ML20038C425

Board Notification BN-81-40 Re Integrity of Reactor Pressure Vessels When Subj to Thermal Shock & Subsequent Repressurization During Overcooling Transient.Forwards NRC Re ORNL Rept on Pressurized Thermal Shock

ML20038C425
Person / Time
Site:	Oconee, Rancho Seco
Issue date:	11/16/1981
From:	Novak T Office of Nuclear Reactor Regulation
To:	Atomic Safety and Licensing Board Panel
Shared Package
ML19252B719	List: ML19252B718 ML20038C425
References
TASK-AS, TASK-BN-81-40 BN--81-40, BN-81-40, NUDOCS 8112110019
Download: ML20038C425 (3)

Text

NOVECER C 1331

).,

C

,I DISTRIBUTION:

NSIC MJollensten-4 i,.

Docket File ORB #4 Rdg MCutchin s

NRC PDR ORB #4 Memo File JStolz S

T%

N s

L PDR ECase ACRS-15 j

O n1 3

g D@d

\\

TERA DEisenhut d'iray

~f O

IP TNovak MPadovan S~

Docket flo. 50-312 DSnyder GVissing M

.(p u es RIngram gj

!;Ei10RA!!DU!! FOR: Atonic Safety and Licensing Board FR0:!:

Thomas !!. flovak, Assistant Director for Operating Reactors, Division of Licensing, NRR SUCJECT:

BOARD f 0TIFICATI0ft - RAllCit0 SEC0 !! EARING (BN-SI-40 )

This notification concerns the integrity of reactor pressure vessels when subjected to thermal shock and subsequent repressurization during an over-cooling transient. This is a follow up to the Board i;otification (BN-81-30) dated October 19, 1981 concerning the same subject.

Enclosed is a copy of the October 30, 1981 letter fron Willian J. Dircks (EDO) to the Comission concerning "The Staff Review of the ORflL Report on Pressurized Thornal Shock". This memorandum enclosed a report by the NRC staff on the NRC staff assessment of the Draft Interin Report by Oak Ridge National Laboratory Entitled " Evaluation of the Threat to PilR Vessel Integrity Posed by Pressurized Themal Shock" and a copy of Duke Power Company's letter dated October 20, 1981, to Mr. R. fl. Bernero, NRC, Olch also prsvides an assessnent of the ORNL report.

The staff has previously concluded and discussed with the Connission that the probability of occurrence of severe pressurized overcooling transients is sufficiently low that imediate corrective action is not warranted but that corrective actions may be required for some plants within a year.

In the staff's judgnent the ORNL reporc does not preser.t any significant new infomation that would change that conclusion.

Original tiyed by Thonas M. flovak, Assistant Director for Operating Reactors Division of Licensing, t!RR Enclosures :

1.

10/30/81 l'ern 2.

10/20/81 DPC Ltr.

8112110019 811116 PDR ADOCK 05000312 P

PDR l

ORB #4:DL C$R DL L

D-ORB #4jDL orric,,

.6...iN94ch. PNh.b.J.Sbb

.TR.'

81 11/.....81

.t s enhu.t..

sumur b l

1

/81 II/m /81 41/g.81 11/

OFFICIAL RECORD COPY usw nei--mm nac ronu 2.a o >8ci.wcu cm

~l

t.

Sacramento Municipal Utility Rancho Seco, Docket No. 50-312 District cc w/ enclosure (s):

David S. Kaplan, Secretary and Christopher Ellison, Esq.

General Counsel Dian Grueuich, Esq.

6201 S Street California Energy Commission P. 0. Box 15830 1111 Howe Averue Sacramento, Galifornia 95813 Sacramento, California 95825 Sacramento County Ms. Eleanor Schwart:

Board of Supervisors California State Office 827 7th Street, Room 424 600 Pennsylvania Avenue, S.E., Rm. 201 Sacramento, California 95814 Washington, D. C.

