SGTR r.oE (IC IfIC 1.0175E-04 LH Small NoSGTR noifIC IfIC 4.9372E-04 LH None SGTR EfIC IfIC 5.6715E-03 LH None SGTR noEfIC IfIC 1.7779E-03 LH None SGTR noEfIC noIfIC 2.7580E-04 LH None NoSGTR EfIC IfIC I

3.1067E-03 LH None NoSGTR noEfIC IfIC 9.8701E-04 LH None NoSGTR noEfIC noIfIC 1.8390E-04 IHS Small SGTR noEfIC IfIC 1.8390E-04 IHS Small SGTR noEfIC noIfIC 2.2498E-03 IHS Small NoSGTR EfIC IfIC 9.9902E-03 IHS Small NoSGTR noEfIC IfIC 9.1348E-03 IHS Small NoSGTR noEfIC noIfIC 2.8307E-04 IHS None SGTR EfIC IfIC 5.4002E-03 IHS None SGTR noEfIC IfIC l 5.4002E-03 IHS None SGTR noEfIC noIfIC 6.6107E-02 IHS None NoSGTR EfIC IfIC 2.9240E-01 IHS None NoSGTR noEfIC IfIC 2.6733E-01 IHS None NoSGTR noEfIC noIfIC 2.8990E-04 ID Small SGTR EfIC IfIC 1.9032E-03 ID Small NoSGTR EfIC IfIC

8.5074E-03 ID None SGTR EfIC IfIC 5.5757E-02 ID None NoSGTR EfIC IfIC 2.2915E-04 EH Small SGTR noEf1C IfIC D-9

SEQUOYAH - TNNNNN - DH and Alpha (Binned by Containment Release Mode)

RESULTS FOR THE CENTRAL WALKTHROUGH (CONTINUED) 2.2915E-04 EH Small SGTR noEfIC noIfIC 1.9269E-04 EH Small NoSGTR EfIC IfIC 1.2271E-03 EH Small NoSGTR noEfIC IfIC 1.3838E-03 EH Small NoSGTR noEfIC noIfIC 6.7506E-03 EH None SGTR noEfIC IfIC 6.7506E-03 EH None SGTR noEfIC noIfIC 5.7824E-03 EH None NoSGTR EfIC IfIC 3.6251E-02 EH None NoSGTR noEfIC IfIC 4.0879E-02 EH None NoSGTR noEfIC noIfIC 1.6594E-04 CI Large SGTR noEfIC IfIC 1.3915E-04 CI Large SGTR noEfIC noIfIC 3.3053E-04 CI Large NoSGTR EfIC IfIC 9.0181E-04 CI Large NoSGTR noEfIC IfIC 8.4190E-04 CI Large NoSGTR noEfIC noIfIC i

.9989530 TOTAL FREQUENCY OF BINS PRINTED TOT \L PROBABILITY IN ALL BINS = .9999997 TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 3.1826E-04 TOTAL PROBABILITY FOR THIS WALKTHROUGH - 1.000318 D-10

SEQUOYAH - TNNNNN - DH and Alpha (Binned by Contai. ment Release Mode)

