ML20128K206

Technical Evaluation Rept of Pilgrim IPE Back-End Submittal Final Rept

ML20128K206
Person / Time
Site:	Pilgrim
Issue date:	05/31/1996
From:	ENERGY RESEARCH, INC.
To:	NRC
Shared Package
ML20128K194	List: ML20128K187 ML20128K206 ML20128K213 ML20129H377 ML20129H390
References
ERI-NRC-96-102, NUDOCS 9610100300
Download: ML20128K206 (42)

Text

. ,

b i

APPENDIX B PILGRIM NUCLEAR POWER STATION INDIVIDUAL PLANT EXAMINATION TECHNICAL EVALUATION REPORT (BACK-END) ,

l l

I I

l 4R totolo O 303 % r p.

1, j ERl/NRC E102 i

1 l

TECHNICAL EVALUATION REPORT OF THE i

t PILGRIM INDIVIDUAL PLANT EXAMINATION 1

l BACK-END SUBMITTAL I

i -

1 t

Final Report l

i f

1 i

l May 1996 i

4 l

l f Energy Research, Inc.

P.O. Box 2034 Rockville, Maryland 20847 l

dI j Prepared for.

SCIENTECH,Inc.

) Rochvme, Marytend j

i J Under Cortract NRC.04-916 Wtn the unhod States Nucteer Regentory Commenelon Weehrgton.D.C 20555 j

4 4

s .

ERI/NRC 96-102 i

i l

l TECHNICAL EVALUATION REPORT OF THE PILGRIM INDIVIDUAL PLANT EXAMINATION BACK-END SUBMITTAL l

l FINAL REPORT l I

May 1996 i

i R. VLjaykumar, A. S. Kuritzky, and M. Khatib-Rshbar Energy Research, Inc.

P. O. Box 2034 Rockville, Maryland 20847-2034 i

Prepared for:

SCENIECH, Inc.

Rockvme, Maryland 20ss2 Under Contract NRC 04-91068 l With the U.S. Nuclear Regulatory N=h l Wasideston, D. C. 20555 l

I

. E. EXECUTIVE

SUMMARY

l This Technical Evaluation Report (TER) documents the findings from a review et 'he back-end l portion of Pilgrim Nuclear Power Station Individual Plant Examination (IPE) Back 'ad submittal of the Boston Edison Company (BECo). The primary intent of the review is to ascenain whether or not, and to what extent, the back-end IPE submittal satisfies the major intent of ,

l 4

Generic letter (GL) 88-20 and achieves the four IPE sub-objectives. The review utilized both,  ;

the information provided in the IPE submittal, and additional information provided by the j licensee in response to NRC questions.

The back-end portion of the IPE submittal supplies a substantial amount of information with f

i regards to the subject areas identified in Generic letter 88-20, and NUREG-1335.

i

!. E.1 Plant Characterization j The Pilgrim plant is a General Electric Company BWR/3 plant with a Mark I containment located at Plymouth, Massachusetts. The rated thermal power is 1998 MWt (687 MWe). The l

mean containment failure pressure is 98 psig.

j j E.2 Ucensee's IPE Process 4

I

! The IPE was a cooperative utility-contractor effort, with most of the work being performed by j

BECo staff. Tenera, Fauske and Associates, and Gabor, Kenton, and Associates were the contractors. Five BECo senior engineers were involved in the IPE. It appears that BECo staff i

! were involved in many, if not all, aspects of the study. However, since no personnel breakdown l

by task is provided, it is impossible to ascertain the actual level of utility involvement. Itis stated in Section 2.1.2.1 of the submittal, that BECo staff were involved in all areas of the IPE l

(both level 1 and level 2), and that a complete transfer of technology was accomplished. As stated in Section 2.1.2.2 of the submittal, the Pilgrim IPE underwent four levels of review. The l first level of review involved a review of the IPE assumptions and results, principally performed by the BECo PRA/IPE staff, with limited support from consultants. The second level of review involved a review of other similar industry studies and plant-specific information from the Peach l

Bottom /NUREG-1150 study. An internal peer review provided the third level of review for the Pilgrim IPE. As stated in the submittal, "[t]his independent, in-house review was conducted to l

ensure the accuracy of the documentation contained in the report, and to validate the IPE process and results." The final level of review was performed by an external peer review team, i Specifically in regards to the back-end analysis, it is stated in the submittal that an independent

}

review was performed by the primary back-end contractor. While Section 2.1.2.2 of the i submittal lists a number of important comments / insights that were obtained from the review process, however, these comments do not pertain to the back-end portion of the submittal.

l

! The methodology employed in the Pilgrim submittal for the back-end evaluation is clearly

described, and the IPE is logical, traceable, and consistent with GL 88-20. The definition of j Plant Damage States (PDSs) involved using the core damage sequences as input to Containment j

Pilgrim IPE Back-End Review ij ERI/NRC 96-102 l

1

l 1

l Safeguards Event Trees (CSETs). The CSET nodes represent those critical safety functions which describe the post core damage status of systems important to accider progression, l

j containment response, and radiological release. Probtbilistic quantification of se ere accident progression involved development of a relatively small Containment Phenomer. Event Tree

! (CPET), and use of supporting fault trees to tillor each CPET question to the specific PDS being evaluated. The results of the CPET analyses lead to an extensive number of end-states, which l

j were in turn binned into a manageable number of release categories, based on similarities in i accident progression and source term characteristics, t

i The front-end analyses in the IPE submittal report a Core Damage Frequency (CDF) of 5.8 x

! 10-3 per reactor year. The dominant contributors to core damage are high pressure core damage

~

sequences (initiated by loss of offsite power or by transients such as loss of feedwater) with loss j of coolant makeup (79.3%), followed by Anticipated Transient Without Scram (ATWS) sequences (7 %), low pressure core damage sequences with loss of coolant makeup (6.2 %), and

LOCAs (3.2%).

After the IPE was submitted to the NRC, the licensee revised the models fer COF, including l a number of enhancements. Specific examples include: elimination of the depenocncy of HPCI

! on room cooling, reduction in the number of Safety Relief Valves (SRVs) required for the I success of the ADS, modification of DC power success criteria, development of recovery actions for SBO, common cause breaker failures, etc. In addition, revisions of the initial estimates for Loss of Offsite Power (LOOP) were also made in the IPE submittal. The net impact of these changes is the reduction of the CDF from 5.85 x 10'5 per reactor year to 2.84 x 105 In addition, the relative contributions of the various PDSs to the CDF, were also found to be different than in the original submittal. However, the containment analyses were not revised.

The interface between the front-end (level 1) analysis and the back-end (level 2) analysis is accomplished through the propagation of front-end (core damage) sequences through the Containment System Event Trees (CSETs). The CSETs are used to define the status of systems l l

which are important for analyzing containment response to accident challenges. Since the actual cutsets from the level 1 analysis are input to the CSETs, system dependencies are accounted for l between the level 1 systems and the containment systems. The core damage /CSET sequences '

l are grouped together into Plant Damage States (PDSs), based on functional characteristics important to accident progression, containment failme and source term definition. Only 12 of the PDSs actually comain cutsets. l 1

Probabilistic quantification of severe accident progression for the probabilistically significant PDSs was performed using a Containment Phenomena Event Tree (CPET). The methodology employed in the Pilgrim submittal invcived development of a relatively small CPET, and use of small, supporting fault trees to tailor each CPET question to the specific PDS being evaluated.

The Pilgrim CPET contains the following nine nodes:

e Debris Cooled In-Vessel e Small Lower RPV Head Failure Pilgrim IPE Back-End Review iii ERI/NRC %I02

.

• C j

i  !

l l 4

• No Pedestal Failure at RPV Failure j e No Drywell Failure at RPV Failure e Debris Cooled Ex-Vessel e No Drywell Over-Temperature Failure j e Containment Not Challenged By Over-Pressurization i .* Wetwell Vapor Space Failure or Venting i e No Pool Bypass i A phenomenological fault tree was developed to suppor a m quantification of each CPET heading

' listed above. The quantification of the fault tree basic ed.nts was obtained through one or more i of the following sources:

4

}

- 1. Pilgrim-specific MAAP calculations,

2. Peach Bottom or Grand Gulf /NUREG-1150 supporting documentation [3,4], or l

i

3. Pilgrim-specific " hand calculations."

j The Pilgrim CPET includes most of the relevant phenomena for BWR's U Mark I containments; nevertheless, as discussed in this TER, the quantification of theirampact on the

! Pilgrim containment is weak.

j The results of the CPET analyses lead to an extensive number of end-states, which are classified into a manageable number of release categories, categorized by similarities in accident progression and source term characteristics. The sequence characteristics which were identified

) in the IPE submittal as having the greatest impact on fission produc.t release at Pilgrim are the 4

l following:

I j - Containment Bypass

{

- Debris Cooled In-Vessel

- Time of Containment Failure / Venting (Relative to Core Damage) i  !

1

- Mode /Imation of Containment Failure (or Venting)

- Suppression Pool Bypass

- Type of Ex-Vessel Core / Concrete Interactions (Dry, Wet or None)

Using these characteristics as headings, a Source Term Category Grouping Diagram was l

i developed. The end points of this diagram represent the individual release categories. A total

! of 34 release categories are defined by the submittal, of which only 20 are myensi as non-zero

! (i.e., frequency > 10*/ry). MAAP calculations were performed to determine the source terms i

i for representative sequences of 12 of the 20 non-zero release categories. Of the remaining eight release categories, the source terms for seven were determined to be similar to one of the l previously calculated source terms. 'Ihe final release category involved no containment failure '

j or venting, and was assigned no source term. A summary of the releases associated with each non-zero release category is provided in Table 4.7-5 of the submittal. The source terms reported i

i in this table do not take credit for any fission product retention in the reactor building.

i i ERI/NRC %-102

} Pilgrim IPE Back-End Review iv 1

i 1

O I e

E.3 Back-End Analysis l

' The conditional probabilities of early and late containment failure calculated by th.: tubmittal are O.216 and 0.61, respectively (see Table E.1). The conditional probability of intacc containment

is about 0.17.