20003 Business and Municipal Department Docketing and Service Section Sacramento City-County Library Office of the Secretary 828 I Street U.S. Nuclear Regulatory Commission Sacramento, California 95814 Washington, D. C.

20555 Resident Inspcctor/ Rancho Seco c/o U. S. N. R. C.

14410 Twin Cities Road Herald, CA 95633 Dr. Richard F. Cole Atomic Safety a Licensing Board Panel U.S. Nuclear Regulatory Commission Washington, D. C.

20555 Regional Radiation F.epresentative EPA Reaion IX Mr. Frederick J. Shon 215 Fremont Street Atomic Safety and Licensing Board San Francisco, California 94111 Panel U.S. Nuclear Regulatory Commission Mr. Robert B. Borsum Washington, D. C.

20555 Babcock & Wilcox Nuclear Power Generation Division Elizabeth S. Bowers, Esq.

Suite 220, 7910 Woodmont Avenue Chairman, Atomic Safety and Bethesda, Maryland 20814 Licensing Board Panel U.S. Nuclear Regulatory Commission Thomas Baxter, Esq.

Washington, D. C.

20555 Shaw, Pittman, Potts & Trowbridge 1800 M Street, N.W.

,.oshington, D. C.

20036 Herbert H. Brown, Esq.

Lawrence Coe Lancher, Eso.

Hill, Christopher and Phillips, P.C.

1900 M Street, N.W.

Atomic Safety and Licensing Board Washington, D. C.

20036 Panel U.S. Nuclear Regulatory Conmission Helen Hubbard Washington, D. C.

20555 P. O. Box 63 Sunol, California 94586 1

f.

a i

Sacramento Municipal Utility

- ?-

District Atomic Safety and Licensing Appeal Board Panel U.S. Nuclear Regulatory Commission Washington, D. C.

20555 Alan S. Rosent.lal, Chairman Atomic Safety and Licensing Appeal Board U. S. Nuclear Regulatory Cornission

as;iington, D. C.

20555 Dr. John H. Buck Atonic Safety and Licensing Appeal Board U. S. "uclear Regulatory Conmission Washington, D. C.

23555 Christine N. Kohl Atonic Safety and Licensing Aoneal Soard U. S. :;uclear Regulatory Comnission Washington, D. C.

20555 California Decartment of Health ATTN-Chief, Environmental Radiation Control Unit Radiological Health Section 714 P Street, Room 498

^acramento, Ca. i fornia 95814

mtc E

UNITED STATES

. J.

.:-.,. e I*1 NUCLEAR REGULATORY COMMISSION e

ii o

W ASHINGTO N. D. C. 20555

.,*, WAL/.. I.

a. V

.5

'%.*.. s' OCT 3 01981 MEMORANDUM FOR:

Chaiman Falladino Co.missioner Gilinsky C:amissioner rac..tord c

Commissicner Aheame Ctenissioner Roberts F R0'i:

Willia., J. Dircks Executive Director for Operations SUEJE;T:

STAFF REVIEW OF ORNL REPORT ON PRESSURIZED THERMAL SHOCK In my memorandu: :o the Ccr: ission'on October 9,1981, I tran'smitted a cecy cf the draft preliminary report on pressurized themal shock prepared oy Cak Ridge Naticnal Laboratory.

The-staff has completed its review of the report, anc the staff's evaluation is enclosed for your informatier, Also enciesed 's a letter from Duke Power Company presenting results of

-heir eva'uation cf the ORNL report.

principal conclusion from the ORNL analysis 'il that an overcooling transient similar :c the ecs; severe transient that has occurred ( ne

=...-.,.

e=.r.r.,,=...e.

,,2--

m.,

.:a m:n v. m...-.

.^e=.

a

.'sr=.=*.

.h.a.

...a

,=-.e.

.~r n.... =. x,

t... e. s. nv =. r.

.-..=.=..-.=..ee.,=.

~. =..u. =.1 e, c e =. ~. =...= 1 e... n., =.

s.

e

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

.e.=.

c...-.,. : =. r. e.

y. =...

e..

