ML20059H984

Summary of 930915 Meeting W/Science & Engineering Assoc,Inc Re Potential for Loss of ECCS in BWRs Due to LOCA Generated Debris.Draft, Expert Judgment Elicitation on Component Rupture... Done by Pnl Encl

ML20059H984
Person / Time
Issue date:	09/21/1993
From:	Serzik A NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
To:	Baer R NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES)
References
NUDOCS 9311100298
Download: ML20059H984 (73)

Text

_

/

o SEP 211993 MEMORANDUM FOR:

Robert L. Baer, Chief Engineering Issues Branch Division of Safety Issue Resolution FROM:

Aleck W. Serkiz Senior Task Manager Engineering Issues Branch Division of Safety Issue Resolution Office of Nuclear Regulatory Research

SUBJECT:

MEETING MINUTES

REFERENCE:

September 15, 1993 Meeting to discuss Potential for loss of ECCS in BWRs Due LOCA Generated Debris A meeting was held on September 15, 1993, to discuss planned analyses to be carried out by Science and Engineering Associates, Inc. for assessing the probability of loss of Emergency Core Cooling Systems (ECCS) in BWRs due to LOCA generated debris and to solicit views from knowledgeable parties in these technical areas.

Meeting attendees are listed in Enclosure 1.

The meeting purpose, safety issue background, SEA's work scope and meeting agenda are in.

SEA's presentation material is contained in Enclosure 3.

SEA's slides focused discussions on a wide range of considerations and approaches to estimating the probability of pipe breaks, debris generation and transport and blockage of RHR intake strainers.

These minutes are not designed to report such discussions in verbatim detail.

comments summarize the more important topics discussed.

The following 1.

The ability to select a single BWR as a generic plant for these analyses was questioned.

The staff has selected Duane-Arnold for use to model and estimate pipe break probabilities and attendant debris generation in a manner similar to the PWR analysis reported in Appendix D of NUREG-i 0869, Rev. 1.

This choice was questioned by a number of the attendees.

The discussions regarding generic applicability ranged from the variability of piping layouts in the drywell, differences in Mark I, II and III containment designs, to the RHR intake strainer design, location in the suppression pool and RHR/HPCI flow rates and distributions.

One option discussed for dealing with this issue was to perform sensitivity analyses to cover plants with different piping configurations, transport i

properties and strainer sizes.

this meeting.

A consensus approach was not reached at i

2.

Pipe break probabilities and attendant debris generation received considerable discussion.

NUREG/CR-4792, Volume 1, " Probability of failure in BWR Reactor Coolant Piping," was identified as the staff's reference to arrive at BWR pipe / weld failure estimates in a manner equivalent to Table 1 in i

9311100298 930921 PDR ORG NRRB

- I)'

PDR 0500n

l sgp

{

Robert L. Baer, Chief 2

Appendix 0 of NUREG-0869, Rev. 1.

Mr. Serkiz requested that if any attendees were aware of more recent, or better pipe failure data, that they should forward such information within the next two weeks to allow for review and assimilation into SEA's analyses.

Although no conclusions were reached on the above subjects, open i

discussion and sharing of analysis approaches by the participants occurred.

Dr. John O'Brien suggested use of expert judgement elicitation such as has been done for PWRs.

He also provided a draft report (Enclosure 4) for use as an example.

3.

The analysis methods described by SEA staff were questioned and discussed.

It is my perception that there was agreement that the problem should be analyzed in terms of three regions: a) the drywell (piping and insulation layout plus break probabilities), b) recognition of the variability of containment deign layouts and including such potential effects in estimating debris transport to the suppression pool, and c) incorporation of more recent information on insulation debris physical characteristics, transport and pressure drop characteristics.

There was also considerable discussion regarding the range of plant parameters to be studied by SEA.

The subject of suppression pool chugging following a LOCA and how such phenomena could impact debris transport and suspension was also discussed.

Other discussions referred to Barsebeck and other European studies, information and data acquired by U.S insulation vendors, industry groups and licensees.

NRC staff noted that effort was underway to obtain permission to release Barsebek and Sulzer information into the public domain and also recommended that all interested parties openly share pertinent information in a timely manner.

Mr. Serkiz noted that the SEA effort is underway with an initial findings report scheduled for December 1993 and that the current work scope is scheduled for completion in May 1994.

There appeared to be a consensus to hold a follow-up open meeting to discuss the results of the SEA initial analyses (such as this meeting) perhaps in January 1994 and attendees are advised that such a meeting would be scheduled.

The BWR thermal insulation types and ECCS characteristics which have been compiled by NRR staff are shown in Enclosure 5 and will be used in selecting parameter ranges along with feedback received at this meeting.

j 4.

Regulatory decision considerations and criteria were discussed.

l Attendees questioned planned analyses given the Commission's current guidance on adequate safety. Dr. R. Barrett, NRR noted that the NRC review process must decide if this is a safety enhancement issue or a l

compliance issue.

If, for a broad spectrum of breaks, conclusions are reached that this issue is a credible challenge to ECCS, then the issue will be in compliance space.

If results indicate findings are viewed as safety enhancements, then backfit cost benefit analysis rules will

1 f

=

I Robert L. Baer, Chief 3

SEP 2Ig apply.

These discussions again focused the strong dependence on estimated pipe / weld break probabilities and why more recent or other relevant information is needed quickly if these studies are to be modified.

5.

Several attendees noted that a Boiling Water Reactor Owners Group (BWROG) meeting for discussion of these BWR debris related concerns was i

being scheduled and that they would be participating.

NRC staff encouraged such participants to provide feedback from such meetings.

Mr. Serkiz thanked all participants for a frank and open exchange of ideas and views. He stated that information that can be used in SEA's analyses should be sent directly to Gil Zigler, Science and Engineering Associates, SEA Plaza, 6100 Uptown Blvd. NE, Albuquerque, NM 87110 and he should be copied with enclosures if not overly extensive (such as a large stack of drawings). He also recommended that written material be accompanied with text and figures in electronic format in Wordperfect 5.1 or ASCII format to assist in sharing such information with interested parties and storage in the PDR.

It was also noted that all material received would be placed in the public domain unless noted as requiring special handling.

Michael Marshall (301-492-3713) was identified as the NRC/RES backup contact for this issue.

OMGR#.L SGNED W Aleck W. Serkiz Engineering Issues Branch Division of Safety Issue Resolution Office of Nuclear Regulatory Research

Enclosures:

1.

Attendees 2.

Background, Work Scope and Agenda 3.

SEA Presentation Material l

4.

Draft Report:

" Expert Judgement Elicitation on Component Rupture Probabilities for Five PWR Systems" 5.

BWR Thermal Insulation Types and ECCS Characteristics i

DIST:

Central Files Baer R. J. Clark /M. J. Davis, NRR/PDI-2 PDR Norian J. R. White, RGN1/RPB2 Branch:RDG Serkiz G. S. Barber, RGN1/Susquehanna Site Minners A. Thadani L. Gifford, General Elec. Co.

l Meeting Att dees L. A. England, BWROG DSIR:EIB S'IYElB B

SERKlZ NORIAN BAER 9/p/93 9/JJ/93 942//93

=

j

,p' " %

., e t

UNITED STATES NUCLEAR REGULATORY COMMISSION

[g.kif j WASHINGTON. D.C. 205554101 g

j

\\.7,[.+#

SEP 211993 MEMORANDUM FOR:

Robert L. Baer, Chief Engineering Issues Branch Division of Safety Issue Resolution FROM:

Aleck W. Serkiz Senior Task Manager Engineering Issues Branch Division of Safety Issue Resolution Office of Nuclear Regulatory Research i

SUBJECT:

MEETING MINUTES

REFERENCE:

September 15, 1993 Meeting to discuss Potential for i'

Loss of ECCS in BWRs Due LOCA Generated Debris A meeting was held on September 15, 1993, to discuss planned analyses to be l

carried out by Science and Engineering Associates, Inc. for assessing the l

probability of loss of Emergency Core Cooling Systems (ECCS) in BWRs due to LOCA generated debris and to solicit views from knowledgeable parties in these i

technical areas. Meeting attendees are listed in Enclosure 1.

The meeting purpose, safety issue background, SEA's work scope and meeting agenda are in.

SEA's presentation material is contained in Enclosure 3.

SEA's slides focused discussions on a wide range of considerations and-approaches to estimating the probability of pipe breaks, debris generation and transport and blockage of RHR intake strainers. These minutes are not designed to report such discussions in verbatim detail. The following comments summarize the more important topics discussed.

1.

The ability to select a single BWR as a generic plant for these analyses i

was questioned. The staff has selected Duane-Arnold for use to model and estimate pipe break probabilities and attendant debris generation in a manner similar to the PWR analysis reported in Appendix D of NUREG-0869, Rev. 1.

This choice was questioned by a number of the attendees.

The discussions regarding generic applicability ranged from the variability of piping layouts in the drywell, differences in Mark I, II 4

and III containment designs, to the RHR intake strainer design, location in the suppression pool and RHR/HPCI flow rates and distributions. One option discussed for dealing with this issue was to perform sensitivity analyses to cover plants with different piping configurations, transport properties and strainer sizes. A consensus approach was not reached at i

this meeting.

2.

Pipe break probabilities and attendant debris generation received considerable discussion.

NUREG/CR-4792, Volume 1, " Probability of Failure in BWR Reactor Coolant Piping," was identified as the staff's reference to arrive at BWR pipe / weld failure estimates in a manner equivalent to Table 1 in

. _ ~

.m J

Robert L. Baer, Chief 2

{

Appendix D of NUREG-0869, Rev. 1.