From review of Table E.1, it is seen that the major difference between Pilgrim and the Peach j Bottom plant (NUREG-1150 analyses) in the table, is that Pilgrim has a lower conditional j probability of early containment failure, and accordingly, a higher probability of late containment failure. It appears that Pilgrim has understated the contribution fror.1 carly drywell i

j failure due to overpressurization (see Section 2.1.3.3 of this review). Also, for BWRs with Mark I containments, the dominant cause of early containment failure is drywell liner melt t through. The submittal (p. 4.7-10) states that the principal reason for the difference in early

containment failure is due to the Pilgrim assumption that for sequences with high vessel pressures at RPV failure, the probability of a liner melt through in a dry cavity is only 0.1, if l

) in-vessel injection is available following RPV failure. Also, as discussed in Sxtion 2.1.2.2 of l this review, the contribution of drywell liner melt through reported in the subnuttalvis somewhat

, lower due to the incorporation of the results from a more recent study. However, since the

submittal does not provide the fraction of core damage frequency that is associated with a dry

! cavity at the time of RPV failure, as well as other event probabilities ===-iami with the

probability of liner melt through, it is not possible to determine the exact reasons why Pilgrim j j exhibits a much lower probability of early containment failure. l l The relatively high probability of late containment failure reported in the submittal may be partially the result of the Pilgrim containment flooding strategy, which directs the opemtors to i flood the containment if RPV water level cannot be maintained above the top of active fuel for non-ATWS sequences, or above two-thirds core height for A'IWS sequences. Once containment i

i Table E.1 Containment Failure as a Percentage of Internal Events CDF: Comparison of I Pilgrim IPE Results to Peach Bottom NUREG-1150 Results Containment Failure Peach Bottom NUREG-1150 Pilgrim IPE

! CDF (per year) 4.3 x 104 2.84 x 10-5 j Early Failure 46 21.6 i

Bypass - 0.4 Iate Failure 26 61.0

\

Intact 3 1.2 i

1 Intact, No Vessel 25 15.8 i Breach i

i Pilgrim IPE Back-End Review y ERI/NRC 96-102 1

?

1 water level reaches the bottom of the recirculation lines, the operators are instructed to initiate RPV venting (to the condenser), which is classified as a late containment failure It should be noted, that as stated in Section 4.8.2.1.11 of the submittal, the licensee c.. considering l

alternatives to the current procedure for containment flooding /RPV venting, si- '.c sensitivity l

1 analyses in the IPE submittal have shown that the current procedure has a negative impact on

! containment performance. .

In spite of these comments, the licensee's process for the evaluation of containment failure probabilities and failure modes is consistent with the intent of Generic I.etter 88-20, Appendix i I. The licensee has considered the failure of the containment isolation system and containment

, bypass scenarios. A number of sensitivity analyses have also been performed. All of the j phenomena of relevance to BWR severe accident phenomenology have been included in the submittal, as well as the principal phenomenological uncertainties. In Section 4.8.2 of the l

! submittal, it is stated that based on recommendations in NUREG-1335 and the EPRI " Guidance Document" for using MAAP [10), the following uncertainty issues wett investigated through

! performance of sensitivity analyses with the MAAP code:

! - Core Melt Progression /In-Vessel Hydrogen Generation l

- Amount of Core Debris Retained in RPV RPV Pressure at Vessel Failure Containment Pressure Load due to RPV Failure f - Direct Containment Heating j - Shell Failure by Liner Melt Through

- Debris Spread in Containment

- Ex-Vessel Debris Coolability l

Containment Failure Location i - Containment Failure Area

- Containment Flooding 4

,i

- Saturated Pool Decontamination Factor

A brief description of each sensitivity case is provided in Table 4.8-3 of the submittal, and a summary of insights and conclusions from these analyses is presented in Table 4.8-15. In i addition, a number of uncertainties in the Pilgrim containment performance were treated i

indirectly through sensitivity analyses of several CPET event probabilities. As stated in the submittal, sensitivity analyses were gdww.ed for those parameters which were judged to have large uncertainties or were expected to significantly influence the final results. The sensitivity

! analyses performed include the following:

I - Probability of In-Vessel Cooling i

- Probability of I.arge RPV Failure Due to Lower Head Thermal Attack j - Probability of I.arge RPV Failure Due to In-Vessel Steam Explosion

- Probability of Pedestal Failure Due to Overpressurization at RPV Failure

{

- Impact of Water in Drywell at RPV Failure

- Probability of Liner Thermal Failure at RPV Failure Pilgrim IPE Dack-End Review vi ERI/NRC 96-102 i

3 4

1 1

1

! i i )

,1 i

- Probability of Core / Concrete Interactions Fraction of PDS Sequences with RPV Depressurized and With In-Vnsel Injection l Available i - Containment Flooding and Venting i - DCH Drywell Overpressure Failure i

. The principal insights from these sensitivities are provided in Section 4.8.1 of the submittal, and summarized in Section 4.8.3.1 of the submittal.

1 E.4 Cont =Innent Performance Improvements Generic letter 88-20, Supplement Numbers 1 and 3 identified specific Containment Performance j improvements (CPIs) to reduce the vulnerability of containments to severe accident challenges.

For BWRs with Mark I containments, the following improvements were identified:

- Alternative water supply for drywell spray / vessel injection, ,

l - Enhn=d reactor pressure vessel depressurization system reliability, l

- Implementation of Revision 4 of the BWR Owners Group EPGs, and l .

- Installation of a hardened vent.

]

At the time of issuance of the Generic letter, the licensee had already implemented the l

j following procedural and hardware modifications that are consistent with the recommendations i of the CPI program:

I

e Installation of a hardened vent path.

i j

e Modification of the existing plant systems to provide an alternate source of water injection into the vessel through the fire water cross-tie.

i e Installation of a third diesel generator to enhance the reliability of AC power.

l

• Implementation of Revision 4 of the BWROG EPGs. l 1

j e Installation of a backup nitrogen supply system to provide long term pneumatic control capability to the Automatic Depressurization System (ADS).

In summary, the licensee's modifications are consistent with the CPI recommendations.

Pilgrim IPE Back-End Review vii ERI/NRC %-102

i i E.5 Vulnerabilities and Plant Improvements l

Section 5.1 of the submittal provides the following criteria, used to detanine if any i vulnerabilities exist at the plant:

i j a) Are there any new or unusual means by which core damage or containment 3

failure occur as compared to those identified in other PRA's?

b) Do the results suggest that the Pilgrim core damage frequency would not be able

to meet the NRC's safety goal for core damage?

l It is stated in the submittal that, based on the above criteria, no potential vulnerabilities were l

identified for the Pilgrim nuclear Station.

l E.6 Observations i

! The back-end portion of the Pilgrim IPE submittal provides a substantial amount e.finformation in regard to the subject areas identified in Generic I.etter 88-20 and NUREG-1335. The PSA methodology used for the backend analysis is basically sound, capable of identifying plant-specific vulnerabilities to release of radioactivity to the environment, and includes all the relevant phenomenological issues. The quantification of accident progression is based in large part on numerical estimates provided in NUREG-1150 supporting documentation; however, the submittal does not always contain adequate documentation as to the applicability of these estimates to the Pilgrim plant.

The important points of the technical evaluation of the Pilgrim IPE back-end analysis are summanzed as follows:

e The back-end portion of this IPE submittal, for the most part, is relatively well performed and well written.

• The number of non-zero PDSs was limited to 12, and the impact of each PDS on containment performance was specifically analyzed with a CPET. ,

o The submittal includes all phenomena of relevance to severe accident progression for BWRs with Mark I containments.

e The submittal considers the impact of severe accident conditions on the operability of equipment.

• All CPI recommendations have been addressed, either in the submittal, or previously as part of the licensee's Safety Enhancement Program.

Pilgrim IPE Back-End Review viii ERI/NRC 96-102

e Several minor weaknesses (with regards to their overall impact on the IPE results) exist, and they include the following:

- The PDS definitions in the submittal are not very transparent, due to the omission of some key accident characteristics (e.g., initiating event type, or status of AC electrical power).

- In a number of instances, the submittal makes use of parameter or probability values from the Grand Gulf or Peach Bottom NUREG-1150, but does not provide a basis for their applicability to the Pilgrim plant.

- The treatment of early drywell failure due to overpressurization is inadequate.

In spits of these identified weaknesses, the licensee's process for the evaluation of containment failure probabilities and failure modes is cor.sistent with the intent of Generic letter 88-20, Appendix 1. The dominant contributors to containment failure are consistent with the insights obtained from the NUREG-1150 analyses for the Peach Bottom plant. In swamary, it is concluded that the IPE submittal provides a substantial amount of information id regard to the subject areas identified in Generic letter 88-20 and NUREG-1335.

l l

Pilgrim IPE Back-End Review ix ERI/NRC 96-102 ,

I .. .

l; 1 TABLE OF CONTENTS 2

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

...... I

- 1.1 Review Process .............................