...=..ca....

-.. =.

3.,...

.c....=.

w

~...

cccurrence pecbaci 1 Tes; anc, using racture mechanics calcu.atiens s

l thcugn: by the staff to ce conservative, credictec failure cf the Occnee-

... =. e.e =.1

4..... -.e...

.e c y =.. =... :. c 4. =. m. s w a. r a.

^.. " - *,d. c,v.

s

.s,

.iv e.jy. - n C i ". C c' #

c' *. *. ^ #. S " '..C.E S '

V. # *. *i

• h. *. w" ^~.~. 4. C

• i o.n r, e.., e

- ' = e w.:

e. Wge m

.. w w

w.

.a

. o g r. w g g]

.) c.

...... =...

w.. w., j 4 4.. y

r.. - o., r. a... c e. r.

e gi, = r a.

.,=. c g u r. j c.3 5

.i v

. w,. 4.. 3

4. 3..,..

e. e e e,. e =... ] y. ;,.. H.... 4 - =. - 4. 2. =.,.. - - a. r. i y =.

.....=...e

4..e e

.. - r.

i..

e

.. r.r.:.. =..-

%:. e.:.

- - e. r e.

. 4. y e.

-. 4. -...e

. =,y %. =.. =. g i. r. =..--

e. g r. e r- =.

1 g...e.

wi:nir..a yea r.

c.iucC..

w w.

s.--

t,iu r l ey, n K.s -

i...

-:.- 2 i c u l

I E

\\

a i

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

4

~.,

=

The Commissioners 2-In the staff's judgment the ORNL report does not present any significant new inferration that would chance that conclusion.

Willi am J. Di rc'<s Executive Director for Operations Encl osu res:

1.

Staff's Assessment of CANL's Draf t Freliminary ' Report on Pressurized Thermal Shoci 2.

Letter from A. C. Thies, Duke O cw e r ~ omp any, t o R.

9..

Se,nero, NRC, date: Octecer 20, 1981 cc:

w/ encl osu res ACRS PCR SECY Of

'S Y 9

8

/

..n:a4

u.. : n,w. --

..o.- R 7,.

M :.r0 u,.

Y

o. e.qq.,-u-vr 4 an., Since ti :

ri n

r o

'.'.'u'": C ,' L Ls' ~: O "v.u C ' Y I T ' "t D, ~

CJm K 'ne'~

a r

n ta v:

...-.n.-

i O ~r n.R \\7 q q,:.,

Itu,,.R7.Y P0.c.D,,

,, E,.,. L.,. l C.a. C r in:

tt:.i t M i nn: i

.,n i

c. e.no..r

3. i n::..,. L e.~n e C, ;.u..., S n

. n A

.in I.

Fu ro Cs e The purpcse cf this dccu ent is to assess the regulatory i clicatiens of the draf t Oak Ricge Naticnal Laboratory Report, N'JREG/CR-2083,

. y c l u.=. #. c a.

c '. *..". a. Th ra. = *.

• o ?W R ". a.s s e'i T r. '.a.a -i ',v, Da=ed 4..

es.,

nr y.

m

.L, v O. r. a..c.e ". i - a. s.

T 5 a.... a i

<.h oc.k.v a..*.s " (C. R"'.

. a. p o. ) wi *.h ra c., e +. '.o e...v

^.,#.

c..*.4. au ed r a..* *. o r ca a r.= ' ' s.. c e n s i. d. a. -i..a,

• 5a.

,w

.m-,

...~

o o

,, r a. - u. 1.. s..a

.. a.. o a 1 c_ ' nr. k 1.e. c... o..

e., r.

+ L. 3 +.,... - n. s o,

• L. o.

r~: C

e.. s, e.

• w

...e

.s b as -.3"e = --eli.inary review Of'*he ~ report, and the staff's

's n# a... a * *. r. a. = ~ ~r *. a c e r. '. e d.h a.. a..

d -

r.,, n r 'i u e i v *.. s 2 " d.