Mr. Serkiz requested that if any attendees were aware of more recent, or better pipe failure data, that they should forward such information within the next two weeks to allow for review and assimilation into SEA's analyses.

[

Although no conclusions were reached on the above subjects, open l

discussion and sharing of analysis approaches by the participants d

occurred. Dr. John O'Brien suggested use of expert judgement elicitation such as has been done for PWRs. He also provided a draft report (Enclosure 4) for use as an example.

l t

3.

The analysis methods described by SEA staff were questioned and discussed.

It is my perception that there was agreement that the problem should be analyzed in terms of three regions: a) the drywell I

(piping and insulation layout plus break probabilities), b) recognition of the variability of containment deign layotts and including such i

potential effects in estimating debris transport to the suppression pool, and c) incorporation of more recent information on insulation l

1 debris physical characteristics, transport and pressure drop 1

characteristics. There was also considerable discussion regarding the i

range of plant parameters to be studied by SEA. The subject of l

suppression pool chugging following a LOCA and how such phenomena could

)

impar.t debris transport and suspension was also discussed.

l Other discussions referred to Barsebeck and other European studies, e

information and data acquired by U.S insulation vendors, industry groups and licensees. NRC staff noted that effort was underway to obtain permission to release Barsebek and Sulzer information into the public domain and also recommended that all interested parties openly share pertinent information in a timely manner.

j Mr. Serkiz noted that the SEA effort is underway with an initial findings report scheduled for December 1993 and that the current work scope is scheduled for completion in May 1994. There appeared to be a consensus to hold a follow-up open meeting to discuss the results of the SEA initial analyses (such as this meeting) perhaps in January 1994 and attendees are advised that such a meeting would be scheduled.

The BWR thermal insulation types and ECCS characteristics which have j

been compiled by NRR staff are shown in Enclosure 5 and will be used in selecting parameter ranges along with feedback received at this meeting.

i 4.

Regulatory decision considerations and criteria were discussed.

Attendees questioned planned analyses given' the Commission's current 1

guidance on adequate safety. Dr. R. Barrett, NRR noted that the NRC review process must decide if this is a safety enhancement issue or a j

compliance issue.

If, for a broad spectrum of breaks, conclusions are reached that this issue is a credible challenge to ECCS, then the issue j

will be in compliance space.

If results indicate findings are viewed as safety enhancements, then backfit cost benefit analysis rules will j

a

~.. - >

_.+

a Robert L. Baer, Chief 3

apply. These discussions again focused the strong dependence on estimated pipe / weld break probabilities and why more recent or other l

relevant information is needed quickly if these studies are to be j

i modified.

i 5.

Several attendees noted that a Boiling Water Reactor Owners Group (BWROG) meeting for discussion of these BWR debris related concerns was i

being scheduled and that they would be participating. NRC staff encouraged such participants to provide feedback from such meetings.

j i

Mr. Serkiz thanked all participants for a frank and open exchange of ideas and l

views. He stated that information that can be used in SEA's analyses should be sent directly to Gil Zigler, Science and Engineering Associates, SEA Plaza, 6100 Uptown Blvd. NE, Albuquerque, NM 87110 and he should be copied with enclosures if not overly extensive (such as a large stack of drawings). He i

also recommended that written material be accompanied with text and figures in electronic format in Wordperfect 5.1 or ASCII format to assist in sharing such information with interested parties and storage in the PDR.

j i

j It was also noted that all material received would be placed in the public 4

domain unless noted as requiring special handling.

t Michael Marshall (301-492-3713) was identified as the NRC/RES backup contact for this issue.

N' y

Aleck W. Serkiz Engineering Issues Branch l

Division of Safety Issue Resolution Office of Nuclear Regulatory Research j

4

Enclosures:

1.

Attendaes 2.

Backgrcund, Work Scope and Agenda 3.

SEA Presentation Material 4.

Draft Report:

1

" Expert Judgement Elicitation on Component Rupture Probabilities for Five PWR Systems" l

S.

BWR Thermal Insulation Types and ECCS Characteristics 4

i

2 1

\\

Meeting Attendees.

September 15,1993 Potential for Loss of Emergency Core Cooling Systems in BWRs l

Due to LOCA Generated Debns i

NAME ORGANIZATION PHONE i

i Al Serkiz NRC/DSIR 301-492-3942 Bob Walsh SEA, Santa Fe 505-2.63-1875 Gilben Zigler SEA, Albuquerque 505-884-2300 l

D.V. Rao SEA, Albuquerque 505-884-2300 l

Frank Sciacca SEA, Rockville 301-468-7371 l

Gordon Han Performance Contracting 913-441-0100 l

i Hiram Reppen Centerior Energy (Perg) 216-259-3737 Paul Roney Centerior Energy (Perg) 216-259-3737 i

Bmce Alpha Transco Products 312-427-2818 Len Barnes Sequoia Consulting Group 617-961-0033 Hans Wolff SNC/Leibstadt 408-438-6444 Alan Bilanin CDI 609-734-9282 Mark Mjaatveot PP&L 215-774-7795 i

R.R. Sgarro PP&L 215-774-7914 Roben Pulsifer NRR 301-504-3016 Jim Kinsey IELP 319-851-7177 Phil Rush NRR 301-504-1415 Richard Barrett NRR 301-504-3627 C

Many Virgillo NRR 301-504-3226 Daniel Maret Consultant to EPRI 704-547-6010 i

Bill Houston EPRI 704-547-6059

(

P.K. Eapen NRC/ REGION I 215-337-5150 B. Siegel AMS 703-836-0300 l

John O'Brien NRC/RES 301-492-3894 Richard Lobel NRC/NRR 301-504-2865 Laura Comes SEA, Alexandria 703-549-8884 Michael Marshall NRC/RES/DSIR 301-492-3713 Paul Norian NRC/DSIR 301-492-3910 Mike Hayner CE1 216-259-3737 1

~_,

i.

Potential for Loss of ECCS in BWRs Due to LOCA Generated Debris l

I l

9-15-93 Meeting Aleck W. Serkiz USNRC RES/DSIR/EIB 301-492-3942 MS NLS-314

Meeting Purpose Discuss planned BWR analysis to assess probability ofloss of ECCS due to LOCA generated debris and solicit views on planned analyses from knowledgeable parties.

1

Background

i 1.

USI A-43,

" Containment Emergency Sump Performance" was resolved in 1985 (SECY-85-349) and resulted in GL 85-22 (12-3-85), RG 1.82,

" Water l

Sources for Long Term Recirculation Cooling Following l

a LOCA", NUREG-0897, Rev.1, NUREG-0869, Rev.1 and SRP 6.2.2.

2.

Backfit action was not selected in 1985, and since that time a number ofinformation notices addressing debris effects have been issued. These were listed in the meeting notice.

3.

The Barsebeck strainer plugging (July 1992) and Perry strainer clogging and damage by debris (March 1992 -

January 1993) events have resulted in a need to more closely evaluate BWR loss of ECCS due to debris blockage.

4.

Planned work is focused on analyzing a BWR in a manner similar to the PWR analysis described in Appendix D of NUREG-0869, Rev.1, evaluate debris transport models, estimate CDF impacts and develop potential mitigating action (s).

/

I

Planned Work Contractor:

Science & Engineering Associates Albuquerque, NM Time Frame:

9/93 - 5/94 (Initial Findings 12/93)

Tasks i

Task 6.1 Break Model Development & Plant Characteristics Obtain piping and insulation information Program piping / break, debris model & checkout Parametric debris generation / flow blockage cales (Q= 5,000 to 15,000 gpm, L/D = 3 - 7)

Task 6.2 Transient Debris Transport & Blockage Modeling (Plant layout and materials dependent)

Transport of LOCA generated debris Estimation of strainer blockage & pressure loss Task 6.3 Assessment of Regulatory Impacts Estimated CDFimpacts Impacts of potential mitigating guidance Need for revising RG 1.82, Rev.1 Task 6.4 Effects of Corrosion Products and Other Foreign

?

Materials on RIIR/LPCI Intakes i

This work will not commence until findings from i:

preceding tasks have been evaluated and a well defined i

scope has been prepared for this subtask.

l Agenda 9-15-93 Meeting 9:00 - 9:15 A31 Introduction & Background (A. Serkiz) t i

9:15 - 9:30 A51 Introductory Comments from Attendees 9:30 Ah!

Selection Criteria & Candidate BWRs Data Needed (Gil Zigler & Bob Walsh)

[

Piping layouts & insulation distribution Break probability selection and assignment Review of available information for BWRs Break model and estimating debris generation (Bob Walsh)

NUREG/CR-0896 methods

~

Needed modifications for B%R studies Other Approaches (attendees) 11:30 Ah!

Hydraulic model development (D.V. Rao)

LOCA debris generation & transport Suction strainer blockage Blockage modeling vs. material dependence l

Generic analyses vs. plant specific dependence 12:30-1:30 Lunch Break 1:30 Regulatory Analysis (Bob Walsh)

Estimating CDF Impacts Identification of remedial action (s)

Cost / Benefit estimation 2:30 - 3:30 Wrap-up Discussions (A. Serkiz)

Attendee comments What changes are netded to work planned ?