RESULTS FOR THE PESSIMISTIC WALKTHROUGH FREQUENCY CFM CI SGTR IceBVB IceBCCI 4.5754E-04 LP None SGTR EfIC IfIC 5.2996E-04 LP None NoSGTR EfIC IfIC 8.6143E-04 LH Small SGTR EfIC IfIC 3.8918E-04 LH Small NoSGTR EfIC IfIC 1.3941E-02 LH None SGTR EfIC IfIC 6.3174E-03 LH None NoSGTR EfIC IfIC 2.3415E-04 IS Small NoSGTR EfIC IfIC 1.0305E-03 IS None SGTR EfIC IfIC 3.7910E-03 IS None NoSGTR EfIC IfIC 9.5313E-04 IHS Small SGTR EfIC IfIC 1.2550E-02 IHS Small NoSGTR EfIC IfIC 1.5429E-02 IHS None SGTR EfIC IfIC 2.0308E-01 IHS None NoSGTR EfIC IfIC 1.7147E-03 ID Small SGTR EfIC IfIC 9.2236E-03 ID Small NoSGTR EfIC IfIC 2.8233E-02 ID None SGTR EfIC IfIC 1.4923E-01 ID None NoSGTR EfIC IfIC 4.0549E-04 IA Small SGTR EfIC IfIC 2.4916E-03 IA Small NoSGTR EfIC IfIC 6.5657E-03 IA None SGTR Ef1C IfIC 4.0333E-02 IA None NoSGTR EfIC IfIC 3.6477E-03 EH Small SGTR noEfIC IfIC 4.0460E-04 EH Tmall SGTR noEfIC noIfIC 3.1445E-03 EH Small NoSGTR EfIC IfIC 1.9455E-02 EH Small NoSGTR noEfIC IfIC 2.3120E-03 EH Small NoSGTR noEfIC noIfIC 5.3035E-02 EH None SGTR noEfIC IfIC 6.5593E-03 EH None SGTR noEfIC noIfIC 5.0991E-02 EH None NoSGTR EfIC IfIC 3.1485E-01 EH None NoSGTR noEfIC IfIC 3.7478E-02 EH None NoSGTR noEfIC noIfIC 4 2.7715E-04 CI Large SGTR EfIC IfIC 2.5071E-04 CI Large SGTR noEfIC IfIC 1.9116E-03 CI Large NoSGTR EfIC IfIC 1.3313E-03 CI Large NoSGTR noEfIC IfIC 1.5260E-04 CI Large NoSGTR noEfIC noIfIC

.9995883 TOTAL FREQUENCY OF BINS PRINTED TOTAL PROBABILITY IN ALL BINS = 1.000001 TOTAL PROBABILITY LOST DUE TO TERMINATION OF BRANCHES 1.4222E-04 TOTAL PROBABILITY FCR THIS WALKTHROUGH = 1.000143 I

l l

l D-11 i I

1 l

(

l

! l 1

DISTRIBUTION:

i U. S. Government Printing Office Receiving Branch (Attn: NRC Stock)

~8610 Cherry Lane Laurel,-MD- 20707 l

l-I t

-3141 S. A. Landenberger (5) l_ 3151 W. L. Garner i 6400 D. J. McCloskey I 6410 N. R. .Ortiz 6414 A. S. Benjamin (25)-

6412 A. L. Camp

6412 F. T. Harper

! 6412 D.-M. Kunsman

6512 D. M. Ericson, Jr.

6415 F. E. Haskin 6415 J. L. Sprung

. 6513 D. D. Carlson 6420 J. V. Walker 6422 D. A. Powers 1 6427 M. Berman 6440 D. A. Dahlgren 6442 W. A. von Riesemann 6449 K. D. Bergeron

. 6449 D. C. Williams

! 8024 P. W. Dean l

3 i

)

i i

DIST-1

Frank Abbsy

-U.K. Atomic Ensrgy Authority Wigshaw Lane, Culcheth Warrington, Chesire.

United-Kingdom, WA3 4NE I

. Steven Aftergood Committee to Bridge the Gap 1637 Butler Avenue, #203 l

Los Angeles, CA 90025

! C. N. Amos Technadyne,.Inc. ,

300 Virginia SE l Albuquerque, NM 87108

-George Apostolakis University of California, Los Angeles Boelter Hall, Room 5532 Los Angeles, CA 90024 James W. Ashkar Boston Edison Company 800 Boylston Street Boston, MA 02199 Donald H. Ashton Bechtel Power Corporation 15740 Shady Grove Road Gaithersburg, MD 20877 Mark Averett Florida Power Corporation P.O. Box 14042 St. Petersburg, FL 33733 Michael B. Aycock TERA Corporation 7101 Wisconsin Avenue Bethesda, MD 20814 Raymond O. Bagley Northeast Utilities P.O. Box 270 Hartford, CT 06101 J. Bagues Consejo de Seguridad Nucleare Sorangela de la Cruz 3 Madrid 38046 l Spain  !

l DIST-2

H. Bairiot Balgonucleaire S A Rue de Champ de Mars 25 B-1050 Brussels Belgium Fred J. Balkovetz EG&G Idaho, Inc.

P.O. Box 1625 Idaho Falls, ID 83415 ,

Patrick Baranowsky U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research Washington, DC 20555 l Robert A. Bari