1.2 Containment Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . I l 2. CONTRACTOR REVIEW FINDINGS . . . . . . . . . . . . . . . . . . . . . . . . . . 3

2.1 Review and Identification of IPE Insights . . . . . . . . . . . . . . . . . . . . . 3 j 2.1.1 Completeness and Methodology . . . . . . . . . . . . . . . . . . . . . . 3

' 2.1.2 As-Built /As-Operated Status . . . . . . . . . . . . . . . . . . . . . . . . 3 2.1.3 Licensee Participation and Peer Review of IPE . . . . . ....... 3 2.2 Containment Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4 l

2.2.1 Front End/Back End Dependencies . . . . . . . . . . . . . . . . . . . . 4 2.2.2 Containment Event Tree Development .................. 6 2.2.3 Containment Failure Modes and Timing . . . . . . . . . . . . . . . . . 9~

l 2.2.4 Containment Isolation Failure ...............~......10 2.2.5 System / Human Response . . . . . . . . . . . . . . . . . . . . . . . . . 10

l. 2.2.6 Radionuclide Release Categories and Characterization . . : . . . . . 11 2.3 Quantitative Assessment of Accident Progression and Containment i 13 Behavior . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

2.3.1 Severe Accident Progression . . . . . . . . . . . . . . . . . . . . . . . 13 j

14

, 2.3.2 Dominant Contributors to Containment Failure . . . . . . . . . . . .

2.3.3 Characterization of Containment Performance . . . . . . . . . . . . 16 j 2.3.4 Impact on Equipment Behavior .....................18 j 2.4 Reducing the Probability of Core Damage or Fission Product Release . . 18 2.4.1 Definition of Vulnerability . . . . . . . . . . . . . . . . . . . . . . . . 18 18

2.4.2 Plant Modifications . . . . . . . . . . . . . . . . . . . . . . . . . . . . .

! 2.5 Responses to the Recommendations of the CPI Program . . . . . . . . . . 19 21

3. OVERALL EVALUATION AND CONCLUSIONS . . . . . . . . . . . . . . . . .

f 4. REFERENCES . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23 i

APPENDIX . . . . . . . . . . . . . . . . . . . ...........................25 Pilgrim IPE Back-End Review x ER1/NRC 96-102

1. . .

4 i

LIST OF TABLES Table 1 Radionuclide Release as a Pcreentage of Internal Events CDF . . . . . . . . . 12

' l Table 2 Containment Failure as a Percentage of Internal Events CDF: Comparison with Other 15 PRA S tudies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . l

'l l l 4 )

)

4 1

1 1 4 l

! I 4

l I

I

)

t I

i e

Pilgrim IPE Back-End Review xi ERUNRC %102 4

l l

1 NOMENCLATURE l

} AC Alternating Current j ADS Automatic Depressurization System

) ATWS Anticipated Transient Without Scram BECo Boston Edison Company

BWR Boiling Water Reactor BWROG Boiling Water Reactor Owners Group CCI Core Concrete Interactions CDF Core Damage Frequency CRD Control Rod Drive 1 CET Containment Event Tree

! CHR Containment Heat Rejection CPI Containment Performance Improvement DC Direct Current i DCH Direct Containment Heating ,

DG Diesel Generator i ECCS Emergency Core Cooling Systems

- EOP Emergency Operating Procedure

EPRI Electric Power Rescasch Institute EVSE Ex-Vessel Steam Explosion FPS Fire Protection System GE General Electric GL Generic Letter HPME High Pressure Melt Ejection HRA Human Reliability Analysis IPE Individual Plant Examination ISLOCA Interfacing Systems I. mss of Coolant Accident LOCA Loss of Coolant Accident LOSP Loss of Offsite Power LPCI Low Pressure Coolant Injection LPCS 14w Pressure Core Spray LT-SBO Iong Term Station Blackout MAAP Modular Accident Analysis Program MOV Motor Operated Valve MSIV Main Steam Isolation Valve NRC Nuclear Regulatory Commission PDS Plant Damage State PRA Probabilistic Risk Assessment RCS Reactor Coolant System RHR Residual Heat Rejection RPV Reactor Pressure Vessel SBO Station Black-Out Pilgrim IPE Back-End Review xii ERI/NRC %102

._ - = _ - __ . .

)

1

)

NOMENCLATURE (Continued)

SORV Stuck-Open Relief Valve SRV Safety Relief Valve ,

SSW Station Service Water TER Technical Evaluation Report TW Loss of Decay Heat Removal

USI Unresolved Safety Issue t

1 1

i l

l 4

4 I

l i

1 i

Q i

i i

1 4

l 4

Pilgrim IPE Back-End Review xiii ERI/NRC 96-102 1

1. INTRODUCTION This Technical Evaluation Report (TER) documents the results of a review :' the Pilgrim Nuclear Power Station Individual Plant Examination (IPE) Back-End submittal [1;. based on the following review objectives set forth by the NRC:

e To determine if the IPE submittal essentially provides the level of detail requested in the " Submittal Guidance Document," NUREG-1335, o To assess the strengths and weaknesses of the IPE submittal, e To provide a preliminary list of questions bssed on this limited review, and e To complete the IPE Evaluation Data Summary Sheet.

The remainder of Section 1 of this report describes the technical evaluation rrocess employed  ;

in this review, and presents a summary of the important characteristics of the P? grim nuclear plant related to containment behavior and post-core-damage severe accident progression, as derived from the IPE. Section 2 :vmmarues the review technical findings, and briefly describes the submittal scope as it pertains to the work requirements. Each portion of Section 2 corresponds to a spxific work requirement as outlined in the NRC contractor task order. A summary of the overall IPE evaluation and review conclusions are summanzed in Section 3.

Section 4 contains a list of cited references. Appendix A to this report contains the IPE evaluation data summary sheets.

1.1 Review Process The technical review process for back-end analysis consists of a complete examination of Sections 1, 2, 4 through 6, and Appendices C through G of the IPE submittal. In this examination, key findings are noted; inputs, methods, and results are reviewed; and any issues or concerns pertaining to the submittal are identified. The primary intent of the review is to ascertain whether or not, and to what extent, the back-end IPE submittal satisfies the major intent of Generic Letter (GL) 88-20 [3] and achieves the four IPE sub-objectives. A draft TER based on the back-end portion of the submittal was submitted to the NRC in December 1994.

A list of questions and requests for additional information was developed to help resolve issues and concerns noted in the examination process, and was forwarded to the licensee. The final TER is based on the information contained in the IPE submittal [1], and the licensee responses to the NRC Requests for Additional Information (RAls) [10].

1.2 Containment Analysis The Pilgrim plant is a General Electric Company BWR/3 plant with a Mark I containment located at Plymouth, Massachusetts. The design features of the Mark I containment are described in Section 4.1.1 of the submittal. The drywell, a steel pressure vessel enclosed in Pilgrim IPE Back-End Review 1 ERI/NRC 96-102

l i

reinforced basaltic concrete, is shown in Figures 4.1-2 and 4.1-3. As stated in the submittal, the drywell has a removable head which is held in place by bolts, and is sealed with a double j gasket. The drywell internal design pressure is 56 psig (at a temperature of 28.'*F).

The suppression chamber is a torus-shaped steel pressure vessel, and is supported by the concrete foundation slab of the reactor building. Eight vent pipes connect the drywell to the suppression chamber vent header and its downcomer pipes, which discharge well below the water level in the toms.

9 The various containment systems considered in the Containment System E*,ent Trees are described in Section 4.1.2 of the submittal.

1 I

I l

i e

i l i

i Pilgrim IPE Back-End Review 2 ERUNRC %-102

' i i

q 4 2. CONTRACTOR REVIEW FINDINGS i

The present review compared the Pilgrim IPE submittal to the requirements of t'eneric Letter

! (GL) 88-20, according to guidance provided in NUREG-1335. The findings f the present

! review are reported in this section. The review findings reported in Section 2.1 follow the i structure of Task Order Subtask 1.

I

! 2.1 Review and Identification of IPE Insights j 2.1.1 Comoleten*== and Methodolorv

. The IPE submittal contains a substantial amount of information in accordance with the

recommendations of GL 88-20 and NUREG-1335.

t The methodology employed in the Pilgrim submittal for the back-end evaluation is clearly described, and the IPE is logical, traceable, and consistent with GL 88-20. The definition of

! Plant Damage States (PDSs) involved using the core damage sequences as input tc. Containment j Safeguards Event Trees (CSErs). The CSET nodes represent those critical safety functions a which describe the post core damage status of systems important to accident progression, containment response, and radiological release. Probabilistic quantification of severe accident progression invohed development of a relatively small Containment Phenomena Event Tree

! (CPET), and use of small, supporting fault trees to tailor each CPET question to the specific i PDS being evaluated. The results of the CPET analyses lead to an extensive number of end-i states, which were in turn binned into a manageable number of release categories, based on similarities in accident progression and source term characteristics.

l I

j 2.1.2 As-Built /As-Onerstad Stanie t

i i A description of the plant walkdown process undertaken at Pilgrim is provided on pages 2.1-6 and 2.1-7 of the submittal. Two stages of walkdowns are discussed. The first set of walkdowns )

was performed by a group of consultants as part of the limited scope Individual Plant Evaluation  !

l

! Methodology (IPEM) study, performed for Pilgrim prior to issuance of the generic letter.

j According to the submittal, the second set of walkdowns was performed by two members of the PRA team, both of which have substantial operations experience, and included the inside of the primary containment, in order to check physical parameters required for the MAAP computer

code modeling.

2.1.3 Liceneae Particination and Peer Review of IPE i

! The 1PE was a cc0pestive utility-contractor effort, with most of the work being performed by BECo staff. Tenera, Fauske and Associates, and Gabor, Kenton, and Associates were the

- contractors. Five BECo senior engineers were involved in the IPE. A list of the BECo staff which participated in the performance of the Pilgrim IPE is presented in Table 2.1-3 of the

submittal. It appears that BECo staff were involved in many, if not all, aspects of the study.