_ _ _ _., ~. _

l Potential for Loss of ECCS in BWRs Due to LOCA Generated Debris Discussion ofData Requirements Gilbert Zigler / Bob Walsh SCIENCE AND ENGINEERING ASSOCIATES Albuquerque, New Mexico Presented to NRC Meeting on Potential for Loss of ECCS in BWRs Due to LOCA-Generated Debris Rockville, MD September 15,1993 J

TOPICS Surrogate " Generic" BWR Piping Layout & Insulation Distribution i

Break Frequency Selection and Assignment

'l Review of Available Information for BWRs

(

e v-v

^

Surrogate " Generic" BWR DUANE ARNOLD 1

CURRENT SELECTION AS THE " GENERIC"BWR SURROGATE e

y

(

PIPING LAYOUTS & INSULATION DISTRIBUTION Surrounding Piping Layouts Vary With Containment Type Applicability of target segments for each break segment from reference plant SW,

e

j i

1 i

t t

i i

r 3

I i

?

a i

hI O.T ES I

!f j

__.l l.

t

-l-I t

99 o

7 2

i

)

I

,/

~

w

{

Plal45 tel36JtJ3tS @

?,--

i 3

-a g

. tassaA aJnsssJd J013f aN 4 :n sspa, 4

~~

1

[

N O, N,,,D AJewlJd l

N

t s

,E_

89:

e L

tf 4.

4 i

~

e..

I r

N m

2 i

e"

• = =,

s em l "

L m

lO!

M L

R

,k

,v

/"

M M

xkc r-Y O~

O Hh k

__S s

_ f-V g

3-b i

ii i

n w"m e

I N

f g

E t

t t

I I

{

i 1

i

. I

g n

n

~

l n-

,+ n a

E

~

/

. :n s

n g.

g; r- ;

-n a.

,/

/

w i...i 1.,

.... =I

,N I_, e

": ~

g s if.-

[.9-W' E>

y i

t a

d

~

.,n _j$

g g

vi;>

o C

Y

{

N o

N e

i J

=

q

a r

m Break Frequency Selection A-43 used pipe break frequency of 3.76E-4/ reactor l

year distributed over 5 pipe size groups NUREG/CR-4792 pipe break frequency neglecting IGSCC is several order of magnitudes less than A-43 NUREG/CR-4792 presents a pipe break frequency value of 1E-3/ reactor year for IGSCC breaks in risers and bypass i

M>

I b

F 3

NUREG-0869, Revision 1 Piping failure probability estimates Pipe Failure Weld failure probability distribution size Diameter probability Weld Weld Weld (inches) class (1/Rx-Yr) type 1 type 2 type 3 D

J P)

W, W,

W, 2 to <6 1

3x10-4 0.7 0.15 0.15 6 to <10 2

4x10-5 0.5 0.30 0.20 10 to <16 3

3x10-5 0.5 0.30 0.20 16 to 28 4

3x10-s 0.5 0.30 0.20

>28 5

3x10'S 0.5 0.30 0.20 IP) = 3.76x10-4/Rx yr Weld Type 1 = fabricated and non-standard joints Weld Type 2 = high restraint joints and tees with joints Weld Type 3 = elbows, reducers, and straight piping runs with joints 5

1 4

L 1

.m m.m.

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

-e

- -. ~,..

m m

a L

r 3

i o

i I

i l

w i

B I

I i

~e

-n t

_ g e

- g.

EF 3o 4on o

4 4

1 l

i y

T a

o

[f o

o o

(dool Jed) 8030 Jo AlligegoJd eAlielnmno i

l

(

)

i e

i

/

a I

3 i

g i

l I

l i

t I

.f5 kN I7 l

l

)

E 8

AN%%\\'%3R%%W l

_I.

4 I

1 j

J ac

-o

%_ _cB *=

m o e z 3* 8

_.5 n n -

E x

s m

i 4

4 4

4 o

~m e

n, o

6 o

6 6

o (803Old to uopeJd

l 1

i i

~

Break Frequency Assienment A-43 assigned break frequency either by length or by weld weighting factor NUREG/CR-4792 details will be reviewed for possible weld weighting factors 1

Applicable BWR Information Brunswick schematic piping drawings NUREG/CR 4792 Barsebeck related analysis Other communications l

J 1

I!

Potential for Loss of ECCS in BWRs Due to LOCA Generated Debris Break Model and Estimated Debris Generation BOB WALSH SCIENCE AND ENGINEERING ASSOCIATES Albuquerque, New Mexico i

Presented to NRC Meeting on Potential for Loss of ECCS in BWRs Due to LOCA-Generated Debris Rockville, MD September 15,1993 M;

)

TOPICS NUREG/CR-0896 Methods Needed Modifications for BWR Studies SW >

e

r 3

NUREG/CR-0896 Methods Break Frequency by Pipe Segment l

Debris Generation Given a Segment Break Does Segment Break Cause Blockage?

A i

r 3

Break Frequency by Pipe Seement PWR data For A-43 estimated from IAEA (1977)

- LOCA freq by diameter class

- weighting factors by weld type l

List of pipe segments from reference plant drawings 4

LOCA frequencies distributed among pipe segments

- weighted by length or

- weighted by weld type factors 1

[M >

a

\\

I j

r N

2 E

P I3 M

\\ **

1 le31 P

~

R

'k

,n /

/-

m x

x%_e g

Lo 3 s

S r

kg 0-I f

w I

imi i

n w"m o

N s

-1 a.

m i

k

~

Debris Generation Given a Segment Break Select L/D (varied parametrically for A-43)

List of target segments for each break segment from reference plant drawings Select target segments that have fibrous insulation i

Calculate and sum fibrous insulation volume l

A s

e i

jl

~

)

J 3

e D

(

U(s\\

75 m

DD I

\\

.s UU e

} t J_

rF.

4 I

3 h

PC R

no i

I t

areneg s

l i

r le b

W k

/

e I

\\

/

d en

)/

r o

O G

o rC

'C S

f 1

\\

d e4 z9 P

3 5 7

t 3 C

R

=

i3

\\

t -

D D D tR U U U uC

\\

/

eG cE nR eU y

uN l

f n ni

)

V-i D

s

(

f e re ot t

a le e

sm s

p m

ei m

0e nt se os i

1 V

u tD Z e P

e e

r gk o

t o

r n

r u

t s

a a eo Tr r

l s

e o B

e n o o

t y r

r P e C b

e e

G v

r g cd ro o e ne o

'm t

t L

s c a c ed s

t e

a t

iM o e

e t

e o e

r R S R H DD C V G P L D O

P S C H U C

R R

e re h

W

(

4

Does Segment Break Cause Blockage?

Select ECCS flow rate, screen area, and allowable head loss (all varied parametrically for A-43) 1 Estimate head loss if all fibrous debris reaches screen, using equation of NUREG/CR-2982 (1982) for high density fiberglass Compare head loss estimate with allowable head loss

Needed Modifications for BWR Studies LOCA frequencies Replacement for head loss formula?

V 1.54 QF 1.84

~

~

H=1653.AZ.

AZ.

1 I

Potential for Loss of ECCS 1

In BWRs Due to LOCA Generated Debris Review ofApplicable Analyses and Data D. V. Rao SCIENCE AND ENGINEERING ASSOCIATES Albuquerque, New Mexico Presented to NRC Meeting on Potential for Loss of ECCS in BWRs Due to LOCA-Generated Debris Rockville, MD September 15,1993 c

i

= -

m ACCIDENT PROGRESSION SEQUENCE

1. GENERATION OF DEBRIS BY LOCA IN THE DRYWELL REGION.

2. SHORT-AND LONG-TERM TRANSPORT OF DEBRIS TO THE ECCS SUPPRESSION POOL.

3. DEBRIS TRANSPORT IN THE SUMP. EFFECT OF FLOW PATTERNS.

I

4. DEBRIS ACCUMULATION ON THE STRAINERS. HEAD LOSS DUE i

TO PRESSURE DROP INDUCED BY DEBRIS LAYER.

i TIME PERIODS OF ACCIDENT PROGRESSION PROBABILITIES OFINDIVIDUAL EVENTS i

y q

?

~

r m

DEBRIS GENERATION BY A LOCA PRESENT UNDERSTANDING:

DEBRIS GENERATION BY LONGITUDINAL BREAKS IS NEGLIGIBLE DEBRIS GENERATED BY PIPES < 6 IN. DIAMETER IS ACCOUNTED FOR IN THE ANALYSIS CONSERVATISM.

HDR TESTS AND CANADIAN TESTS DEMONSTRATE EXTENT OF DEBRIS GENERATION ALL THE INSULATION WITHIN L/D OF 7 MAY BE DISLODGED FROM THE STRUCTURES. SHREDDED INTO SMALL PIECES?

UNCERTAINTIES:

HOW MUCH PAINT AND RUST DEBRIS WILL BE RELEASED?

WHAT SHAPE AND SIZE WILL SUCH DEBRIS BE RELEASED?

WILL DEBRIS DISLODGMENT DEPEND ON THE L/D AS WELL AS DEBRIS PACKAGING AND ATTACHMENT TO THE STRUCTURES?

A>

e

r m

DEBRIS TRANSPORT NUREG/CR-3394 ASSUMES THAT ALL GENERATED DEBRIS IS TRANSPORTED TO THFs SUMP IN SHORT-TERM.

KKL ANALYSES ASSUME THAT ONLY ABOUT 50%-60% REACH THE SUPPRESSION POOL. REST DEPOSITED ON DRYWELL STRUCTURES AND CONTAINMENT STRUCTURES.

1 2

SOME ANALYTICAL MODELS WERE PRESENTED IN NUREG/CR-2791.

THEIR APPLICABILITY WAS NOT CHECKED.