4-cl.' u-a, r o a rig i.

v u

o a-n

. e.

't.

q....,*v 3 e..a u s..,e t,, e n.. s u

. a.., :.~ e..:. e. e e:

e.

.a.

• : '. ; s.

e. e.,

...,.., a.

.z., e a.. J 3 g....>. ey

c. e.. -. a_

w.

i,

- c_ e. s.,..

e....3

.- f. -... s. e.e. r i =. -., c _ 3.,.... - p-3, r

.,a r. '..',.e.g e ^ r. *. s *..; '. o. c,

e.g e, i

• e C i. c. -

J. e 6 L.. p

. i..

o

...g

',,, e. *. a.

c. =. v *.

2. r a.

_me ry

s..

'.e.1..,.

.e..,. a. e.

s s.=.,..

e. '3-

. =,. :.-., e a

t u.

2 n e.v. e. r.,. g] j r.

-,,r t.. e 4. v,.,...

ee.

........a : ". u t

.3.,..

4..e. w 3...

2,

..i, y

=..:....;-...-. - :. -. e..

c. a. e. e. z.

.. 1,. e : =...

.-..n..

n.

-- n,..

e.a. ;.

r.

e s.

s....

.8 2 r

• h.

'J O., ' C 's c' I, W 'e 'i l *. *v *.

,. 3 C

.t o.,. - e i, o. r.

.v m. -, e.

r s. a.

c'

'. *1 r a. 2 '. '. O *. " a.

a^:

~

3 u.

. a. c.e c.i.

e r.

- c. y e.,. 3. }. e, r s.

w

,o 2_.7-

. e.2.

s,

2

,,m. e.g.. r. c.

y 4 v 2

w e

a m,n,.,.,. 5. :-,

4.

r r. 2 r....,. o., r.. C ] U s 1 v.en C.

. w o. v :..n... C r. 1 s

+.6. 2 +

4. e. r e r. a l,o a

..v

.m events -' Ore severa tr.an the RanChC SeCO Over0Coling event Were to

.w c i.w

. p o.

re..a r. e r, re_ e. s "s r a. v a. c.c e l

v. a. ". e '. O

1. c-. a d r.

a'*

4e n,,,,..J,..c;gy, w

w.

high Dressure Or ' e re*ressurized, a"d if fracture Techanics CalculatiCnS D

believed to be conservative are used, then vessel failure may be predicted for the Ocenee-1 vessel. The NRC staff had previously reached this srne conclusien (Refs.1, 2, 3, 4, 5), which is the reascn that the pressurized thennal shock issue is under consider-ation today.

The NRC staff also had previously cencluded, and discussed with the Commissicn, that the probability of occurrence of pressurized evercccling events, nore severe than the Rancho Seco event, is sufficiently low that immediate corrective action is not warranted (Ref. 1), although longer tenn corrective actions may be repuired for sane plants within a year.

The 0"NL recort does not present any significant new information that wculd c:1ange this conclusien by the staff.

The recor: presents the completed results c# analyses for four overccoling transients postulated for Oconee-1.

These are:

a large break less-cf-cccian; acciden (LOCAS, a ain stean line break (".SLE), an :verCCCling event e+ich actually ccCurred en "a-ch 2C, 1: 75 a: : arche Sect, anc a costulated Overcecling even:

cre severe : nan :ne Earcnc Seco event, re#errec to as One runa< cay feedwater ransient I FT).

In additicn, ORNL re.iewed partial c % u l c' ' i - a. e... w a., a v-. 2. 4. c... e.

.= c c. ;... 2%

re.;,

w.,.

c.

..c1 8-

.w.

Cameleted for the 00"L recert.

Table 3.7 cf ne OR1L report c-ese' s rest ics sm:..ing e##ective For the fuli pcwer years before predictec Oconee-1 vessel failure.

large break LOCA anc the Rancho Secc transient, ;revious results are confinned that many years of c;eration remain before these events would present a potential for failure of the cressure vessel.