Pilgrim IPE Back-End Review 3 ERI/NRC 96-102

J i

! However, since no personnel breakdown by task is provided, it is impossible to ascertain the 1 actual level of utility involvement. It is stated in Section 2.1.2.1 of the submi ral, that BECo staff were involved in all areas of the IPE (both level 1 and level 2), and that a coicalete transfer of technology was accomplished. As stated in Section 2.1.2.2 of the submittal, tiv Pilgrim IPE l 1 underwent four levels of review. The first level of review involved a review of the IPE assumptions and results, principally performed by the Boston Edison Company (BECo) PRA/IPE staff, with limited support from consultants. The second level of review involved a review of other similar industry studies, principally insights from the limited scope IPEM, and plant-specific information from the Peach Bottom /NUREG-1150 study [2]. An internal peer review l provided the third level of review for the Pilgrim IPE. As stated in the su5mittal, "[t]his j independent, in-house review was conducted to ensure the accuracy of the documentation l

contained in the report, and to validate the IPE process and results." The final level of review

was performed by an external peer review team, which was comprised of " industry experts who l

reviewed the IPE for completeness and correctness of the methods used. They also contrasted the Pilgrim results with other PRA's they were familiar with to assure consistency."

l '

Specifically in regards to the back-end analysis, it is stated in the submittal that an independent review was performed by the primary back-end contractor. However, while Section 2.1.2.2 of l

the submittal lists a number of important comments / insights that were obtained from the review process, however, these comments do not pertain to the back-end portion of the submittal.

2.2 Containment Analysis i

This section provides a review of PDS binning, CET analyses, release category definitions, severe accident analyses, and the containment structural analyses in the submittal.

2.2.1 Front End/hk Fad Danendencies The fmnt-end analyses in the IPE submittal report a Core Damage Frequency (CDF) of 5.85 x 105 per reactor year. The dominant contributors to core damage are high pressure core damage sequences (initiated by loss of offsite power or by transients such as lost of feedwater) with loss of coolant makeup (79.3%), followed by Anticipated Transient V'ithout Scram (ATWS) sequences (7%), low pressure core damage sequences with loss of coolant makeup (6.2 %), and LOCAs (3.2 %).

After the IPE was submitted to the NRC, the licensee revised the models for CDF including a number of enhancements. Specific examples include: elimination of the dependency of HPCI on room cooling, reduction in the number of Safety Relief Valves (SRVs) required for the success of the ADS, modification of DC power success criteria, development of recovery actions for SBO, common cause breaker failures, etc. In addition, revisions of the initial estimates for loss of Offsite Power (LOOP) were also made in the IPE submittal. The net impact of these changes is the reduction of the CDF from 5.85 x 104 per reactor year to 2.84 x 10~5 In addition, the relative contributions of the various PDSs to the CDF, were also found to be different than in the original submittal. However, the containment analyses were not revised.

Pilgrim IPE Back-End Review 4 ERI/NRC 96-102

_ _ _ _ _ _ _ _ _ _ _ _ _ ._ _ _ _ = ~ _ _ _ _ _ _ _ _ _ . _

. t 3

w The interface between the front-end (level 1) analysis and the back-end (level 2) analysis is accomplished through the propagation of front-end (core damage) sequencea through the

! Containment System Event Trees (CSETs), as described in Section 4.3 of the .umittal. As stated in the submittal, the CSETs are used to define the status of systems which n important l

i for analyzing containment response to accident challenges. Since the actual cutsets from the i j

level 1 analysis are input to the CSETs, system dependencies are accounted for between the level I systems and the containment systems. The core damage /CSET sequences are grouped together into Plant Damage States (PDSs), based on functional characteristics important to accident

progression, containment failure and source term definition. As in the submittal (page 4.3-7),

j these characteristics are listed below by functional category.

o CONTAINMENT STATUS PRIOR TO CORE DAMAGE

! Containment B ?passed i Containment Failed /Not Isolated i e INITIATING EVENT TYPE .

j ATWS 4

Non-ATWS l e CONTAINMENT MITIGATION FEATURES l Containment Heat Removal i Drywell Spray i Containment Vent Vapor Suppression

! e REACTOR PRESSURE VESSEL STATUS l In-Vessel Injection Recovered During Core Damage

! Injection Available Following Vessel Failure

- RPV Pressure (High or Iow)

In the submittal, FigmL 4.3-3 presents the Plant Damage State Grouping Diagram. This i diagram defines 63 different plant damage states based on the characteristics listed above.

However, only 12 of the PDSs actually contain cutsets. Each PDS which contributes greater than one percent of the total core damage frequency is described in Section 4.3.3 of the

! submittal.

The PDS binning process appears to be reasonable and relatively complete, and includes most indicators of interest to the back-end analysis. However, the PDS definitions are deficient in two regards. First, for initiating event type, the PDS definitions only discriminate between ATWS and non-ATWS initiators. Secondly, the PDSs do not define the availability, or potential 1

Pilgrim IPE Back-End Review 5 ERI/NRC 96-102 l

4 J

k

availability, of plant AC and DC electrical power systems. Without the specific initiating event type and status of AC and DC power systems included in the PDS definitions, E is difficult to determine what operator recovery actions are possible for preventing RPV o containment i

failure, or for mitigating effects of a radiological release.

J 2.2.2 Containment Event Tree Develonment Probabilistic quantification of severe accident progression for the probabilistically significant PDSs was performed using a Containment Phenomena Event Tree (CPET). The methodology employed in the Pilgrim submittal involved development of a relatively small CPET, and use of small, supporting fault trees to tailor each CPET question to the specific PDS being evaluated.

The Pilgrim CPET contains the following nine nodes:

e Debris Cooled In-Vessel e Small Lower RPV Head Failure l

e No Pedestal Failure at RPV Failure j e No Drywell Failure at RPV Failure j e Debris Cooled Ex-Vessel e No Drywell Over-Temperature Failure e Containment Not Challenged By Over-Pressurization e Wetwell Vapor Space Failure or Venting e No Pool Bypass A phenomenological fault tree was developed to support tr.E quantification of each CPET heading listed above. A description of each phenomenological fault tree, and its quantification, is provided in Section 4.5.2 of the submittal. As stated in the submittal, the quantification of the fault tree basic events was obtained through one or more of the following sources:

1. Pil;; rim-specific MAAP calculations,

2. Peach Bottom or Grand Gulf /NUREG-1150 supporting documentation [3,4], or

3. Pilgrim-specific " hand calculations."

The first node in the CPET considers the probability of recovering vessel injection and cooling the core debris in-vessel. The values used for the conditional probability of debris cooling in-l i

Pilgrim IPE Back-End Review 6 ERI/NRC E102

- - - - . . - - - - - - _ - - - - . - - ~ - - - _ - -

vessel are stated to have been obtained from the Grand Gulf /NUREG-1150 supporting ,

documentation. ATWS sequences with failure to initiate standby liquid control are assumed to be not coolable in-vessel.  ;

Given that core damage was not arrested in-vessel, the second CPET node is used to determine the size of the initial RPV lower head failure (i.e., either a small instrument tube or control rod drive peneuation failure, or a large breach). Conditional probabilities of large vessel breach given an in-vessel steam explosion (IVSE) or direct thermal attack are initially taken from Grand Gulf NUREG/CR-4551, then reduced by 60 to 90 percent. It is stated in the submittal (pages 4.5-10 and -11), that these lower values for the probability of a large breach are based on recent studies, with references provided. However, no rationale is provided to support the use and applicability of this general literature for Pilgrim. Note, that the effect of assuming a higher probability of a small RPV lower head breach results in a higher probability of ex-vessel debris cooling (due to the increased probability of high pressure melt ejection).

The third CPET node assesses whether or not the pedestal fails at the time of RPV failure. In the submittal, pedestal failure at RPV failure is considered to result from eithe- the dynamic loading associated with an ex-vessel steam explosion (EVSE), or from quasi-static overpressurization. As stated in Section 4.5.2.3, due to the relatively small depth of water resulting from flooding of the containments, it is considered unlikely that an EVSE would be of sufficient force to fail the pedestal walls, and a probability of 0.001 is assigned. However, water depth is not the most relevant parameter for determining the loading associated with an EVSE. Of more importance are parameters such as vessel failure mode, debris mass, degree of superheat, etc. Also, while it is inferred that this value is obtained from the Peach Bottom NUREG/CR-4551 study, no basis or actual origin for this value is provided. Similarly, the specific basis and origin are not provided for the value used for the probability of pedestal failure due to static overpressure given a low pressure case with a large RPV failure (page 4.5-14 of the submittal).

In addition, as discussed on page 4.5-15 of the submittal, for calculating the probability of static overpressurization of the pedestal under high pressure conditions, Peach Bottom analysis in NUREG/CR-4551 showed peak pressures ranging from 413 to 518 psid for large RPV failures, and peak pressures of 207-403 psid for small RPV failures. Pilgrim-specific MAAP calculations for these two cases are stated to indicate peak pressures of 109 psid and 30 psid, respectively.

No supporting documentation is provided for the Pilgrim-specific peak pressures, nor is an explanation provided for the substantial differences in reported peak pressures between the Peach Bottom and Pilgrim analyses.