TRANSPORT IS COMPLICATED BY (a). DIFFERENCES IN MARK I, 1

MARK II AND MARK III DRYWELL STRUCTURES; (b). DIFFERENCES l

IN FLOW PATH. SOME DRYWELLS HAVE GRATING ON TOP OF THE RISER TUBES AND SOME HAVE CONSTRUCTED LIPS AROUND THE l

RISER PIPES.

FLUID VELOCITIES MAY HAVE AN IMPACT ON THE TRANSPORT. IF i

INSULATION IS TO BE SHREDDED INTO FIBERS, TRANSPORT IS EASY.

l TIME SCALE FOR TRANSPORT SHOULD BE ASSESSED.

SW,

g i

1

r 3

DEBRIS TRANSPORT WITHIN THE SUMP PRESENT ASSUMPTION:

ALL THE DEBRIS TRANSPORTED TO THE SUPPRESSION POOL WILL ULTIMATELY DEPOSIT ON THE SUCTION STRAINER.

THE DEBRIS LAYER FORMATION WOULD RESULT IN A UNIFORM DEBRIS LAYER ON THE STRAINER.

UNCERTAINTIES:

WOULD THE POOL RECIRCULATION PATTERNS INFLUENCE DEBRIS TRANSPORT WITHIN THE SUMP?

HAS THERE BEEN ANY EXPERIMENTAL INVESTIGATION OF THIS t

PHENOMENON ?

9 i

RESOLUTION WILL PROBABLY IMPRO E TIIE ACCURACY A PARAMETRIC STUDY MAY BE USED TO ASSESS ITS EFFECT A>

e l

t HEAD LOSS DUE TO STRAINER BLOCKAGE NUREG 0897, REV.1 PROVIDES FOLLOWING RELATIONSHIP:

AP= A.Vh.tc V

= Strainer Approach Velocity, t

= Fibrous Debris Layer Thickness, A,b,c = Empirical Material Constants.

For Mineral Wool

A= 123; b=1.51 and c=1.36 For Low Density Fiberglass

A= 68.3; b=1.79 and c=1.07 j

For High Density Fiberglass

A=1653; b=1.84 and c=1.54 EMPIRICAL EQUNTIONS DERIVED FROM SEPARATE EFFECTS i

EXPERIMENTS. HAVE NOT BEEN CHECKED FOR APPLICABILITY.

)

ANALYTICAL SOLUTIONS MAY BE POSSIBLE. NOT EXPLORED J

COMPLETELY.

I j

q i

r m

HEAD LOSS DUE TO STRAINER BLOCKAGE i

DEBRIS ANALYSES SUGGEST DEPOSITION OF RUST AND FINE PARTICULATE TOGETHER WITH THE FIBERS. THIS MAY FURTHER l!

INCREASE HEAD LOSS.

i APPLICABILITY OF CORRELATIONS TO ACTUAL CASES MAY POSE SEVERAL UNCERTAINTIES.

- EXPERIMENTAL RANGE OF PARAMETERS Vs. ACTUAL CASES

- REPRODUCIBILITY OF EXPERIMENTAL DATA q

J

r m

HEAD LOSS DUE TO STRAINER BLOCKAGE INDEPENDENT TF, STING BY KKL EXHIBITED A MORE SEVERE PRESSURE DROP BEHAVIOR.

KKL AP= 318.V113.t 14 i

I NUREG-0897 AP= 123.V1.si.t.36 1

THE DIFFERENCE IS 200-300% IN THE OPERATING PARAMETER RANGE OF INTEREST.

DIFFERENCE IN EXPERIMENTAL CONDITIONS MAY EXPLAIN SOME OF THIS DEVIATION:

I KKL P= 4.5 Bar; T = 15 C j

ADL Experiments P= 1 Bar

T Varied ALSO, KKL MEASURED AT MAX. BED COMPACTION.

i

\\

DIFFERENCE MAY BE EXPIAINEDIF BED l

COMPACTION FACTORS ARE TAKENINTO CONSIDERATION k

J

Potential for Loss of ECCS in BWRs Due to LOCA Generated Debris Regulatory Analysis BOB WALSH SCIENCE AND ENGINEERING ASSOCIATES 4

Albuquerque, New Mexico l

s Presented to NRC Meeting on Potential for Loss of ECCS in BWRs Due to LOCA-Generated Debris i

Rockville, MD September 15,1993 j

o

~

TOPICS Estimating CDF Impacts Identification of Remedial Actions Cost / Benefit Estimation t

. ~.. -.

r Estimating CDF Impacts Perform Parameter Study

- Flow rate 5000 to 15000 gpm

- L/D 3 to 7

- Screen area 50 to 150 sq ft

- Allowable head loss 1 to 5 ft H2O Assume 50% of blockage events result in core damage l

m

Identification of Remedial Actions Replacing Insulation Increasing Suction Screen Area Other???

• F Cost / Benefit Estimation i

Use Parameter Study to Estimate " Generic" CDF Reduction for Each Proposed Remedial Action l

Use 5 E6 Person-Rem per Core Melt for Mark 1 & 2 i

Use 5 E5 Person-Rem per Core Melt for Mark 3 l

Use FORECAST to Estimate Cost and Exposure Impacts for Proposed Remedial Action i

i SW,

w i

i

WRAP-UP DISCUSSIONS 1

Attendee Comments What Changes are Needed to Work Planned t

J

?

ENCLOSURE 4 DRAFT

~

EXPERT JUDGMENT ELICITATION ON COMPONENT RUPTURE PROBABILITIES FOR FIVE PWR SYSTEMS Truong V. Yo Nuclear Systems and Concepts Department Pacific Northwest Laboratory Richland, Washington 99352 Fredric A. Simonen Automation and Measurement Sciences Department Pacific Northwest Laboratory Richland, Washington 99352 l

Bryan F. Gore James V. Livingston Nuclear Systems and Concepts Department Nuclear Systems and Concepts Department Pacific Northwest Laboratory Pacific Northwest laboratory Richland Washington 99352 Richland, Washington 99352 ABSTRACT which is being conducted at Paci6e Northwest Laboratory (PNL)

In the work sponsored by the Nuclear Regulato y Commission for the Nuclear Regulatory Commission (NRC), are to detertmne (NRC), Pacific Northwest Laboratory (PNL) has developed a the rehabihty of current inservice inspections of pressure risk-based method for estabbshmginspection pnorities for nuclear boundary systems and components and to develop recommenda-power plant components. In this method, the results of probabilis-tions that can ensure a suitably high inspection reliabihty.

tie risk assessment (PRA) are used to esumate the safety conse-In meeung the program objectives, a risk-based method has quences of component failures. he method also requires esti-been developed (Vo et al.1989). In this regard, results of mates of structure and component failure probabilities. Because probabihstic risk assessment (PRA) have been useful for esti-neither suf5cient data from operating experience nor detailed matmg the safety consequences resulung from the failure of fracture mecnerucs analyses are available, expert judgment was parucular system components; however, the methodology also needed to surnlement and interpret the available data on failure requires estimates of the probabihties of failure for the com-probabih6es ponents of interest. In performing trial applications for the risk-The plant selected for the chestation was the Surry Nuclear based methodology, a signi6 cant lack of failure probabihty data Power Stanon. Urut 1. To iruuate the analysts, PNL conducted has been encountered at the required level of detail. Histoncally the first of two expert judgment eheitation workshops in May there have been very few structural failures of the pressure 1990 at Rockville, Maryland, to address four selected safety sys-boundary components, giving a high level of uncertamty to tems at Surry-1. A second workshop was held in February 1992 stanstical esumates of failure probabihties. Probabihsuc fracture in Washington, D.C., to address the remammg safety systems and mecharucs models have addressed a few components of interest, balance-of-plant systems. This paper gives results for the Eve but these calculations also have a high degree of uncertainty.

systems addressed in this second workshop, includmg residual To overcome this data de6ciency, PNL has conducted expert heat removal, high pressure injection, power conversion, service judgment workshops to address the issue of failure probabihties.

water, and component cooling water systems.

The Surry Nuclear Power Stanon, Urut 1 (Surry-1), was selected The chcita6on used a formal set of procedures, which was for the study. To initate the analysis, PNL conducted the first edapted from the NRC Severe Accident Risk Program capert judgment ehcitanon workshop in May 1990 at Rockville.

(NRC 1989). This paper describes the process and resuhs of the Maryland, to address four selected safety systems at Surry-l. To elicitation. The estunated rupture probabibbes will be used by complete the analysis PNL organized a second workshop, which PNL in an ongoing pilot study on the risk-based appbcanons to was held in February 1992, at Washington, D.C., to address the estabhsh inspecuen pnonties for rystems and components at remazmng safety systems and balance of-plant systems at Surry-1.

nuclear power plants.

The goal of the workshops was to obtain numerical esumates for probabihties of catastrophic or disruptive failures in pressure boundary systems and components in pressurized water reactors

1.0 INTRODUCTION

(PWRs). He Sve systems for the second clicitation were the The goals of the multi-year program Nondestructive Evaluation high-pressure injection, service water, component cochng water, Rehabibty for Inservice inspection of Light Water Reactors, 1

Truong V. Vo

S the residual heat removal, and power conversion systems. This been possible if the discussions considered only a

• generic" paper describes the clicitation process and gives the results of the PWR. Nevertheless, experience at other PWRs was an expert judgment elicitation for these five systems, important consideration at the workshop.

i Based on their knowledge and experience, the expert partici-i pants provided input that was used to assign failure probabilities 2.0 OBJECTIVES OF THE WORKSHOP for use in risk-based calculations and for later use in developing His section outlines the types of information sought from the risk-based inservice inspection plans (Vo et al.1989 and Vo et al.

expertjudgment clicitation. He specific objective was to develop 1990).

numerical estimates for probabilities of catastrophic failures or ruptures for pressure boundary components at the Surry-1 plant.