Time-to-credictec vessel failure results for the small break LOCA are not cresented in the CRNL report, but c her calculations have been mace which indicate this event is nct of immediate ccncern (Refs. 5 and 6) and it will not be further ciscussed.here.

I

M 3.-

The time-to-predicted vessel failur'

_:sults presented in Table 8.7

]

of the ORNL report for the RFT and Ine '4SLB are the principal focus of this report since they raise the questien of whether or not there is an immediate safety arn at Oconee or other plants.

According to the ORNL report, the MSL3 is.the icwest probability event which has been analyzed. The MSLS is stated (small table in Section 3.1) to have an occurrence frequency of 5 x 10-6 er reactor year.

Not mentioned in the. report, but apparently included J

in the qucted c:currence frequency (in order to produce over:coling conditiens sufficiently severe to potentially. fail the pressure.

vessel), is the probability that the operator fails to isolate I

^

feedwater to the steam generator with the brcken line.

That human error probability (HEP) has been ruiticlied by the MSL3' occurrence f requency to obtcin the' estimated frequency of an overcooling event that would challenge the pressure vessel at. 0conee-1, -(i.e., the i

estimated frecuency of the MSLB is not stated in the ORNL report, but apparently ORNL assumed the value cuot_ed in WASH-1400 and then used a paritcular HEP to cbtain the cucted value cf 5 x 10-6 er reat:Or year sho<in in :ne ORNL re: Ort for the everccciing event).

1: 33::::k and "il cx (312) Olants c her

.'an Oconee Units 1, 2, er

^, an auto atic fescwater isolaticn system and a steam line isolation

~

system are installed.

Prc:er operation cf these systems folicwing 4

a *1S'_5 acui t prevent an over:coling event severe enougn to challenge l

he pressure vessel.

Therefore, at other 5&W olants, estimated

' #re:Lency cf this event is a:Oroximately 5 x 10 cer reactor year itimes the pr:bability that au;omatic feedwater is0lation fails, l

times the estimated frequency that the steam line isolation system fails.

(Details Of hcw the exact systems vary from plant-to-olant l

are not completely described in this brief summary.) This co-bined estimated frecuency for other B&W plants would be below the value stated in the CRNL report of 5 x 10 cer reactor year.

In the small table in Section 3.1 cf the ORNL report, the F5T is

(

stated to have an occurrence frepuency of 1.0 per reactor year.

l l

l l

l 3

s.

i However, additional failures wculd be necessary to cause a severe overcooling event as a result of the RFT.

Therefore, the statement immediately below the reported occurrence frequency of 1.0 in the j

table must be censidered as a vital qualification of that frequency, i.e., "...for Ocenee-1 it appears that multiple independent failures 4

1 a re requi red... " The occurrence frequency for a mild transient initiated 'by the feedwater sys, tem is indeed close to 1.0 per 4

reactor yecr since such transierits are frequent, but such transients are of no consecuence to piant safety unless there are subsequent f ail u res.

The probabilities of all the other failures must be combined in crder to arrive at the actual estimated frequency of a severe cv 2rcooling event.

That ?stimated cccurrence frequency is believed to be icw,. as discussed below.

The estimated cccurrence frequency of the particular,' cetailed RFT scenario presented in the reccet is very icw since the total amount of feedwater assumed te be pucced into the steam generators is cenriderably greater than the maximum condensate :hysically avail-able in the system at a iccation where it can be a source cf feed-water.

-ssuming the a cun; actually available ir. stead cf the

1. icticicusly larger a cun; acuid decrease cne cccling anc make the actual transient less severe.

In adciticn, feedwater flow rates wcule rcbably :e reduced beicw those assumed in the recort, even

.;hile water is still avaiia:le in the syste ; :: be Ourced inte the steam cenerators.