The fourth CPET node considers drywell failure at RPV failure. As described in Section 4.5.2.4 of the submittal, the four drywell failure mechanisms considered include " alpha" mode (in-vessel steam explosion), pedestal structural failure, drywell liner melt-through, and drywell overpressurization. These failure mechanisms represent all of the early drywell failures of concern for BWR Mark I containments. Drywell liner melt-through, which has typicaliy been shown in other studies to be the dominant early containment failure mode for Mark I Pilgrim IPE Back-End Review 7 ERI/NRC 96-102

containments, is of somewhat lesser importance in the Pilgrim IPE, due to the incorporation of

'the results of a more recent study by Theofanous, et al. [5].

The probability of cooling the core debris in the pedestal following RPV failure . evaluated in the fifth CPET node. Values are provided for different cases depending on water availability in the cavity, the energy level of debris dispersal, and the quantity of initial debris release. It is important to note that for a large debris mass, non-energetic release to an initially dry cavity, the probability of non-coolability is assessed to be indeterminate, and therefore assigned a value of 0.5. Since, as stated in the submittal (pr.ge 4.5-23), if all debris fills the inner sumps, initial coolability is "unlikely," assigning a probability of 0.5 may be optimistic. Also, it is not apparent whether or not consideration was given to the probability that less th:m the maximum (or relatively little) heat might be removed from the debris bed, due to crust formation.

The sixth CPET node involves the probability of drywell over-temperature failure. As stated in the submittal (page 4.5-24), in the absence of drywell sprays and in-vessel injection following RPV failure, drywell gas temperatures will rapidly exceed 500*F, resulting in substmtial leakage through the drywell head silicone closure seals. .

The final three CPEI' nodes involve overpressure challenge to the containment from steam or non-condensible gas generation, wetwell failure or venting, and suppression pool bypass. As stated in the submittal, the containment will be subject to gradual overpressure if containment heat removal is unavailable, or if there is no ex-vessel debris cooling. Wetwell failure or venting will be prevented given drywell or RPV venting, or in the event of a drywell head overpressure failure. Suppression pool bypass (other than drywell failure or venting) is assumed to occur by either a stuck open wetwell/drywell vacuum breaker, or by loss of suppression pool inventory below the downcomer quenchers. Based on the Pilgrim containment structural evaluation, this last failure mode (i.e., loss of pool inventory) was deemed to be of insignificant probability.

In summary, the methodology employed in the Pilgrim IPE submittal is relatively well orgamzed and easy to comprehend. However, even though the Pilgrim CPET includes most of the relevant phesumena for BWRs with Mark I containments; nevertheless, the quantification of their impact on the Pilgrim containment is relatively weak. For the most part, the quantification is based on the NUREG-1150 study, but the applicability of many of these values to the Pilgrim plant is not adequately documented in the submittal. For instance, high pressure melt ejection / direct containment heating (HPME/DCH) is only considered indirectly, by using Peach Bottom NUREG/CR-4551 calculations for the drywell pressure rise at vessel breach (page 4.5-18 of the l submittal). Also, no truncation value is reported for the quantification of the PDSs; therefore, it is not possible to ascertain whether all important sequences have been analyzed through use of CPETs. l l

1 8 ERI/NRC 96-102 Pilgrim IPE Back-End Review 4

_.e- -- , w

3 2.2.3 Containment Failure Modes and Timing

]

As discussed in Section 4.4 of the submittal, instead of performing a detailed stru: nral analysis of the Pilgrim containment overpressure capacity, a review was performed of p.or structural l >

1 analyses of Mark I cort:ainment components, and the following potential failure lccations were

considered

l Drvwell Drywell Shell j Equipment Hatch i Personnel Airlock 2

Mechanical Penetrations

! Electrical Penetrations

, Drywell Head Closure Wetwell 1

Drywell to Wetwell Vent Line Bellows Wetwell Shell Below Downcomers l

Above Downcomers l

After comparing the mean failure pressures for all major failure locations (Table 4.4-1 of the l

i submittal), the three failure modes identified as potentially dominant were the vent line bellows, the drywell closure head, and the drywell shell. Fragility curves for each of these failure modes l

were constructed, and these were combined into a composite containment fragility curve (Figure 4.4-1 of the submittal). The median containment failure pressure was calculated to be 98 psig (at design temperature), at which pressure the drywell closure head contributes 36 percent to the i containment failure probability, and the wetwell vent line bellows contributes 64 percent. Note, that no reference is given for the origin of the mean failure pressures given in Table 4.4-1, nor j does the submittal contain any discussion as to the applicability of these values to the Pilgrim

! containment.

4 As stated in the submittal (page 4.4-4), at temperatures greater than 500*F, significant leakage j

will occur past the drywell closure head seal at pressures as low as 62 psig. Figure 4.4-2 of the l

j submittal shows the Pilgrim drywell failure pressure as a function of temperature. According  ;

to the submittal, tests performed by Sandia National Laboratories indicate that the BWR Mark I containment electrical penetration assemblies will not degrade at temperatures below 700*F [6].

As such, the drywell closure head seal is judged to be the controlling failure location for containment overtemperature failure.

i From information contained in Reference [7), the submittal concluded that the dominant failure locations are the drywell to wetwell vent line bellows (with a mean failure pressure of 100 psig) ,

i Pilgrim IPE Back-End Review 9 ERI/NRC %-102 i

4 k

4

and the drywell head closure (with a mean failure pressure of 125 psig). According to.the submittal, the MAAP analyses performed for the Pilgrim IPE considered both o' these failure modes, as well as drywell failure due to overtemperature. Also, for specific accide.t conditions, the drywell shell was assumed to fail due to direct contact by core debris.

In summary, the submittal appears to identify and analyze all relevant potential containment failure modes. All applicable containment failure modes from Table 2.2 of NUREG-1335 have been considered in the CPET analysis. In addition, the issue of containment overtemperature on penetration seals has been addressed. The procedure for obtaining an overall containment fragility curve using the data from Table 4.41 is reasonable, and standard for many IPEs.

However, no reference, or basis, for the use of the mean containment failure pressures of Table 4.4-1 has been provided.

2.2.4 Containment Isolation Failure In the Pilgrim IPE submittal, containment isolation failure is accounted for in the definition of plant damage states. In the PDS Grouping Diagram (Figure 4.3-3 in the submittal), the first two nodes are " Containment Bypassed" and " Containment Failed Prior to Core Damage." The latter node encompasses both sequences with containment failure prior to core damage, as we!! as sequences with loss of containment isolation. All containment bypass sequences are assigned to PDS 63, which has a frequency of 2.4 x 104per reactor year (ry). Sequences with the containment failed (or not isolated) prior to core damage are included in PDSs 44 and 51, which 4

have frequencies of 8.8 x 104/ry and 4.0 x 10 /ry, respectively, and their frequencies were calculated using fault trees.

All containment bypass sequences were m= pal directly into Release Category 34, and contribute 0.4 percent to total core damage frequency. Sequences with containment failure prior to core damage (including isolation failure) are mapped into Release Categories 17 through 25, and contribute 8.6 percent to total core damage frequency.

2.2.5 System / Human Pamanne The three operator actions accounted for in the CPETs are listed below:

OPER301 - Operators turn on drywell sprays prior to RPV failure (page 4.5-13)

OPER501(N) - Operators (fail to) initiate drywell sprays when required (pages 4.5-19 and -20)

OPER801 - Operators initiate drywell or RPV venting (page 4.5-26)

Probabilities are not provided in the submittal for these operator actions. Adiscussion describing the situation or factors that could affect human performance is also not included. l Pilgrim IPE Back-End Review 10 ERI/NRC 96-102 4

. l l

r .

I i

i Throughout the submittal (e.g., pages 4.8-23 to 4.8-25, pages 5.0-19 and -20, etc.), drywell and RPV venting are stated, and shown, to have a significant impact on the timing md magnitude j of fission product release. Due to the importance of these actions, more dedis regarding

venting procedures should be provided in the submittal. It is important to spec..y whether or not any local operator actions are required for initiation of wetwell or drywell/RPV venting, and, l

j include an analysis of the effect of the potentially degraded environment on human performance.

I.astly, since the availability of AC electrical power is not included as part of the PDS  ;

j definitions, or discussed as part of the CPET fault tree descriptions (Section 4.5.2), it is

! impossible to tell where, or if, credit is taken for AC power recovery.

i I 2.2.6 Radionuclide p.1,ne, rateenries and Characterintion y

The results of the CPET analyses lead to an extensive number of end-states, which are in turn l

binned for source term analyses. This process is analogous to the definition of PDSs for the i level 1 to level 2 interface. Outcomes of the CPETs are classified into a manageable number of release categories, which are categorized by similarities in accident progressica and source term characteristics.

i As discussed in Section 4.7.1 of the submittal, the sequence characteristics which were identified j as having the greatest impact on fission product release at Pilgrim are the following:

- Containment Bypass

]

- Debris Cooled In-Vessel

! - Time of Containment Failure / Venting (Relative to Core Damage)

! - Mode / Location of Containment Failure (or Venting)

- Suppression Pool Bypass

- Type of Ex-Vessel Core / Concrete Interactions (Dry, Wet or None)

Using these characteristics as headings, a Source Term Category Grouping Diagram was developed (Figure 4.7-1 of the submittal). The end points of this diagram represent the individual release categories. It is interesting to note that the Pilgrim IPE definition of release categories does not consider two characteristics typically considered in other IPEs; namely, the operation of drywell sprays, and the vessel pressure at the time of RPV failure.

A total of 34 release categories are defined by the diagram in Figure 4.7-1, of which only 20 att reported as non-zero (i.e., frequency > 10*/ry). As described in Section 4.7.3 of the submittal, MAAP calculations were performed to determine the source terms for representative sequences of 12 of the 20 non-zero release categories. Of the remaining eight release categories, the source terms for seven were determined to be similar to one of the previously calculated source terms. 'Ihe final release category involved no containment failure or venting, and was j assigned no source term. A summary of the releases associated with each non-zero release category is provided in Table 4.7-5 of the submittal. The source terms reported in this table do not take credit for any fission product retention in the reactor building.