! Brookhaven National Laboratories i Building 130 Upton, NY 11973 l

Richard Barrett Baskin

.; U.S. Nuclear Regulatory Commission j Office of Nuclear Reactor Regulation Washington, DC 20555 S. Kenneth S. California Edison Company P.O. Box 800 Rosemead, CA 91770 f

l Werner Bastl Gesellschaft Fur Reaktorsicherheit Forschungsgelande 8046 Garching Federal Republic of Germany Paul Baybutt

Battelle Columbus Division

! 505 King Avenue -

l ColumbusOli 43201 Bruce B. Beckley <

Public Service Company l

! P.O. Box 330 I Manchester, NH 03105 Myer Bender Querytech, Inc.

408 Cedar Bluff Road l Knoxville, TN 37923 DIST-3

.-- -.. . - , _ . _ - - . . - . . - . . . . , - - . - - - - _ _ . . . - . . . - . - - . _ , ~ . _ - ..

R. C. Bertucio

Energy' Incorporated

~

1851 So. Central Place, Suite 201 Kent, WA 98031 Peter Bieniarz Risk Management Association i 2309 Dietz Farm Road, NW Albuquerque, NM 87107 Adolf Birkhofer Gesellschaft Fur Reaktorsicherheit Forschungsgelande 8046 Garching Federal Republic of Germany James Blackburn i Illinois Dept. of Nuclear Safety 1035 Outer Park Drive l Springfield. IL 62704

! Dennis C. Bley

- Pickard, Lowe & Garrick, Inc.

2260 University Drive Newport Beach, CA 92660 Roger M. Blond Science Applications Int. Corp.

1710 Goodridge Drive McLean, VA 22102 A. John Bourne U.K. Atomic Energy Authority Wigshaw Lane, Culcheth Warrington. Chesire j

United Kingdom, WA3 4NE G. J. Boyd Safety & Reliability Optimization Services

! 9724 Kingston Pike Knoxville. TN 37922 K. J. Brinkmann Netherlands Energy Res. Fdtn.

P.O. Box 1 1755ZG Petten NH Netherlands I R. H. Buchholz Nutech 4

6835 Via Del Oro l San Jose, CA 95119 l

I I

I DIST-4

Fred Buckman D211an Corporation 295 Duff Road Monroeville, PA 15146 Robert J. Budnitz Future Resources Associates 734 Alameda '

Berkeley, CA .94707 1

A. R. Buhl International Technology Corp.

575 Oak Ridge Turnpike Oak Ridge. TN 37830 a

Gary R. Burdick i U.S. Nuclear Regulatory Commission i Office of Nuclear Regulatory Research l Washington, DC 20555 Kenneth S. Canady Duke Power Company 422 S. Church Street Charlotte, NC 28217 Lennart Carlsson Swedish Nuclear Power Inspectorate l Box 27106 i Stockholm, Sweden 08-635560 Sweden l Annick Carnino Electricite de France 32 Rue de Monceau 8EME Paris, France F5008 France 4

i N. G. Cathey Idaho National Engineering Lab.

P. O. Box 1625 Idaho Falls, ID 83415 i Ivan Catton ,

i 22436 Galilee Street i

Woodland Hills, CA 91364 S. Chakraborty i Radiation Protection Section i Div. De La Securite Des Inst. Nuc. '

5303 Wurenlingen, Switzerland i Switzerland i

1 I

DIST-5

San-I Chang Institute of Nuclear En9rgy Research P.O. Box 3 Lungtan, 325 Taiwan Robert F. Christie Tennessee Valley Authority 400 W. Summit Hill Avenue, W10D1 90 Knoxville, TN 37902 4

} Frank Coffman U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation-Washington, DC 20555 l

Larry L. Conradi Energy Incorporated 1851 South Central Place, Suite 201 Kent, WA 98031 Peter Cooper U.K. Atomic Energy Authority

! Wigshaw Lane, Culcheth l Warrington, Chesire United Kingdom, WA3 4NE 1

! Michael Corradini

University of Wisconsin 150r Johnson Drive i Mr.dison, WI 53706

Michael Cullingford

IAEA 4 P.O. Box 200 4

A-1400 Vienna, Austria Garth Cummir i Lawrence Lit ire Laboratory

L-91, Box 80.

l Livermore, CA 94526 Mark A, Cunningham U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research j, Washington, DC 20555 i

l I <

1 f

1 DIST-6 I

Jamns J. Curry-7135 Sclem Park Circle Mechanicsburg, PA 17055 Peter Cybulskis Battelle Columbus Division 505 King Avenue

Columbus, OH 43201 Peter R. Davis Intermountain Technologies, Inc.