The components of interest were the pipe segments, which in-3.0 EXPERT CREDENTIALS cluded straight lengths of pipe, elbows, couphngs, fittings, He selected experts needed to have knowledge of the subject flanged joints, and welds. Functional failures of active compo-matter as well as an understanding of the rules governing the nents (e.g., pumps and valves) were excluded from the analysis.

form on which they were to respond. Experts had demonstrated Steam generator, tank, heat exchanger, etc. are considered in expertise by publications, hands-on experience, and managing or other studies. Some important considerations that guided these performing research in areas related to the issues. Experts were estimates were selected to be versatile enough to address several issues and have

• Failures that are not catastrophic in nature (e.g., leaks and sufficient PRA understanding to consider how these issues would cracking) were to be clearly distinguished from catastrophic be used in a risk analysis. Finally, the experts needed to represent failures (e.g., pipe ruptures). Less severe failures were of a wide perspective of the issues and be willing to be elicited interest because they can be viewed as precursor events that under the methodology used. Table I lists the meeting attendees, suggest locations with higher rupture probabilities.

including substantive and normative experts from PNL and the

• Failure probabihties and their uncertainties were requested at NRC observers. Normative experts are those staff members with a very detailed component level, as this information was experience in clicitation techniques. Substantive experts are those needed to pnoritize sampling plans for inservice inspections.

staff members familiar with the technical subjects.

Therefore, estimates for system-level failure probabilities needed to be decomposed to estimate the individual contribu-tions for each of the individual components making up the 4.0 EXPERT JUDGMENT ELICITATION systems.

A systematic procedure has been developed for conducting

• De evaluations nee.ded to be comprehensive to address not clicitations (Vo et al.1991, Wheeler et al.1989, and NRC 1989).

only welds (the traditional approach), but to also address all ne procedure calls for care in enlisting a suitable panel of locations of potential failures, including those base-metal loca-experts, training of expens, preparing the panel to provide nns where stresses and degradation mechanisms could lead to responses to a collection of sell-posed questions, and allowing seucturalintegrity failures.

sufficient time for experts to document the rational. Figure 1 Ti.e estimates of failure probabihties needed to be quantitative shows the expen clicitation procedure.

in nature, while recognizing that there are large uncertainties In this study, the panel of experts provided experience in in nu nerical values. Estimates of these uncertamties were re-structural integnty problems at operating plants as well as an quested (e.g., 5 and 95 percentiles). Quantitative estimates of understanding of response from structural materials to service failure probabilities are necessary for risk-based prioritization, environments. Prior to the workshop, reference materials were because the method is designed to compare priorities on an sent to the experts, including data sources, reports, models, and overall plant basis and, thereby, address a range of systems recent PRA results. Panel members were asked to study these with potentially large differences in their safety consequences materials and to provide irutial estimates of failure probabihties.

of failure.

These estimates would be discussed, refined, and compiled at the It was important to know the reasons that experts had for

meeting, estunatmg specific failure probabilities. ne clicitation process, Training in probability assessment was the first scheduled aetiv-therefore, reviewed failure probabihties in a systematic manner ity for the expert panel. De purpose of this training was to help by identifying potential degradation mechanisms and causes of the experts to better express their knowledge and opinions into a fadure. This process permitted the experts to reconsider their form that could be incorporated into risk models. Training initial estimates and helped PNL in interpreting the information included informing the experts about the methods that would be collected during the clicitation.

ur,ed to process and propagate their subjective beliefs, introducing It was recognized that many of the invited experts did not reg-them to the assessment tools, practicing with the tools, and ularly work in the areas of probability and risk analysis; how-introducing them to the psychological aspects cf probability ever, these experts brought other important knowledge and clicitation. Included in the training were presentations of the experience to the meeting. The clicitation meeting permitted potential biases that might be experienced in this type of meeting.

input from different sources to be integrated into the numerical The clicitation training meeting was designed to anticipate, moni-estimates of failure probabihties.

tor, and address this potential problem and to revise the process Consideration of Surry-1 as a sample plant served to focus the as needed. Initial documentation of the experts

• judgments and meeting at a greater level of detail than would otherwise have 2

Truong V. Vo

e w.

TABLEI. LIST OF EXPERT ATTENDEES

)

Area of Expertise Oraanization i

Name i

f d

Expert Panel Structures / Fracture Mech.

Westinghouse l

Bishop, Bruce Structures /PRA f

Rolls Royce & Associates Chapman, Vic PRA/ Reactor Design University of Missouri-Rolla Edwards, Ray Operating History / Data Base U.S. Department of Energy l

Jamale, Kamiar PRA/ Structures Nuclear Regulatory Commission t

O'Brien, John Reactor Systems /PRA f

Idaho National Eng. Lab.

Phillips, Jerry ISI/ Reactor Systems l

Martin Marietta Energy Puthns, Steve Structures / Service Exp.

Rodabaugh, Everett Battelle Columbus f

Materials / Structures Shewman, Paul Ohio State University /ACRS 1

i Observers Statistics / Expert Bicit.

Nuclear Regulatory Commission Abramson, Lee Structures / Fracture Mechanics ABB Combustion Engineering Ayres, David ISI/ System / Materials Virginia Bectric and Power Co.

McNeil, Alex Decision Analysis / Expert Judgment Westinghouse Sectric Corporation i

Perdue, Bob i

PNL Normative / Substantive Experts Reactor Systems /PRA I

Pacific Northwest Laboratory Gore, Bryan Fracture Mechanics Simonen, Fred Pacific Northwest Laboratory PRA/ Reactor Systems / Aging Res.

Pacific Northwest Laboratory I

Vo, Truong e

t i

l Selection of Selection Fam.hanzation i

1

issues,

---c-7 of Expens of issues Parameters i

Elict:ation Recomposition Review by

+

cf Experts and Aggrepation +

Experts and I

and Training of Resutts Documentation f

FIGURE 1. EXPERT JUDGMENT PROCEDURES in an unambiguous manner to avoid the poteada! for precondsnon-their suppornng reasomng were obtained during this traumng ed or biased responses.

%e pmsentations were followed by discussion sessions. The session.

To address the issues from many viewpoints, the clicitation of A= Man of each issue invohed the experts, the observers, and the experts was desi ned as a formal, face-to-face meetag. Forprojectteam analysts (a team composed of normative and substan-F each system addressed, a formal presemation was provided. Pre-tive analysts). Knowledge from the experts regarding plant design sentanons covered technical descriptions, historical co iv ;

and operanon, failure history,insterial degradation mechanisms, failure mechanisms, clicitation statements, suggested approaches, dtion and aggreFanon of the data, etc. was brought into and questionnaire forms. %e issues were presented to the expertsi%

3 Truong V. Vo

._.. _, ~.

4

A

• " r m

~..

k d to use this information to revise l

involved %e pane was as e the discussions. Because the process was designed to take advant-their original estimates as needed. The revised individual judg-i age of the diversity of the knowledge, experts were encouraged ments were once again recomposed and aggregated to provide a to provide their own estunations, and an effort was not made to single, final composite judgment for each issue.

seek a consensus on the est mated rupture probabihties.

Each panel member developed rupture frequency (or proba-bihty) estunates, which incorporated his opiruons regarding the 5.0 RESULTS AND DISCUSSIONS uncertainty of the estunate for pipe rupture. For most compo-De results provided by the expert panel members were com-nents, panel members considered the fact that no ruptures have piled into distributions.%c distributions were then used to deter-j t

occurred, but information from the hterature and discussions ses-mine a best estunate of rupture probability for each component.

l sion did suggest to them that certain failure modes were possible.

his section presents a summary of this data and discusses the l

Data exist regardmg the structural degradation of welds or piping estunates in terms of Surry-1 plant operanng experience.

due to erosion / corrosion, fatigue, etc., but several experts ex-i pressed their belief that other mechanisms could contnbute to 5.1 Method of Recomposition and Accreaation of Results f

%e best estimates obtained from the population of experts were l

future failures.

Each expert then completed questionnaire forms that addressed summartzed in a series of plots (i.e., boxes and whiskers). %ese locanon-speci5c rupture probabih6es for the systems ofinterest.plots represent the distribution of esumated failure (rupture) prob-L Dis data covered best estunates of probabitioes, uncertainty esti-abibties provided by the experts. An individual plot graphically motes, and the ranonale for these W'~ De quesuonnaire displays the following features of the distribution.

forms for each issue were collected at the end of each session. In-

• De *whisken* idenufy the extreme upper and lower values of formation that was used to obtain the destred inputs from the ex-the distribr 'on.

perts is shown in Figure 2.