Pilgrim IPE Back-End Review 11 ERUNRC %-102

---u._ _ . _ _ _ . ,_

- - . - - - . - . - - - ~_.- . - . - . - .-. - - .. _ . - . - - -

.* o

=

i Generic Letter 88-20 states that "any functional sequence that has a core damage frequency j greater than or equal to 104 per reactor year and that leads to containment faibre which can i" result in a radioactive release magnitude greater than or equal to the BWR-3 or P VR-4 release categories of WASH-1400," or "any functional sequences that contribute to a conte, unent bypass j frequency of 104 per reactor year," should be reported. From Table 4.7-5 of t.ie submittal, three release categories are identified as having frequencies greater than 104 per reactor year and ,

having Csl releases comparable to the BWR-3 category of WASH-1400 [8). All of these release j

3 categories (RC-14, RC-15 and RC-16) principally involve early containment failure caused by

! drywell liner melt-through (with the principal difference being the level of core / concrete I interaction). From Table 4.7-1 of the submittal, it is seen that there are three instances of a j PDS which contributes at least 104 per reactor year to core damage frequency for at least one j of these release categories. Specifically, PDS-21 contributes 2.17 x 104/ry to RC-16, and PDS-l 22 contributes 1.20 x 104/ry to RC-14 and 1.10 x 104/ry to RC-15. However, since the

dominant functional sequence frequencies for these PDS are not reported, it cannot be ascertained as to whether or not any functional sequences exceed the reporting criteria. In l

1 addition, the containment bypass release category (RC-34) has a frequency of 2.,40 x 104/ry, and consists only of PDS-63. Again, however, it is impossible to tell if any " functio $1 sequence" in PDS-63 exceeds the reporting criteria. In response to an NRC question, the licensee

confirmed that it was not possible to identify functional sequences that meet the reporting l

criterion, due to the complexity of the methodology used.

l

! The 34 release categories are further binned into five groups based on the magnitude of cesium -

s iodide release. The definition of the release groupings and the conditional probability of each ,

l group (expressed as a percentage of the CDF), are provided in Table 1.

It is important to note, that the submittal (on page 4.7-14) determined that only noble gas and i Cd source terms are "necessary or desirable" to be reported for the purposes of the IPE.

i Source terms associated with other radionuclide species are n~~~y for development of a full awareness of severe accident behavior. For example, the releases associated with refractory

fission products are necessary for determining the extent of ex-vessel releases in BWR Mark I containments. However, it is consistent with the NRC position on the reporting criterion.

i 2

Table 1 Radionuclide Release as a Percentage of the Internal Events CDF f

1 Release Category Grouping l Definition Percentage of CDF t

High (H) > 10% Csl Release 14.5 j Medium (M) > 1% Csl Release 64 j Low (L) > .1% Cs! Release 2.8

Low-Low (LL) > .01% Csl Release 1.7 Negligible Negligible Release 17 i

i Pilgrim IPE Back-End Review 12 ERI/NRC 96-102

}

i l

l

, o 9

• 1 i

Also, as stated above, one requirement of GL 88-20 is the identification of all functional sequences with frequency greater than 104/ry, and releases comparable to those < the WASH-l l, 1400 BWR-3 category. However, the functional sequences with frequency .iat meet the j' reporting criteria could not be determined from the submittal.

i i As part of this review, a comparison was made between the releases reported in Table 4.7-5 of the submittal and Figures B.3-2a and B.3-2b in Peach Bottom NUREG/CR-4551 [3]. Figure B.3-2a gives the range of expected radionuclide releases for accidents which result in early j drywell failure at high RPV pressure, which roughly corresponds to Pilgrim release categories

RC-14, RC-15 and RC-16. The Csl releases reported in the Pilgrim submittal for these three release categories all compare reasonably well with the mean iodine (I) release given in Figure i B.3-2a for Peach Bottom. Figure B.3-2b in Peach Bottom NUREG/CR-4551 gives the range of expected radionuclide releases for accidents resulting in containment venting, which roughly corresponds to Pilgrim (non-zero) release categories RC-2 to RC-5, and RC-17 to RC-20.

Again, the Csl releases reported in the Pilgrim submittal for these release categoHes compare reasonably well with the mean I release given in Figure B.3-2b for Peach BoNm, with the exception that those Pilgrim release categories which include suppression pool' scrubbing are

. roughly a factor of 10 lower than those in Peach Bottom. This difference is most likely due to I the fact that Pilgrim assumes a decontamination factor of 10 for saturated pool conditions.

2.3 Quantitative Ah of Accident Progression and Containment Behavior 2.3.1 Severe Accident Progression MAAP-BWR 3.0B Revision 7.03 was the principal tool used to analyze postulated severe

! accidents at Pilgrim. Even though MAAP-BWR 3.0B Revision 8.00 became available at the time of the analysis, it was not utihad since it had yet to receive widespread utility use or acceptance. As stated in the submittal, a few modifications were incorporated into the code version for the analysis. These modifications are discussed in Section 4.2.1.3.1 of the submittal. i Section 4.2.1.3.3 of the submittal discusses the incosporation of EOP modelling into the MAAP '

analyses. This was accomplished using MIPS, the MAAP input / output processor utility [9].

The MIPS input file is listed in Appendix D, which is not provided. The MAAP plant parameter file is briefly discussed in Section 4.2.1.3.4. A listing of the actual MAAP input file is included in Appendix G, which is not provided. Section 4.6.1 of the submittal includes a description of the accident progression, and MAAP results, for the three dominant types of sequences at Pilgrim (representing over 92 percent of total core damage frequency). Summary listings of major input parameters and assumptions, and of the results, for all MAAP runs are provided in submittal Tables 4.6-7 and 4.6-8, respectively. The complete input and output for all of the MAAP analyses are included in Appendix E, which is not provided.

i All of the phenomena of relevance to BWR severe accident phenomenology have been included in the submittal, as well as the principal phenomenological uncertainties. In Section 4.8.2 of the submittal, it is stated that based on recommendations in NUREG-1335 and the EPRI Pilgrim IPE Back-End Review 13 ERI/NRC %-102 i

),

" Guidance Document" for using MAAP [10), the following uncertainty issues were investigated j

through performance of sensitivity analyses with the MAAP code

s

- Core Melt Progression /In-Vessel Hydrogen Generation ,

Amount of Cote Debris Retained in RPV

! - RPV Pressure at Vessel Failure i

j - Containment Pressure I.oad due to RPV Failure i

- Direct Containment Heating

! - Shell Failure by Liner Melt Through l Debris Spread in Containment I - Ex-Vessel Debris Coolability

! - Containment Failure Imation l

- Containment Failure Area j' - Containment Flooding

- Saturated Pool Decontamination Factor i

A brief description of each sensitivity case is provided in Table 4.8-3 of the su%ittal, and a summary of insights and conclusions from these analyses is presented in Table 4.8-15.

l 4

2.3.2 Dominant Contributors to Containment Failure 1 .

Table 2 shows a comparison of the conditional probabilities of the containment failure modes provided in the Pilgrim IPE submittal, together with the results of the IPE submittals for the l

Fitzpatrick, Oyster Creek and Browns Ferry plants, as well as the NUREG-1150 study for Peach i Bottom. Note, that the results reported for Peach Bottom do not include internal flood events.

1 From review of Table 2, it is seen that the major difference between Pilgrim and the other

] BWRs with Mark I containments in the table, is that Pilgrim has a significantly lower conditional i probability of early containment failure, and accordingly, a much higher probability of late

- containment failure (with the exception of Oyster Creek). It appears that Pilgrim has understated the contribution from early drywell failure due to overpressurization (see Section 2.1.3.3 of this l

review). Also, for BWRs with Mark I containments, the dominant cause of early containment i

failure is drywell liner melt through. The submittal (p. 4.7-10) states that the principal reason

for the difference in early containment failure is due to the Pilgrim assumption that for sequences

with high vessel pressures at RPV failure, the probability of a liner melt through in a dry cavity l is only 0.1, if in-vessel injection is available following RPV failure. Also, as stated previously in Section 2.1.2.2 of this review, the contribution of drywell liner melt through reported in the l submittal, is somewhat lower due to the incorporation of the results from a more recent study.

j l

However, since the submittal does not provide the fraction of core damage frequency that is i associated with a dry cavity at the time of RPV failure, as well as other event probabilities i i

e associated with the probability of liner melt through, it is not possible to determine the exact reasons why Pilgrim exhibits a much lower probability of early containment failure.

l a

Pilgrim IPE Back-End Review 14 ERI/NRC 96-102 ,

N I

w - - - - - - -

e ,

. wm --- - - - .~ -- . . + - . - -,

I

?:

3.

3 9

m Table 2 Containment Failure as a Fercentage of Internal Events CDF:

g Comparison to Other BWR Mark I IPEs and Peach Bottom NUREG-1150 Results

?

@ Oyster Creek Browns Ferry Peach Bottom Pilgrim

," Containment Fitzpatrick IPE IPE IPE NUREG-Il50 IPE g Failure 5'

8 CDF (per year) 1.9 x 104 3.2 x 10 4 4.8 x 104 4.3 x 104 5.8 x 105 Early Failure 60.4 16.4 55.7 46 21.6 Bypass NA 7.3 NA NA 0.4 Late Failure 26.0 26.4 16.0 26 61.0 U Intact 2.5 0.0 18.4 3 1.2 No Vessel 11.1 50.4 9.9 25 15.8 Breach NA - Not Available i

M k

n .