P.O. Box 1604 Idaho Falls, ID 83401 Jose DeCarlos Sor Angela De la Cruz, No. 3, Planta 8 Madrid 28016, Spain Richard Deem New York Power Authority 123 Main Street White Plains, NY 10601 Richard S. Denning Battelle Columbus Division 505 King Avenue Columbus, OH 43201 1

) Raymond DiSalvo

Battelle Columbus Division J 505 King Avenue Columbus, OH 43201 Adel A. El-Bassioni U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation Washington, DC '20555 J. Mark Elliott International Energy Associates. Ltd.

600 New Hampshire Ave. NW, Suite 600 Washington, DC 20037 Energy Issues Rep.

Sierra Club 1228 North Street #31 Sacramento, CA 95814 Malcolm L. Ernst U.S. Nuclear Regulatory Commission 101 Marietta Street, Suite 3100 Atlanta, GA 30323

, DIST-7

Executive Director Natural Resources Defense Council 122 East 42nd Street New YorkNY 10168 P. Fehrenback Atomic Energy Canada. Ltd.

Chalk River Ontario, Canada A. Fiege Kernforschungszentrum Postfach 3640 7500 Karlsruhe Federal Republic of Germany George F. Flanagan Oak Ridge National Laboratories P.O. Box Y Oak Ridge, TN 37831 Mario Fontana International Technology Corp.

575 Oak Ridge Turnpike Oak Ridge, TN 37830 Ray R. Fray Pacific Gas and Electric Co.

215 Market Street. Room 206 San Francisco, CA 94106 James Fulford NUS Corporation 910 Clopper Road Gaithersburg, MD 20878 J. B. Fussell JBF Associates. Inc.

1630 Downtown West Boulevard Knoxville, TN 37919 Raymond H. V. Gallucci Battelle Pacific Northwest Labs.

P.O. Box 999 i Richland, WA 99352 John Garrick Pickard, Lowe & Garrick. Inc.

2260 University Drive Newport Beach, CA 92660 John Gaunt British Embassy 3100 Massachusetts Avenue, NW Washington, DC 20008 DIST-8 1

J. GionGks B ttoll@ Coluabus Division 505 King Avenue Columbus, OH 43201 James C. Glynn U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research Washington, DC 20555 L. Hammar Statens Karnkraftinspektion P.O. Box 27106 S-10252 Stockholm, Sweden M. R. Hayns U.K. Atomic Energy Authority Wigshaw Lane, Culcheth Warrington, Chesire United Kingdom WA3 4NE Robert E. Henry Fauske and Associates, Inc.

16WO70 West 83rd Street Burr Ridge. IL 60521 Daniel Hirsch University of California A. Stevenson Prog. on Nuclear Policy Santa Cruz, CA 95064 Steven Hodge Oak Ridge National Laboratories P.O. Box Y Oak Ridge, TN 37831 Thomas C. Houghton KMC, Inc.

1747 Pennsylvania Avenue, NW Washington, DC 20006 Dean Houston U.S. Nuclear Regulatory Commission Advisory Comm. on Reactor Safeguards Washington, DC 20555 Der Yu Hsia Taiwan Atomic Energy Council 1 67, Lane 144, Keelung Rd., Sec. 4 1 Taipei, Taiwan l

1 DIST-9 l

R. E. .Jaquith Combustion Engineering, Inc.

1000 Prospect Hill Road M/C 9490-2405 Windsor, CT 06095 Kjell Johannson Studsvik Energiteknik AB S-611 82. Nykoping, Sweden D. H. Johnson Pickard, Lowe & Garrick, Inc.