This wcuid normaliv result from loss cf the steam su::1y :: drive the tur:ine criven.;u :s as a censequence of floodinc of :ne stean generators which are Ine scurce of that steam i

t i

supply.

That is, gross -cverfeeding of tne steam generators might be self-limiting under such extreme conditions.

This was not taken into account in the subject report and may be applicable to the MLB event as well as to the RFT event.

Therefore, the NRC staff would ex:ect that an actual RFT wculd be less severe than the one calculated in this re;crt, and it would i

I i

e,,-w.w,...--,ww,

,---r,,.-ym.,-

., -. - + * -,,

e,ew..w+w_.,,,,.-,sr yv=y,y.

,-y.,w--

-- +, -

w e.,=,-~,e-v,----e

'-w ew--w-=

m-we

,-we g-

5-

~.

a i

still require several failures, including:

the.feedwater. controller or integrated. control e,ystem (ICS) must fail; the BTU limiter must f ail; and the operator must fail to correctly diagnose the problem and take corrective action.

(These items are discussed in further detail in the bcdy of this report.)

4 The above discussions of overcooling event analyses would not be complete withcut mentien of the computer codes used.

These fall-into two categories, fracture mechanics codes and transient codes.

The transient codes are used to calculate the pressure and temperature versus tine that is ' input into the fracture ' mechanics codes; that i

is, they do the systems calculations that predict what pressures 4

and temperatures will result,,given a particular hypothe,tical event.

The fracture hechanics codes assune th'e carticular pressure and temperature versus time history calculated by the transient j

codes (i.e., that a particular event has occurred).

These codes are then used to calculate the prcbability that the pressure vessel will fail if it has a certain size and shapi' crack present at a critical loca icn en the inner surface.

The fricture reebanics ccde used ir the CESL recort, tcgether with the input data (i.e., materials precerties, including fracture teuenness and variatiens cf materials prc;erties with te. perature I

i and expcsure to neutron radiatien) shculd yield satewhat ccnservative

,results.

That is, if they differ fran reality, it is believed-that f

f ailures wculd be predicted when they would not in act occur, t

Detailed evaluatier cf the mechanics of materials aspects of the I

i ORNL report is difficult because of insufficient infonnation for i

many of the values used.

For example, Table 7.1 (page 7-1) lists partneters which must be known in crder to set up OCA-1 for a thermal sheck analysis but many of them were given only by inference (e.g., alcha, E. Poisson's ratio, yield and ultimate strengths) or l

l f

I l

i 6-bv reference to docu.ents (e.c., K and Kla).

With respect to the

.rc heat transfer ccefficient, the values given in the text (page 8 4:

r-r.?

inn 0 zis/n.

.m e,j ar.d in t able e.1 page c-e:

q either,.n0 or : 0

. -r v

ev 9

STU/hr-f t -F *) are contradictory.

The OR!;L recorts on the HSST themal shock experiments (TSE) in the past have failed to follow the dictates cf the ASSiE Secticn XI recr:mendaticns for anal.vsis and tcughness (K.

and K.,2 ) deter-

,;natiend.

4ne ccnclusions given in the,,..,

reports, that.q.

us,u.

calculaticns and cbservari.1s were in agreement, relied en a so ewnat circJiar argument Of neChaniCal procerty deteminations.

In response to a specific tiRR recuest, reanalysis of TSE-Ea by OR!;L in acccrdar.ce witn the A2'E Code shesed the criginal analysis to result ir crack ex er.r.icn cver-estima*es (Ref. 9).

The tiRR inter-creta icn cf the reanalysis led to the conclusion that the Code

.. o. x. - -

.em.d ="c *. 4.:*: " #. c. s e ~.

1' e,

• o. " e - r a. d. i.. 4. c...

. v.

o,.,J ~,. * : ". a.

--a

. a

.~

v o

.~.

o r

t. e. n.. 4 n. e,

C.

~. ~ a.

a

• :"w

e. e. n. e.