• De box uself locates the 25% and 75% quantiles (i.e.,

Following the cheita6on meeting, the informszion provided by quartiles) of the distribution. In other words,50% of the data the expert panel was recomposed and aggre;4ted. A wricen an-points are within the box.

olysis of each issue includmg the initial recomposition, addruonal

• He line within the box intersectmg with the circle or dot is the plant-specific data, and other relevant information was then re-rnedian of the distribution.

turned to each expert panel members for review. De purpose of he 25 % and 75% quartiles were chosen for uncertamty esti-this review was to provide experts sith the opportunity to revise esates to ehmmate the two most extreme values of the their earlier assessments, to ensure that potential misunder-distribution. Figures 3 shows an example of tbe failure probabibry standmgs were identiSed and resolved, and to ensure that the esumates for components within the residual heat removal system documentation correctly reflected the judgment of the experts i

I PRA results and Data from other relevant information Wstoncat fracture rnechanics (system, component priontization.

j failure data analyses system desenptions. etc.)

j I

9P 1r 1

1r Additional inforrnst on Eroert judgment I

ehcrtstron and h

(acditional plant-specific information, etc.)

discussion 1

l l

V i

Estimated rupture J

probaDikties i

FIGURE 2. INFORMATION USED TO SUPPORT EXPERT PANEL 4

Truong V. Vo

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

.-,,-_i.m.,.

w,

.~y,

,.-1

.~ -.

i

}

j l

^

1 l

i

-12

-11

-10

-9

-e

-7

-6

-5

-4

-12

-11

-10

-9

-e

-7

-6

-s

-4 4

1 l

i i

i i

4 i

1. Loop 1 Holleg to WW1700 e4. mm.t m eevices g

d j

es. stvisobe.es a

2. WoV1700 to WDv1701 ap i

4 f

3. M3Vl7D1 to hoes 0 qp

**"8*

• N '

4>

f 4>

.g.m..m 4 tesoe O to 1 -RH-8 4p E "5D0* * ' **-*

18 1-9-139 m 6sep t qp S. 1-AM-8tel-RM-P-1A

-p to esses am t-e+-te qp y

s.1-RM-Riot-RM-11 o

)

28 '-e*3etomoso qi 7.1-RM-11 tel-R,4-12 I

1 SL W W el-CS-B

--O 8.1-RM-12 to hode l p

st. 8-45-3e ee t-O-t->

d>

9. hode1(HXloop) tones-15

-p l

'~

10.1 -#M-15 to t-RM-t-1 A si s.<p-E-aam e O-es e

11,1 -RM-t-1 A to 1 -R**-19 h

361 es estee-cs-me q,

35. 9-EE-ee to1-05-#-t qp 12.1-st>6-19 to HL'Y 1758

26. sesse e ne s>s 721 q-

13. NL"Y 1758 to 8eoes 2 o

I r

i f

1 t

t 1

1 i

t 1

e t

n

-12

-11

-10

-e

-e

-7

-s

-s

-12

-it

-10

-s

-e

-7

-s

-s

-d Log (10)(foilure/yr)

Log (10)(follure/yr)

{

FIGURE 3. FAILURE PROBABILITY ESTIMATES FOR THE RESIDUAL HEAT REMOVA1.

i SYSTEM COMPONENTS l

cs Suny-1. For martability, the probabihties are presented on a in agreement. De component medians generally varied within a i

I log 10 scale, with the probabihties expressed as failures per year, factor of 10 to 1000 within the systems studied.

Table 2 shows the numerical failure probabihty esnmar= for componems within the residual beat removal system. This type of 5.2 Ra-Idaaf Heat Removat System For the residual heat removal system, abe highest failure

-}

plot was produced for all 6ve systems at Surry-1.

j As indvateA in the following 6gures (e.g., Fogire 3), expens probabihty estimates were those pipe segments within residual provided a wide range of iv.fw regarding failure probabsb-heat removal heat exchanger discharge lines to reactor coolant l

i ties, and this trend is entirely consistent with the large uncer-loops (cold legs). Results for this system are shown by Figure 3.

tainties associated with the performance of the components being

,.The high failure estunates were primarily due to the high addressed. In 6 yy the range of iww, there was no operatmg stresses and occurrences of various fatigue failure i

anempt to seek consensus from the espert panet Rather, the mechamsms. Iower failure probabahties were estimated for the responses were evaluated in a statistical eense, with a central pipe segments within pump suction imes, including ponaons l

tendency being established as a single best esnmate probabihty for connectag to the hot les reactor coolant loop.

l Based on information reported by the NRC's Office for Analy-l use in the risk-based studies.

The meAan failure probabihty estimates for components within sis and Evaluation of Operational Data, thenna! strati 6 cation has the systems studied vary between 1.00E-06 and 1.00E-02 failures caused cracks and damaged supports and has contributed to ther-i per year dependmg on the system, components within the system, mal fatigue in the residual heat removal system piping. The report and component locations. For a given componerit within the sys>-

also :sted that through-wall cracks were generally found in the

{

sems, the quartile range generally had a variation between a factor 90 degree elbow of the pipe section coritamn.r strati 6ed flu;d.

I of 10 to 100, indicating that most experts' estimates were Pipe cracks have also been described in the report published by i

Truong V. Vo j

i

--,-__n

,,,-n.~,

g TABLE 2. MEDIAN AND QUARTILE ESTIMATES FOR THE RESIDUAL HEAT REMOVAL SYSTEM COMPONENTS Failure Probability Percentile Component Median 25 75 1.

Loop 1 HL to MOV 1700 1.60E-06 9.47E-07 1.OOE-05 2.

MOV 1700 to MOV 1701 2.00E-06 8.91 E-07 4.45E-06 3.

MOV 1701 to Node O 4.OOE-06 1.48E-06 1.12E 05 4.

Node O to Valve 1-RH-8 5.00E-06 1.78E-06 1.00E-05 5.

Valve 1-RH-8 to Pump 1 A 1.20E-06 1.OOE-06 5.61 E-06 6.

Pump 1 A to Valve 1-RH-11 2.00E-06 1.00E-06 1.12 E-05 7.

Valve 1-RH-11 to 1-RH-12 1.40E-06 1.00E-06 9.87E-06 8.

Valve 1-RH-12 to Node 1 4.00E-06 2.00E-06 2.29 E-05 9.

Node 1 to Valve 1-RH-15 2.00E-06 1.88E-06 1.79 E-05 10.

Valve 1-RH-15 to RHR Hx 1.00E-06 9.93E-07 5.61 E-06 11.

RHR Hx to Valve 1-RH-19 1.00E-06 6.59E-07 5.61 E-06 12.

Valve 1-RH-19 to HCV 1758 4.00E-06 1.59E-06 1.41 E-05 13.

Valve HCV 1758 to Node 2 1.OOE-06 6.59E 07 4.4 5E-06 14.

Node 1 to Valve FCV 1605 1.40E-06 6.81 E-07 6.30E-06 15.

Valve FCV 1605 to Node 2 2.00E-06 6.81 E-07 9.87E-06 16.

Node 2 to MOV 1720A 5.OCE-06 2.83E-OG 2.71 E-05 17.

MOV 1720A to Valve 1-SI-130 2.00E-06 8.75E-07 1.19 E-05 18.

Valve 1-SI-130 to Loop 2 CL 1.00E-06 6.59E-07 5.61 E-06 13.

Node 3 to Valve 1-RH-29 2.00E-06 6.81 E-07 5.61 E-06 20.

Valve 1-RH-29 to MOV 100 3.00E-06 1.4 6E-06 4.00E-05 21.

MOV 100 to Valve 1-CS-35 6.00E-06 3.12E-06 4.*J1 E-05 22.

Valve 1-CS-35 to 1-CD-E2A 1.20E-06 6.81 E-07 5.61 E-06 23, 1-CD-E2A to Valve 1-CS-46 1.20E-06 6.81 E-07 5.61 E-06 24.

Valve 1-CS-46 to 1-CS-48 2.40E-06 1.21 E-06 2.03E-05 25.

Valve 1-CS-48 to RWST 2.00E-06 6.81 E-07 9.87 E-06 26.

Node 4 to Valve RV-1721 4.00E-06 6.81 E-07 4.4 5E-06 the American Society of Mechanical Engineers (ASME) Sec-High failure probabilities were also estimated for the reactor tion XI Task Group on Fatigue in Operating Plants (ASM E 1990).

coolant pump seal injection lines. Fatigue-related problems, sub-Numerous cases of pipe leakage have also been reported, where component weld failure, and valve leakage and failures have been fatigue-related problems have been the principal contributors to experienced at operating PWRs. Relatively high failure probabil-leakage. Other causes include cavitation, water hammer, and ities were also estimated for charging pump discharge lines.

improper installation.

Vibrational fatigue mechanisms have been experienced in these pipe segments at Surry-1 and cracked welds and pinhole leaks 5.3 Hioh Pressure Iniection System have been reported at cther operating PWRs. l ower failure for the high-pressure injection system shown in Figure 4, the probabilities were estimated for pipe segments within the normal highest failure probability estimates were for those pipe segments charging lines and pump suction lines.

within the cold and hot leg injection lines extending from the con-tainment isolation valves to the reactor coolant system loops.

5.4 Power Conversion System These high estimates were primarily due to the high operating For the power conversion system shown in Figure 5, the high stresses and the occurrence of various fatigue failure mechanisms.

estimates of failure probabilities were for pipe segments within Cracked welds and large leakage events within these lines due to the feedwater turbine. driven pump, main feedwater discharge high-cycle thermal fatigue have been reported at operating PWRs lines, moisture separators, and feedwater heaters. These problems (ASME 1990).

6 Truong V. Yo r,-e--

,,-n---

~-

,L

.i t

w.

w.

.2 1

4 g Qt 1 =

I t

i q4 p_

Hf F

l t f--

g 3

4 t_

..u.

I p.,g ib I

l 4 f-f F-

~

~

i -

1 4 p

' Q'

+

F-

.a

_i 4 t.

.=.

T gy_

rw 3

o

..~

3

~'

.a

.s Lego o)paaww/yr)

Logo 0) pow./yr)_ _ _ _.

-. ~

-8 l 4 t-3 l +

r.-

3 l

4F c3

}

• g,g*

+f

.. +... -.. _

E3]

a-}.