2260 University Drive Newport Beach, CA 92660 Vojin Joksimovich NUS Corporation 16885 W. Bernardo Drive j San Diego, CA 92127 H. R. Jun Korea Adv. Energy Research Inst.

P.O. Box 7 Daeduk Danju Chungnam 300-31, Korea Peter Kafka Gesellschaft Fur Reaktorsicherheit Forschungsgelande 8046 Garching Federal Republic of Germany I

M. Kahatib-Rahbar Brookhaven National Laboratory i Associated Universities, Inc.

! Upton, NY 11973 a

l Geoffrey D. Kaiser i NUS Corporation j 910 Clopper Road

Gaithersburg, MD 20878 i

W. Kastenburg i

UCLA 4

5532 Boelter Hall Los Angeles, CA 90024

. W. Kato I

Brookhaven National Laboratory l Associated Universities, Inc.

Upton, NY 11973

] ,

l William Kerr

! University of Michigan, North Campus

Department of Nuclear Engineering Ann Arbor, MI 48109 DIST-10

~G2orge Klopp Connonwaalth Edison Company P.O. Box 767 Room 35W Chicago, IL 60690 Klaus Koberlein Gesellschaft Fur Reaktorsicherheit Forschungsgelande 8046 Garching Federal Republic of Germany E. Kohn Atomic Energy Canada Ltd.

Candu Operations, Mississauga Ontario, Canada Herbert J. C. Kouts Brookhaven National Laboratories l Building 179C l Upton, L.I. NY 11973 Thomas Kress j Oak Ridge National Laboratories P.O. Box Y

< Oak Ridge, TN 37831 Jeanne-Marie Lanore

. CEA B.P. No. 6 i 92660 Fonenay-Au-Roses, France

John C. Lee University of Michigan, North Campus, Dept. of Nuclear Engineering Ann Arbor, MI 48109 i

I 1

i 1

I 1

i I

l i

j otsr-11

. . - . . . . - . . . . . -- . ._ _ _ _. . - . = . _

, Harold W. Lewis University of California Physics Department Santa Barbara, CA 93106 j Librarian Atomic Industrial Forum l '7101 Wisconsin Avenue

Bethesda, MD 20814 1

Eng Lin l Taiwan Power Company j 242, Roosevelt Rd., Sec. 3 Taipei, Taiwan J. P. Longworth

Central Electric Gen' Ting Brd.

t Berkeley Gloucester 1 GL13 9PB United Kingdom William J. Luckas l Brookhaven National Laboratories Building 130 Upton, L.I., NY 11973 l

Hans Ludewig 1

Brookhaven National naboratories Building 130 t Upton, L.I., NY 11973 P. MacDonald l EG&G Idaho, Inc.

i P.O. Box 1625 l Idaho Falls. ID 83415 i

i A. P. Malinauskas l Oak Ridge National Laboratories i P.O. Box Y Oak Ridge, TN 37831

]

r i Giuseppe Mancini Commission European Comm.

J CEC-JRC Eraton j Ispra Varese, Italy 1

i Alan McGowan Scientists Institute for Public Info.

! 355 Lexington Avenue t New York, NY 10017 i

! Ralph Meyer U.S. Nuclear Regulatory Commission l Office of Nuclear Regulatory Research Washington, DC 20555 l

i j

i f DIST-12

. ,__... . _ - _ . _ _ _ . ~ .

Josolyn Mitch311 U.S. Nuclocr Regulatory Commission Office of Nuclear Regulatory Research Washington, DC 20555 Joseph A. Murphy U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research Washington, DC 20555 Sue Niemczyk 1545 18th St. NW, #112 Washington, DC 20036 l

Edward P. O'Donnell Ebasco Services. Inc.