2.

0

The relatively high probability of late containment failure reported in the submittal may be partially the result of the Pilgrim containment flooding strategy, which directs the operators to flood the containment if RPV water level cannot be maintained above the top of -etive fuel for non-ATWS sequences, or above two-thirds core height for ATWS sequences. Onw containment water level reaches the bottom of the recirculation lines, the operators are instructed to initiate

. RPV venting (to the condenser), which is classified as a late containment failure. It should be noted, that as stated in Section 4.8.2.1.11 of the submittal, the licensee is considering alternatives to the current procedure for containment flooding /RPV venting, since sensitivity analyses have shown that the current procedure has a negative impact on containment performance.

2.3.3 Characterindon of Containment Performance The MAAP code was used to calculate some accident progression parameters (e.g., peak pressures from quasi-static overpressurization at RPV failure, in order to assess the probability of pedestal failure), but for many other important phenomena, values were cotained from the Peach Bottom NUREG/CR-4551 study (e.g., probability of drywell overpressure fMure at vessel breach). More importantly, it does not appear that the submittal correctly' characterized containment performance for early drywell failure due to overpressurization. As described in Section 4.5.2.4, ranges of values for the pressure rise at vessel breach were obtained from Peach Bottom, then combiced with the possible range of containment pressures prior to vessel breach, and finally compared to the mean drywell failure pressure obtained for Pilgrim (125 psig). Since the resulting pressure peaks were all less than the mean drywell failure pressure, very small values were assigned to the probability of drywell overpressure failure (i.e.,0.001 for high pressure sequences, and non-zero probability for low pressure sequences). This treatment is approximate, and understates the potential for early drywell failures due to overpressunzation, since it does not account for the uncertainty in the containment fragility and in the containment loading associated with individual severe accident phenomena. Even at peak pressures well below the mean failure pressure, there may be a small probability of drywell failure. Table 2.5-10 of Peach Bottom NUREG/CR-4551 shows early drywell overpressure conditional failure probabilities of between 2 and 19 percent for different PDSs (as opposed to 0.1 percent for Pilgrim). It is indicative of the fact that the Pilgrim treatment of drywell overpressure failure may be optimistic. In addition, in all cases where the containment is challenged by overpressurization, there is a finite probability of failure of both the wetwell and the drywell, as shown in Figure 4.4-1 of the submittal.

Figures 4.7-5 and 4.7-6 of the submittal show the Pilgrim total core damage frequency broken down by containment release time and failure mode, respectively. From comparison of these tables, it appears that sequences with RPV venting are classified as late releases. It is not evident from the brief description of the containment flooding /RPV venting procedures provided in the submittal (pages 4.8-23 to 4.8-25), why RPV venting should be classified as a late release.

A number of uncertainties in the Pilgrim containment performance were treated indirectly through sensitivity analyses of several CPET event probabilities. As stated in the submittal, Pilgrim IPE Back-End Review 16 ERI/NRC 96-102 l

l

i i

sensitivity analyses were performed for those parameters which were judged to have large The sensitivity

! uncertainties or were expected to significantly influence the final results.

analyses performed include the following:

l i - Probability of In-Vessel Cooling

- Probability of Large RPV Failure Due to bwer Head Thermal Attack

- Probability of Large RPV Failure Due to In-Vessel Steam Explosion 4

- Probability of Pedestal Failure Due to Overpressurization at RPV Failure i - Impact of Water in Drywell at RPV Failure

- Probability of Liner Thermal Failure at RPV Failure

- Probability of Core / Concrete Interactions l

- Fraction of PDS Sequences with RPV Depressurized and With In-Vessel Injection Available

- Containment Flooding and Venting l <

- DCH Drywell Overpressure Failure l a

i l The principal insights from these sensitivities are provided in Section 4.8.1 of the 34bmittal, and summarized in Section 4.8.3.1. A brief listing of these insights is provided below:

.i v

- De probability of early containment failure is highly sensitive to assumptions regarding i liner melt through.

i

- - The probability of liner melt through is highly sensitive to the presence of water on the drywell floor at the dme of RPV failure.

4

- The probability of core damage arrest in-vessel is highly sensitive to RPV j depressurization (since, in many cases, depressunzation allows in-vessel injection using low pressure systems).

J

- The probability of containment failure due to drywell/RPV venting is very sensitive to the EOP-directed containment flooding strategy.

- In-vessel core damage arrest is only marginally sensitive to assumptions regarding in-

' vessel debris coolability.

i - The major level 2 results are not strongly sensitive to direct containment heating phenomena, pedestal over-pressurization at RPV failure, ex-vessel debris coolability and core / concrete attack, or initial RPV failure size, l

1 Pilgrim IPE Back-End Review 17 ERI/NRC 96-102 l

I 5

~

4 2.3.4 Imoact on Eauioment Behavior Section 4.1.2 of the submittal provides a discussion of the various contain>ent systems, including identification of some limitations on equipment operability due t. the adverse environment associated with severe accidents. The specific limitations identified include:

- Safety relief valves will be unable to operate to depressurize the reactor if containment pressure exceeds 60 psig.

- For sequences involving containment failure, the only high pressure injection system credited is the feedwater system, and the only low pressure injection systems credited are the condensate system and the fire water cross-Tie (due to uncertainties associated with equipment survivability in the reactor building following containment breach).

- Following containment failure, the only source of water credited for the drywell sprays is the fire water cross-tie. ,

Note, however, that no discussion is provided regarding the adverse effects of a severe accident environment on the flow of suppression pool water to heat exchangers for suppression pool cooling or to the drywell sprays (prior to containment failure).

2.4 Reducing the Probability of Core Damage or Hssion Product Release

2.4.1 Definition of Vulnerability 1

Section 5.1 of the submittal provides the following critena, used to determine if any vulnerabilities exist at the plant:

l i a) Are there any new or unusual means by which core damage or containment l failure occur as compared to those identified in other PRA's?

! b) Do the results suggest that the Pilgrim core damage frequency would not be able

! to meet the NRC's safety goal for core damage?

I I It is stated in the submittal that, based on the above criteria, no potential vulnerabilities were j identified for the Pilgrim Station.

l 2.4.2 Plant Modifications As stated in Section 6 of the submittal, prior to issuance of the generic letter, the licensee had 1 already made a number of plant modifications as part of its Safety Enhancement Program (SEP).

Of these modifications, those with the most impact on the IPE back-end analysis include:

I 1 i Pilgdm IPE Back-End Review 18 ERI/NRC 96-102

}

l

i  !

- Installation of a hardened vent path

- Modification of existing plant systems to provide an alternate source of i eter injection into the vessel (or to the drywell sprays) through the fire water cros? ie (which is

! protected from the harsh environment of a severe accident)

- Implementation of Revision 4 of the Emergency Procedure Guidelines l In addition, as described in Section 5.3 of the submittal, a number of important insights were obtained from the back-end containment analysis, of which the three most impor: ant are briefly l I l described below.

Containment Floodinn As previously described in Section 2.1.3.2 of this review, the Pilgrim EOP's require the operators to initiate containment flooding, often resulting in RPV or drywell venting, whenever adequate RPV water level cannot be maintained. Sensitivity analyses i performed with MAAP

• indicate that the containment flooding strategy ger eral!y results in containment radionuclide releases which are larger and earlier than for sequences w2re flooding i is not performed." As such, the licensee is considering alternatives to the current containment flooding / venting strategy, including limited drywell flooding, as described below.

5 Drvwell Floor Floodine. Since drywell liner melt through is the dominant contributor to early drywell failure, and it has been shown that the probability of liner melt through is much lower

if a pool of water exists on the drywell floor at the time of RPV failure, a strategy of limited j flooding of the drywell floor prior to RPV failure could greatly reduce, or essentially eliminate, the probability of drywell liner melt through. 'Ihis was demonstrated in the submittal through i performance of sensitivity analysis. As such, the licensee is investigating modifications to the

[ EOP's to incorporate limited drywell flooding using the drywell sprays.

9 i

RPV Deoraunrintion. Almost 62 percent of the Pilgrim core damage frequency results from j high pressure sequences. Sensitivity analyses described in Section 4.8.1 of the submittal

indicated that a significant increase in the probability of arresting core damage in-vessel can be obtained if RPV depressurization occurs early enough. Since the dominant cause of failing to j depressurize the RPV is operator error, the licensee is considering further evaluation of the i operator error rates assumed for the analysis, and possibly modifications to procedures or operator training.

[i

! 2.5 Responses to the Beammendations of the CPI %.=

Generic letter 88-20 supplement number 1 [11) and number 3 [12] identified specific

' Containment Performance Improvements (CPIs) to reduce the vulnerability of containments to l severe accident challenges. For BWRs with Mark I containments, the following improvements j, were identified:

ERI/NRC 96-102

) Pilgrim IPE Back-End Review 19 I

f . .

- Hardened vent, Alternative water supply for drywell spray / vessel injection,

- Enhanced reactor pressure vessel depressurization system reliability, ar.d

- Implementation of Revision 4 of the BWR Owners Group EPGs.

At the time of issuance of the Generic letter, the licensee had already implemented the following procedural and hardware modifications that are consistent with the recommendations of the CPI program:

e Installation of a hardened vent path, e Modification of the existing plant systems to provide an alternate source of water injection into the vessel through the fire water cross-tie.

e Installation of a third diesel generator to enhance the reliability of AC power.

e Implementation of Revision 4 of the BWROG EPGs.

e Installation of a backup nitrogen supply system to provide long term pneumatic control capability to the Automatic Depressurization System (ADS).

l In summary, all CPI recommendations have been addressed, either in the submittal, or j, previously as part of the licensee's Safety Enhancement Program.