1 o}

E}-

-:=-

EI-r-T1-i

~

=ma.-

+ F-r --.'

8' :",,"-"

d l

4 t-H

+F

....-w l

4 F-l l 4 F-l

+ t-W 3

l +

t-

.a 5

. LogD C)Cowe/r). -.

LogD.C) Cows /F)-

FIGURE 4. FAILURE PROBABluTY ESTIMATES FOR THE HIGH PRESSURE INJECTION SYSTEM COMPONENTS J

7 Truong V. Vo

an.Res.

1 l...... -.

I

=-

'~n-333 a

.... ~,

• F-.

O y 4p

=-

[D-

~=-

3

• =-

ED-

~.: ::-

3 S

=.,.

3

-+

t-3 l

++-

D 3

l m-D-

3

}

4 t.

1}_

==

0

~ ::-

Df-

~=-

33

~==

+F 3

I 4 F-

=; ;;.

}

L.90 0)(rowe/r)

L.00 0)(f a.we/r)

FIGURE 4. (CONTINUED) resulted from various failure

d. A-such as water hammen, probabilitics, varying within a factor of 2 to 5, dependmg upon erosion, corrosion, and fatigue.

component locations.

Based on the Nuclear Plant Reliability Data System (NPRDS)

Based on the NPRDS information, there have been two rupture data, there has been one rupture event reported in the PWR feed-events reported in the service water systems at operating PWRs.

water system in approximately 500 plant years of operuzion. %e These ruptures were attributed to wear and possible vibrational 16-inch ruptured pipe was located on the main feedwater pump fatigue. At least 50 leakage events have been reported at operas-suction and the cause of the failure was crosion/ corrosion. Prpe ing PWRs. Erosion /corrosionis the principal contributor to leak.

cracks have also been reported in the feedwater systems at operat-age for service water system piping in PWRs. Other causes for ing PWRs. At least 12 pipe cracks have been reported, located leakage include wear, cavitation-induced corrosion, and water predominantly at prpe elbows and reducers. Fatigue related prob.

hammer lems and corrosion are believed to be the principal contributors to cracks in the mam feedwater piping.

5.6 Cornoonent Coolina Water System According to the report published by the ASME Section XI For the component cooling water system shown in Figure 7 Task Group on Fatigue in Operstmg Plants (ASME 1990), at least the most important pipe segments with respect to rupture proba-19 plants have found cracks or defects in the vicinity of the steam bihty are the lines that provide coolmg to the residual heat generator feedwater nozzles. ne cracks were located predom-removal system heat exchangers. High failure probabihties were ir antly in the counterbore region of the weld preparanon, away esumated for the pipe segments within the pump discharge hnes, from the beat-affected zone; at least two cracks were found to be and pipe segments that provide cooling to the reactor coolant j

through-wall and leaking. Corrosion was found to be a contribut.

pump seal were also emmated to have high failure probabibues.

i l

ing factor in assisung the thermal fatigue in most the cracking Several pipe leakage events have been reported wsthm the com-incidents.

ponent cooling water systems at operanng PWRs. Waier hammer-related problems are the pranetpal contributors to pipe leakage.

i 5.5 Service Water System j

Within the service sater system shownin Figure 6, the highest rupture probabihnes were estunated for pipe segments extendmg 6.0

SUMMARY

AND CONCLUSIONS from the condensers, to the component cooling water beat

%e procedures used for the expertjudgement clicitation were exchangers, and to the rectreulation spray heat exchangers. %e adapted from the recent methods developed for the NUREG-1150 high failure esnm.ree were muributed to various failure PRA studies (NRC 1989 and Wheeler et al.1989). These mea.ams such as wear, corrosion, and erosion. High failure methods were modzSed by PNL, based on staff knowledge and probabilities were also esumated for pipe segments within the expenencein the area of expertjudgment clicitation gained from circia.rint water pumps and the bearing cooling water heat ex-recent workshops performed by other national laboratories.

changers. Erosion /corrosionis the major cause of failure for these Generally, the cooperation from industry representatives was very pipe segments. The remammg pipe segments within the service good, and participationin the discussions was very forthright and water system have relauvely the same esumated failure E

Truong V. Vo l

c,,

s 1

A m.

I w.

i i

1

.n

.... u

....n

.u

.n

.u

.n

..u

. *. ---,;-..-+-= ~ D.

~..~ e

{.

O. -

r. ;.

.--~-.*~.-{.

~~

rm

}.

~..

C

' '-*-!:Z;~ --4 4

l

. * =..,

-H 4 l r7 p 4

l

+ ;

i,....,.~.

i l4 ;

i' l

.i

_q 4g r7- -]

_4 4;

=;-

y p

' ' : **~~ ~

+

a,.

W

-4

, F--

{

..z....

C

= p g -a-.~.

--i+

l--.

~

..u

.u

-u

.n

-u

.o

.a

.u.o

.u

-u a.a

.u.a

.u

-..u -u.a.u Logo o)(toe./r) is900)(fowe/r)

......u

.u

.u

.o

.....o

.u

.o

.. ~.,

_q.

j h

.. ~..

..'-....~_4

=.

i 4.

t-

_(

4,

... -.... ~,

44

~

__.4 ;

..,. ~

_q.

..---...~.i 4

.,~

_4

-i +;

a :

-.f T '

.. ~ ~.. ~. q

__q 4 r

..... ~,

-i.;

--i 4 q.

.g

. -.... ~.

.. ~.

_ fT' j

--f I'

.... ~,,

.. ~.. ~,

-i + ;

- -L,_.j

.... ~

.. ~....

= ;,~,.., --

W

... ~,

C.

4 e.

. u -u -u -u.u -u

.u ou u

.u u -u.u.u.u.u -u -u

-u -u

.u tagn o)(vowe/y)

__ __ _ _ LogDC)(foiwe/r)

FIGURE 5. FAILURE PROBABILITY ESTIMATES FOR THE POWER CONVERSION SYSTEM COMPONENTS 9

Truong V. Vo

,i---.,,e,am.-

'se%'v-

.ea we

A m

............-.u

,.. -..... ~

~.

,, _. _.. _ 4.;

' ~

-.. -m

 ;-~.:.='~

-i.

-i.

. 7; y.

-. +

... ~

g

-_+

... - ~... - -

-4 _ :

~.

"***'7.-*""

C

= ;g.y..

4 J

E

....--.-.i.

i i

. ~ ::-:

.m..

.a.---Ii r

c.. e.

o M,

e*-*-*-8='-*-*

....3 4

M

-fT'.

c-.-

a

• .- o L

= ' - * - * * * * + *

--H *

-- i.

"'-*-*='-*-*"*~~'d*I

- -.i.

...a...

• • -**~~-'-*---fT l

-i.

i.

-.H.

l

...u.a

.a

.,a u

.a

.u

.u.a

.a

..n a,

.<.-,'.s.

-u -"

-a

~^

togo c) one./yr) to90 0){aw/r')

FIGURE 5. (CONTINUED)

~

belpful. Their opinions appeared to be unbiased and (e.g., rupture probabilities of the component cooling and service knowledgeable.

water systems at some plants). %e expert opinion process used Training signi5cantly helped the experts to better encode their in this study has established a procedure that can be used to knowledge and beliefs into a form that can be incorporated into address such events.

PRA models. It is recommended that at least one day of formal tram.ng should be included in all future workshops.

An enormous amount of technicalinformation was gathered re-ACKNOWLEDGMENTS gardmg the elicited issues. Knowledge and understanding of the his work was supported by the U.S. Nuclear Regulatory issues were signi5cantly enhanced by the exchanges ofinforma-Commission (NRC) under a Related Services Agreement with the tion between experts and observers at the meeting. Informa: ion U.S. Department of Energy under contract DE-AC06-76RLO gamed from the expert judgment clicitation greatly enhanced the 1830. The authors wish to acknowledge the direction and support cred.bihty of the Surry-1 plant analyses. Other notable beneSts provided by Dr. Joe Muscara. NRC Program Manager and Dr.

from the expert judgment include the enhancement of the risk-S. R. Doctor, PNL Program Manager. Acknowledgements are based methodology for estunanng parameters for specific fauh made to Virginia Electric Power Company staff for their partici-events and guidance for incorporating innovative recovery actions pation in this work. A special acknowledgementis given ta all of in support of a number of the NRC's regulatory programs.

the expert panel members and observers sto attended the in summary, the data provided by the expert panel appeared to elicitanon meetmg.

be rear.onable. he esumates were generally agreed with and reflected Surry-1 plant operanng experience. Typical areas of concern correspond to such factors as high stresses (e.g. places 7.0. REFERENCES stcre mMng of fluit with Ir.rge tenerature differersces occur)

Americsm Society of Mech.mical Enginre (ASME).1990. Mc ts!

l

_ and places where fangue mechanisms and erosion / corrosion Fatirue in Oreratmc Nuclear Power Plants. ASME Section XI mechanisms are susceptible. These esumated resuhs are being

. Task Group on Fatigue in Operanng Plants.

used in an ongoing pilot study which is based on the PRA resuhs

~

and other plant specific informanon. The expert judgment panel U.S. Nuclear Regulatory Comrmuion (NRC).1975. Resetor generated information in other areas that also enhance the plant Safety Studv-An Assessment of Accident Risks in U.S. Com-cnalyses. Previous PRA methods would normally apply compo-mercial Nuclear Power Plants. WASH-1400, NUREG-75/014, nent rupture probabilities from WASH-1400 (NRC 1975) to quan-Washington, D.C.

tify the core damage frequency. The expert panel elicitation for this study has produced a act of component rupture probabihties U.S. Nuclear Regulatory Commission (NRC).1989. Severe i

for various components within selected systems at Surry-1 as a Accident Risks An Apenment for Five U. S. Nuclear Power function of location and failure mechamsms such as stress, Plants. NUREG-1150, Nuclear Regulatory Commasion, thermal fatigue, etc. %ere have been long-standmg issues regard-War.hington, D.C ing how to model spine failure features oflight water tractors 10 Truong V. Vo

=

?

e e.

f N

=W"W,

- =.

i I

e d.