2 World Trade Center, 89th Floor New York, NY 10048 David Okrent University of California, Los Angeles Boelter llall, Room 5532 Los Angeles, CA 90024 William Parkinson Delian Corporation 1340 Saratoga-Sunnyvale Road Suite 206 San Jose, CA 95129 G. Petrangeli ENEA Nucl. Energ. ALT Disp Via V Brancati 48 00144 Roma, Italy Robert D. Pollard Union of Concerned Scientists 1616 P Street, NW, Suite 310 Washington, DC 20036 H. Potter UK Atomic Energy Authority Winfrith Dorchester Dorset DT2 8Dil United Kingdom William T. Pratt Brookhaven National Laboratories Building 130 ,

Upton, NY 11973 David Pyatt U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research l Washington, DC 20555  !

l DIST-13

Ingvard Rtenusssn Rico Nationti Ltboratory Postbox 49 DK-4000 Roskilde, Denmark Norman C. Rasmussen Massachusetts Institute of Technology 77 Massachusetts Avenue Cambridge, MA 02139 Dennis Richardson Westinghouse Electric Corporation P. O. Box 355 '

Pittsburg, PA 15230 Robert Ritzman Electric Power Research Institute 3412 Hillview Avenue Palo Alto, CA 94304 Denwood F. Ross U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research Washington, DC 20555 Frank Rowsome U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation Washington, DC 20555 John Siegal Atomic Industrial Forum 7101 Wisconsin Avenue Bethesda, MD 20814 Chester P. Siess Advisory Comm. on Reactor Safeguards 805 Hamilton Drive Champaign, IL 61820 Melvin Silberberg U.S. Nuclear Regulatory Commission Office of Nuclear Regulatory Research Washington, DC 20555 K. Soda Japan Atomic Energy Res. Inst.

Tokai-Mura Naka-Gun IBARAKI-KEN 319-11, Japan Leonard Soffer U.S. Nuclear Regulatory Commission Office of Nuclear Reactor Regulation Washington, DC 20555 DIST-14

Th0aic P. Spois U.S. Nucloor Rcquietory Cotmission Office of Nuclear Reactor Regulation Washington, DC 20555 Klaus B. Stadie OECD-NEA, 38 Bld. Suchet 75016 Paris, France Wayne L. Stiede Commonwealth Edison Company P.O. Box 767 Chicago, IL 60690 John Taylor i Electric Power Research Institute

! 3412 Hillview Avenue Palo Alto, CA 94303 Merrill A. Taylor l

U.S. Nuclear Regulatory Commission j Office of the EDO Washington, DC 20555 Technical Library Electric Power Research Institute P.O. Box 10412 Palo AltoCA 94304 Mark 1. Temme General Electric, Inc.

! P.O. Box 3508 Sunnyvale, CA 94008 T. G. Theofanous

University of California, S.B.

Dept. of Chemical and Nuclear Eng.

Santa Barbara, CA 93106 I Gordon Thompson Inst. for Resource & Security

{

Studies

( 27 Elsworth Avenue

Cambridge, MA 02139 Long Sun Tong 9733 Lookout Place Gaithersburg, MD 20879 l 1

l A. F. Torri Pickard, Lowe & Garrick, Inc.

191 Calle Magdalena, Suite 290 Encinitas, CA 92024 I

DIST-15

.- .-. . ~. . .- - . . _ _ . . _ - _ = _ . - --

Choo-Chin Tung i AtcDic Enorgy Council 67 Lane 144, Sec 4 Keelung Rd.

Taipei, Taiwan Ian B. Wall Electric Power Research Institute

~'

P.O. Box 10412 Palo Alto, CA 94303 David Ward Savannah River Laboratories

. Building 773-A, A-219 Aiken, SC 29808 Edward Warman Stone & Webster Engineering Corp.

P.O. Box 2325 Boston, MA 02107

, Norman Weber l Sargent & Lundy Co.

55 E. Monroe Street Chicago, IL 60603 l

j Lois Webster i

American Nuclear Society 555 N. Kensington Avenue i La Grange Park, IL 60525 J

Paul Wood '

Delian Corporation j 296 Duff Road Monroeville, PA 15146 John Wreathall Battelle Columbus Division 505 King Avenue Columbus, OH 43201 I Roger Wyrick

, Inst. for Nuclear Power Operations 1100 circle 75 Parkway, Culte 1500 ,

! Atlanta, GA 30339 i Jonathan Young '

Energy, Inc.