(' d o u,,

,.,...t

.. a T.a.eza. ex.a.n., -.w.e sa...4.-.e')

.n i o v oa.

, w.

c e...e :...a 2

.,-3....e

,,. e a. r.. :

4. c. - 2.

- a.., e 2.,..,, = - 4 e j.

2

..i..

a. y e.

. -... e 4.: y.

';+ '.

a.

'. e. -. =.

.1.

.e =. y a. e.z 'i e.. s e.

i:.

2

..: 4.... =. e =,. e. e.,

..e.

2. p a.

c., - c. 4. 2. s.. gau 2,. e., pi g

• r av

-e 4

v.

.v

.c o. i s.

. j :. - : s 4. 4 s,.2]

u.

1., c. c

..c.*>

w.

. w w.. e. a r....

4... e. i. y.

e., 3,, e. 3.i c. s,n..

r. :.i n

.e

,.r-u

.nr...

. r. c. e., -., - c. e c e... a.

r. c. e.....e j -.

= ~..e.

-f.

.e e *.. c 4. *. #. '.' i 'y"

.=.= i.V s # c,

~..~a.

r. ". a.#..'.'4

.~ a. ":. a. n.

~ ~ ". c. a. #v.: *s # e.. # "... e.

1. . "u " *." r a. -

i.

v.

m

. v

,-c...=

4..e

...:..- =. e :..e r a. r. r. e.

.s a.

. 2... 4.. 2 - 4.

a. i.y.

m.., 2.,. 2 e..

s..

i Wi*.h rescect tc the systems Ccdes used tc calculate the primar.v syste cco.i ant cenavi or, tn.e.m-code use,. in any c,. the ca,i cu lati ons is not believed to be appecpriate for such calculatiens for severai reasons.

It dces nct centain a realistic nodel for ficws in the crimary system cnce significant voiding has occurred.

Instead, it uses #1cws that are in;ut as a table and therefore are invariable.

,, hen sic.niticant vc;c;nc. cccurs, as it cces in to,e - _i event cresente,..,

n n.

4

/

7-the ccde ccntinues to assume primary system coolant. flow to the steam cenerators, where it is ccoled and then returned tc the primary s.vstem tnereby making the eressure vessel cverccoling event worse.

In actuality, it is predicted that for certain cases the voids would collect in the system high points.

If sufficient vcids cellect at the tcp cf the het.lec inverted U-bend, a natural circulaticn fica inter ucticn (or "vacer leck," to use e cor. mon analce..v) would occur and overcooling would be greatly decreased.

!n this way, the ccde sculd tent to predict events more se/ere than the actual event.

The co'de also has a different cause for inaccuracy in an unkncwn direction. That is, the code fails to conserve mass and enero..v throughout the calculatien, b.v cbserved amounts as much as ::*.

ne ener:v cr mass ;;cwinc. out c,. ene volume does not e:ual tne total amounc cf energy or ass received at all other volumes as a result Of the #iew, as it must, in reality.

This discrecanc.v can result in errcrs that will var.v in mac.nituce and

' c- *. a r. '.n a '.-. '.'..= c.v..= ri a.

c.... s e. v.= *.i v a.

~. :..,.. 4. - r t,. a.., =... rs d'a. ** *.d e i

. - s r.~-..

.-...e=_.=.'ve_ '.v, a. v i r.3 2. ~ '. *.e i. '

4

.. c. r..,.. =. ; * ". r e.. -..

c.2.e.e

.... ~. e.,, - -. }. n,

3.., e... z a

i."s.

( P. a.... o.< =..O,

. 4i,,_

.c

,.e.

e. =.,/ a. - a.

. v =......"....>c.

v....

.. a. e.a..,.

.m

.. a.

r. - -. a. a _1

v. a. z. c. a ] t. e. r z. av p,, *i

.e ra.

.a..

,.,.. -, z. a.

c

.v.

~.

.u

,,--.., =...

.-=_s

'=. -.c.r.:.a.