-a

--->-s

,,,,..,....s

.s e-m.

.si.

.{

4j D.

q

. '. + - =. -. + * -. -

4

,p 4

41 3-

' ' ' + * ~ * - * * " '

p ED-

= * + " - -

F g._

4y g)

. = = = ~

n__

9-4 ED-c31 E}-

c31 E}-

r w

ED-g}.

I s on e *

.s

.e

.a

.s

.s n.900)4

/r) 4.seOC)D e/r) s y

g rn_

==:;1 73_.

g3_._

g>_

g}_

g__

.. - - -. ~ ~

=.:~::=~.:1 0}-

= = :.~. : =

9--

...-s.

.m

.w y

E}-

, J}-

.w J

a.

....s E. } -

q 4,

-... ~

t 4 l l

.u *- a.s...u-m 9-i

• F

..-a--..-a-~

D---

-.+- ;;--

l 4 F l

}

• ., = ;;-. t ",;r

}

. w

.r 3--

-r n

.. _ ~.. _.

.w

.e 4

.s

.a

..a i.

.a

'**U CI U**#'/F)

A.,rso)e

./r)

FIGURE 6. FAILURE PROBABILITY ESTIMATES FOR THE SERVICE WATER SYSTEM COMPONENTS 4

h e

e Il Truong V. Vo ppr-i--i--,,

,p r,

ivy-p y-*gg-9 q y,--

-yiyym y-w -

--n y g rivy gy,g+

arywy-p*y,-4--'wa-T.

'. se meg y-gM w 9-y

-y-h

l l

I

.s

.s

...a.e-

) fL

]

e s

w n..s>

{L.-

...n.-

I g

...a i

{

u s

m g

g)_

j 3

E}-

i i

.is

.o a.

r LeeOc) Cow./yr)

{

FIGURE 6. (CONTINUED)

I Vo. T. V., B. F. Gore, LJ. Eschbach, and F. A. Simonen Vo, T. V., B. W. Smith, and F. A. Simonen.1990.

• Develop-l 1989. 'Probabilistic Risk Assessment-Based Guidance for Piping ment of Generic Inservice Inspection Guidance for Pressure Inservice Inspection.' Nuclear Technolory, Volume 88, Num-Boundary Systems.* Nuclear Technotory, Volume 92, Num-ber 1. American Nuclear Society, La Grange Park, Illinois.

ber 3. American Nuclear Society, b Grange Park, Illinois.

l r

Vo, T. V., P. G. Heasler, S. R. Doctor, F. A. Simonen, and B.

Wheeler, T. A., S. C. Hora, W. R. Cramond, and S. D. Unwin.

F. Gore.1991. *Esumates of Rupture Probabilities for Nuclear 1989. Analysis of Core Damare Freauenev from Internal Events:

PowerPlantComponena: ExpertJudgmentElicitation.' Nuclear Ewert Judement Elicitation. NUREG/CR-4550, Volume 2.

l Technoincy, Volume 96, Number 3. American Nuclear Society, Sandia National bboratories, Albuquerque, New Mexico.

La Grange Park, Illinois.

l k

12 Truong V. Yo

s.:

m.

... +....

4 4 p

q.

p

.c.u...

{ 4 l

l., p

~ ~.-=--*

--i I.

t-

,4 F

-.I e

}

i t-l

+

t-4 p

44 t

" l.a,,~,,

l 9

F-I +

F l

4 F-

-4 +

F-l 4

t-4p 4

t

---I +

t-

-l 4

f-a.-.=.-=-.=

-4 4

F-

-1

+

t-t

,q 4 g

.q 4

p.-

-4 4

F--

-i 4

F-

........ +.

-4 4

t-

= ~. -.

-4 4

t--

L g'$0)(i w./yr) i test 10)(.omw./pr) 4 q

.p

_4 4 p t-F._

i H

{

. =...- -- *,

i i -=-=. --.

p

_q y

p i

a

..e --.

m M

t-t.

u...,.a r

p t-

- ~

t.-

l

-=-. -- -

E i

J

=. ~.. ~ -

e t-l 4

H i

4 l

4 p

4 p

l

+

F

+

p F

=.

{ 4 p

l i

l F

-.I 4 h-

= -...

...a.

F l

-1

+

t-t

.a. - _

H.

.{

4 p

. - ~ -

F-

-i t-i

.a

.n

-..ag40)(e ir./yr) sag (10)(Doewe/yr)

FIGURE 7. FAILURE PROBABILITY ESTIMATES FOR THE COMPONENT COOLING WATER SYSTEM COMPONENTS a

e d

n I

i 13 1

Truong V. Yo

-+y.#,-,-

v.,,

,,,.,,,.w-...,y.,

c-

.gm,.

,,,,,.,54

,,--.-,,.,,',,ww,._,.,

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

e +*

e W..,

.3

..e 4

_e

.e

.a

s...a is...a.=>

y p

r

.2 i

p

...a y

i t

F-

-4 e

f-I

+

_e_..

d h

..a r_

i ;.

p

..u.i a.

l i

..c g.:

w I

4 L

.cc..s....u 1

I I

..c 4..

f---

-u u

s.==

...w...

l 4 p

{'

F'

.s..a...

.e i

l 4

F

=.

n~.,

u

.c.e s..

4

.a

.e...a

.e H-

.... ~..

..a a.

4 4 p

m q

4 p_

r

.e.......

... _.. ~

r_

H a, [

-m -r --

1 g

a

..e e

.t

.e

.s

.a

.s 2

.s Log (10)(lodure/yr) ang(to)(s wre/pr)

FIGURE 7. (CONTINUED) 6 14 Truong V. Vo

i l

Type of ECCS Pumps Strainer Strainer

NPSH, NPSHg Plant Insulation in Pumps per Area Suction (ft)

(ft) z Drywell (1)

Strainer (ft )

Flow (2) l (gpm)

Millstone 1 Mineral Wool RHR-4 6

69.6 27200 33 28 i

CS-2 Monticello Nukon RHR-4 6

40.1 22040 31 28 CS-2 l

Nine Mile 1 Reflective Metal some Fiberglass Nine Mile 2 Reflective Metal limited Fibrous Oyster Creek Nukon LPCS-4 8

58.1 36000 32.8 23 limited Mineral CS-4 Wool Peach Bottom Nukon RHR-4 1

34.6 10000 CS-4 1

26.1 3125 Perry Primarily Nukon RHR-3 1

42.2 7000 3.4 0.5 Reflective metal on RPV LPCS-1 1

42.2 6600 3

1.7 Pilgrim Nukon RHR-4 1

13.3 5250 30 CS-2 1

13.3 3600 Quad Cities Nukon River Bend Fiberglass RHR-3 1

21.2 4970 5.0 1.3 Reflective Metal LPCS-1 1

21.2 5500 5.76 1.3 i

Susquehanna Nukon RHR-4 1

43.9 10650 12 5

Reflective Metal Fiberglass CS-4 2

19.6 6350 12 2

Vermont Nukon RHR-4 2

47.5 14400 Yankee some Fiberglass CS-2 1

10.2 3000 WNP 2 Reflective Metal RHR-3 1

19.4 7175 36 12 e g ass LPCS-1 1

19.4 7214 36 10 NOTES: (1) RHR = LPSI and/or containment spray pumps:

CS = Core spray pumps:

LPCS = Low pressure core spray j

(2) NPSH, = available net positive suction head: NPSH, = required NPSH.

(3)

Information which is currently unavailable is indicated by a dash.

4 i

BWR THERMAL INSULATION TYPES AND ECCS CHARACTERISTICS t

Type tf ECCS Pumps Strainer Strainer

NPSH, NPSH, Plant Insulation in Pumps per Area Suction (ft)

(ft) 2 Drywell (1)

Strainer (ft )

Flow (2)

(gpm)

Big Rock Nukon RHR-2 2

3.4 800 Point (3)

Browns Ferry Reflective Metal 1.2.3 Brunswick 1.2 Nukon RHR-4 2

32.2 15400 15 Reflective Metal CS-2 1

15.7 4625 12 Clinton Reflective Metal Cooper Reflective Metal Dresden 2.3 Reflective Metal limited Fibrous Duane Arnold Nukon RHR-4 2

14.6 9600 24 CS-2 1

4.21 3020 32 Fermi 2 Reflective Metal some Fiberglass Fitzpatrick 80% Nukon RHR-4 2

26.7 20800 20% Mineral Wool CS-2 1

11.9 4725 Grand Gulf Reflective Metal some Fiberglass Hatch 1 Nukon RHR-4 1

29.3 8400 28 24 CS-2 1

17.3 4625 3

Hatch 2 Nukon RHR-4 1

29.3 8400 10.2 6.25 CS-2 1

20.9 4625 8.6 7.3 Hope Creek Nukon RHR-4 1

14.7 10000 6.02 4.5 CS-4 1

5.1 3715 11.2 10 La Salle 1.2 Reflective Metal Limerick 1.2 Nukon RHR-4 1

27.7 10000 17.5 5

some Fiberglass CS-4 1

13.3 3715

.-.