1851 South Central Place

, Suite 201 i Kent, WA 98031 1

i Steven Sholly i MAB Technical Associates

1723 Hamilton Avenue, Suite K San Jose, CA 95125 i

. I l

DIST-16

l 1

u s auctame.uuvono co== o= i -neo-r =w-eiwe rioc - =..a.*,

geomai ass U'," '8 NUREG/CR-470 i IOGRAPHIC DATA SHEET SAND 86-1135 Vol.2 set etstauCtio45 0% fut styles 3 tit 64 aNosvetiTLt 3484WieLANE CONTAINMENT EVE ANALYSIS FOR POSTULATED SEVERE ACCIDENTS. SEQUOYAH POWER STATION, e e yate aeront cowe6etoo gg y uomfs. vaan l

. , . . , , /

V. L. Behr, A. S. Be jamin, D. M. Kunsman, /"o*'" o ca'<ai oa m swso S. R. Lewis, W. B. M fin / l i february 1987 7 WEwrCaM'NG ORGANila160N hawa shQ wastsNQ ADO S$ natwa te ceses eP JJ4Cf.Taan Woomn whid Nwwgen i f M '* '"'"''**"

Sandia Nat.onal Laborato es l Albuquerque., NM 87185 A1322 4 10 trONSomeNG OmGamedafiom %aws agg Wa LigQ AQOnllt ufarew e cores t1 f.'EO'RtPQAf Division of Reactor System Sifety

Office of Nuclear Regulatory escarch DRAFT FOR COMMENT

' " " ' " ' " " ' " ~ " * "

l U.S. Nuclear Regulatory Commis ion Washington, DC 20555 1 , , .v,, . .. . , a . w e . .

s I

ta assimact aoo.co.a wa, y l A study has been perfetmed as part of the Severe Accident  !

! Risk Reduction Program f$ASFP) to investigate the response of a t

! particular pressurised teatetg reactor with an ice. condenser containment (Sequoyah Antt Q to postulated severe accidents.

A detailed containeerst event gree tot the Sequoyah plant has j been devised to desettbe the Verious possible accident pathways

( that can is,ad to radioactive releases from containment. Data

and analyses from a large numberkof NRC and industry-sponsored 1 j- programs have beery' reviewed and deed as a basis tot quantif ying I the event tree. tie., determinin 6the likelihood of each pathway for a vatiety of acendent kequence initiators. A generalised conpatnment event tree todo, called EvwTat, has been developed to f acilitate the quahtification. The uneettaintytr/theresultshasbeenesaminedbyperformingthe quantificattop three times. using a digterent set of input each

time to reptysent the variation of opinion in the reactor

+ safetyconnipity, in the so called *centrata estimate, the 1

Intelihood at early contalheent failure (Securring befute et at l the timeb of station teactor henout vessel breach) sequences but very was lowtehnd to be tot other h19h tot accident sequence nitiators. Unavailablitty of ignitors and air return fans was the peinetpal reason for the high tbiture probability j for sta on blaeneute. The analysis also shoged that molting or bype a of the ice betete er within a short %ime after vesset breach can be espected to occur with moderate te high i likel ood during station blackouts and during sequences innti ted by very small LOCAS with failure of en*(gency cote cool g in the rectreulation phase aftet success to the

inle tion phase. t et severe This work supports NBC's assessmag\

j acefdent tiots to be published in t#U8t0 1150.

. . oow. 4, a . s , .. . . . . . . m v.o. . . .c . .. n . ,.a.a..4..

j statewe%f i

is secwa itec6assisication o re., eneer e ion %ts. ..... o.im e =os o "**

Un.clasnifind ar . ,

IInc

,, %.ei i a.n o eaas i f i ed L

,4,63 e u a novin8vs47 m=nwo orrece e.or-trao.s ecoss I

120555078877 g gag US NRC-DARM-Ang O!V 0F pud SVCS hh0 Y $ PUB MGT BR-PDR NUREG WASHINGTON DC 20555

.[ 3 7 t

I l

l l

. _ . _ . _ _ _ . _ _ _ _ . _ _ . _ . _ . . _ . . _ _ _ . _ _ . _ ._. __ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ _ _ . . _ _ . _ _ . _ _ . _ _ _ _ _ _ . _ _ _ _ _ _ _ _ _ _