1 1 4 l l

4 i

' l i

i 4

s i

i Pilgrim IPE Back-End Review 20 ERI/NRC %102 9

d i

~

p :-

4-wE4g

3. OVERALL EVALUATION AND CONCLUSIONS The back-end portion of the Pilgrim IPE submittal provides a substantial amount of information

in regard to the subject areas identified in Generic Letter 88-20 and NUREw-1335. The 1 i methodology used for the back-end analysis is basically sound, capable of ioentifying plant-specific vulnerabilities to release of radioactive material to environment, and includes the l

relevant phenomenological issues. The quantification of accident progression is based in large part on numerical estimates provided in NUREG-1150 supporting documentation; however, the submittal does not always contain adequate documentation as to the applicability of these i estimates to the Pilgrim plant.

l The important points of the technical evaluation of the Pilgrim IPE back-end analysis are j summarized as follows:

, e The back-end portion of this IPE submittal, for the most part, is relatively well performed and well written. -

e The number of non-zero PDSs was limited to 12, and the impact of each PDS on containment performance was specifically analyzed with a CPET.,

e The submittal includes all relevant phenomena of interest to severe accident phenomenology for BWRs with Mark I containments.

e The submittal considers the impact of severe accident conditions on the operability of equipment, e All CPI recommendations have been addressed, either in the submittal, or previously as part of the licensee's Safety Enhancement Program.

Several weaknesses (with regards to their overall impact on the IPE results) exist, and they include the following:

e The PDS definitions used in the submittal are not very transparent, due to the omission of some key accident characteristics (e.g., initiating event type, or status of AC power).

This makes it impossible to determine whether AC power is available for all sequences in a PDS, and whether AC power recovery has been credited for sequences involving loss of AC power.

• In a number of instances, the submittal makes use of parameters or probability values from the Grand Gulf or Peach Bottom NUREG-1150 studies, but does not provide a basis for their applicability to the Pilgrim plant.

• The treatment of the probability of early drywell failure due to overpressurization in the submittal is weak and poorly documented.

Pilgrim IPE Back-End Review 21 ERI/NRC %-102

_.. _._ . . _ _ _ _ .. ._.. _ _ _ _ ~ _ _ _ _ . _ _ _ . ~ _ _ _ _ _

t i e l e Generic letter 88-20 states that "any functional sequence that has a core damage frequency greater than or equal to 104 per reactor year and that leads to containment failure which can result in a radioactive release magnitude greater than a equal to the BWR-3 or PWR-4 release categories of WASH-1400," or "any functir.al sequences that contribute to a containment bypass frequency of 104 per reactor year,' should be reported in the IPE submittal. The licensee did not identify sequences that meet with the reporting criteria.

In spite of these identified weaknesses, the licensee's process for the evaluation of containment failure probabilities and failure modes is consistent with the intent of Generic letter 88-20, Appendix I. The dominant contributors to containment failure are consistent with the insights obtained from the NUREG-1150 analyses for the Peach Bottom plant. The licensee has considered the failure of the containment isolation system and containment bypass scenarios.

A number of sensitivity analyses have also been performed. In summary, it is concluded that j the IPE submittal provides'a substantial amount of information in regard to the subject areas identified in Generic letter 88-20 and NUREG 1335.

, )

i i l

r Pilgrim IPE Back-End Review 22 ERI/NRC %-102 I

r

\

4 1

4. REFERENCES

1. " Pilgrim Nuclear Power Station Individual Plant Examination for IntertJ Events per GL-88-20," Boston Edison Company, September 1992.

2. USNRC, " Severe Accident Risks: An Assessment for Five U.S. Nuclear Power Plants," NUREG-1150, Vols. I and 2, December 1990.

3. " Evaluation of Severe Accident Risks: Peach Bottom, Unit 2," NUREG/CR-4551, Vol.

4, Rev.1, December 1990.

)

4. " Evaluation of Severe Accident Risks: Grand Gulf, Unit 1," NUREG/CR-4551, Vol.

6 Rev.1, December 1990.

5. T. G. Theofanous, et al., "The Probability of Liner Failure in a Mark-I Containment,"

NUREG/CR-5423, August 1991. ,

6. FA1/91-76, "'Ihermal Attack of Containment Penetrations," Fauske & Associates, Inc.,

May 1991.

7. FAI/91-84, Rev. 2, " Containment Overpressurization," Fauske & Associates, Inc.,

January 1992.

8. USNRC, " Reactor Safety Study - An Assessment of Accident Risks in U.S.

Commercial Nuclear Power Plants," WASH-1400 (NUREG-75/014), October 1975.

9. Gabor, Kenton & Associates, Inc., "MIPS: An Improved Input and Output Processor for the MAAP Code," Version 1.80, May 1992.

10. Gabor, Kenton & Associates, Inc., ' Recommended Sensitivity Analyses for an Individual Plant Examination Using MAAP 3.0B," EPRI TR-100167,1991.

11. NRC Letter to All Licensees Holding Operating Licenses and Construction Permits for Nuclear Power Reactor Facilities, " Initiation of the Individual Plant Examination for Severe Accident Vulnerabilities - 10 CFR 550.54(f)," Generic Letter 88-20, Supplement No.1, dated August 29,1989.

12. NRC Letter to All Licensees Holding Operating Licenses and Construction Permits for Nuclear Power Reactor Facilities, " Completion of Containment Performance i

Improvement Program and Forwarding of insights for Use in the Individual Plant 1 Examination for Severe Accident Vulnerabilities- Generic Letter No. 88-20 Supplement No. 3 - 10 CFR 650.54(f)," Generic Letter 88-20, Supplement No. 3, dated July 6, 1990. l 23 ERI/NRC 96-102 Pilgrim IPE Back-End Review

13. Response to Request for AdditionalInformation Regarding the Pilgrim Individual Plant Examination (IPE) Submittal (TAC No. M74451), Enclosure to BEco Ltr. #95-127 from E. T. Boulette, Boston Edison Company, to U.S. Nuclear Reguir.tcr. Commission Dated December 28,1995.  ;

i l

l

\

l t

)

i e

l Pilgrim IPE Back-End Review 24 ERI/NRC 96-102 i

l l

l

l APPENDIX IPE EVALUATION AND DATA

SUMMARY

SHEET L

BWR Back-End Facts Plant Name Pilgrim Nuclear Power Station Containment Type Mark I Unique Containment Features None found ll l

Unique Vessel Featuns I

1 None found ,

Number of Plant Damage States 63 (12 non-zero)

Containment Failure Pressure 98 psig (median)

Additional Radionuclide Transport and Retention Structures No credit taken for retention in the Reactor Building or other structures l

Conditional Probability That the Contalpment is Not Isolated 0.090 (includes containment failure prior to core damage) l l

Important insights, Including Unique Safety Features MAAP sensitivity analyses indicate that the current strategy for containment flooding (and RPV venting) results in radionuclide releases which are larger and earlier than if flooding is not performed.

l l

Pilgrim IPE Back-End Review 25 ERI/NRC 96-102 l l

1 i

haplemented Plant Improvements Consideration is being given to alternative containment flooding strategier e.g., limited drywell flooding), and to increasing operator reliability of RPV depress ization.

Note: Plant modifications in response to all CPI recommendations have already been made as part of the BECo Safety Enhancement Program (SEP).

l C-Matrix l

See Table A.1 (attached) l l

l 26 ERI/NRC 96-102 Pilgrim IPE Back-End Review

.!:E j;  :!

• i ' ,, ,

_ . - . r

,: *i g _

1 1 4 3 4 0 t

2 8 9 1 1 4 0 3 5 7 7 9 3 0 6 6 1 0 0 a 1 0 1 0 0 9 9 4 4

• 0 0 2 2 1 0 5 0 -

t 0 0 0 0 0 2 2 0 0 0 0 0 0 0 0 0 0 0 1 0 o

t < < < < 1 0 0 4

~ 0 0 3 0 0

_ 6 1 1 7 7 3 3 0 2 7 7 0 1 2 0 5 1 1 3 3 1

m 7 3 0

_ 4 5 0 4 2 0 4 7, 1

_ 0 8 5 6 1 0 5 6 2 5 5 0

. 2 4 4 0 0

- 1 0

7 2 8 3 0 3 7 3 4 4 0 _

2 4 4 0 0 -

1 x

i r .

0 t 5 9 0 6 a 4 0 e 2 7 2 5 0

_ M t 2 4 4 0 0

- a C 1 t

e s i

r e .

g g _

a 0 _

l i m e 4 0 _

P a 1 9 0 0 D 2 3 6 1 _

t 1 n A l a

e P l 7 7 0 b 8 7 1 0

a 5 1 6 0 0 0 0 T

1 5 4 0 0 0 1

_ 1 7 6 5 0 0 2 6 0 0 0 1 5 4 0 0 0 _

1 _

8 9 7 7 0 _

_ 1 6 0 0 0 _

9 5 4 0 0 0 .

1 _

- 8 7 1 2 2 7 0 5 3 2 7 7 3 0 2 0 0 0 0 0 7 0 1

6 6 5 4 9 0 5 5 7 7 3 0 1

0 0 0 0 7 0 1

4 5 6 7 8 9 0 2 3 4 5 6 4 1 2 3 4 5 7 8 1 1 1 1 1 1 2 2 2 2 2 2 3 e y s r s a og l a

le e t t

o e a T

_ R C

,  ;! , 'll t-a

' ,9

• W* ::30 - r:.

- . %chE JW t

1P_sZAO U c? OJ t

_