.. -.. = *.

i s. - e e d. '. '. a

..2. c.

. =.

. _ ~

7

- s. ~. >.. =.... r =. e =. v. e. <. z.

=. ~. =. e. e. i,.4 '.. +.... e.n.~-. s. '.. :- i,n q ~...,. e

.n.. e n. a.,,, a_

a.

4. c.,

,._,em..

e, c.....

e

=.. 4. -. s..

". a. s * = '. #. '.e m' " " -- o....

1. . e.. " ~ *. *. e v-. u.. a. - c a. '.ca..

e.~*v

1. . a sava.re e

3..

a.3r *. *,v c#. 9ea.

., r a. a w.q. *"%a a

...., l s.

.j o_ --, e 1...

...ci...

.. c~.

a 6

i

• ,. w

..v Oconee *. vessel is sufficiently low that time is available for the

..*. a. '. u 'i 1 v a.v c'i " a a.

.a. -. r~d i. '. '. * ".. a.

v a.s s a.l = r~d "r-osa.

, e. e

.-s v

. v 2..

sciutions to the creciem be# crc further regulatory acticn is needed.

i

~. '

4 III. Discussion t

The sections below previde a acre co prehensive summary of the preliminary review of the CRNL report by the NRC staff.

A.

The Cverall Recert from a Svstems Viewecint (1) Runawav Feedwater Transients (RFT) i i

Of the general classes of-transients identified, the cost complex

(

i

~

fran a system and centrei viewecint is the Runaway Feedwater ine reecwater Cen:rc,i su system c,. the,:ntegrate,.

s

_i ransient (nr i).

Centrol System (ICS) is designed to maintain a tctal feecwater flew ecual tc the feedwater flow cemand. The fics in the feedwater system is centrelied by the ICS in the automatic code by using incut signais and menitoring precess parame't'ers, or it can be r

centrciled by the cperator thrcuch the ICS a.: the Lecp Feedwater Demand level er at the Feedwater Vaive ?c'sition er Feedwater Furp a

Sceee level.

The ccerater can intervere at any time, therefore he can te an initia:ce and/cr a :eminater cf an everecclin; transient.

i The cause cf the RFT may ce internal or external te the ICS (co cenen:

'ailure er ccerater errer).

The severity cf the everecciing transient for Ocense-1 can be reduced or ten.inated in ene cf :nree l

. hays:

(1) The ccerater can take centrol cf

~.e FW cunes er valves, if qis incica:icn (i.e., instrumentation) is cceratir; and he i

l di'agnoses the crcblem correctly, in which case ne event can be teminated cuicki.v.

However, if :ne event..is nc: ciagnosed correctly, t

l he may cake the evercooling ccre severe.

(2) The STU iisiter can

.., demand since it cent;nuousiy ca.culates tne -_i,Js or energy n

i e

1,.m1 :

I centainec in the steam ceneratcr.

(3) The hi-level limit is a i

1. ixed se:pcint which limits the licuid level in the steam cenera-t e rs.

The hi-level trip (present eniy on Oconee Units 1, 2, and 3 r

I

?

s q

4

--,_-.~ _,-._ _ -,_ _ _.---,-_.. _,._.. -,.,_,..., _,. -. _.-._,.-,,,, _,---,_-. - -. m. -,-, - - -,,,-,

_ _ = _

9-i i

but not on other B&W plants) can trip FW pumps at its independent fixed setpoint.

All of these potential mitigation actions were implicitly assumed to fail in the runaway feedwater transient in the ORNL report.

(2) Main'Stcan Line Break (MSLS) i 1

The Oconee Units 1, 2, and 3, and Rancho Seco do not have main steam isolatien valves (MSIVs), whereas all other B&W anits do have MSIVs.

The three units a' Oconee do not have automatic FW isola-tien, whereas all other 3&W units do have FW isola' tion.

All S&W.

{

units except the Ocenee units have some type cf rain steam line break logic that will isolate'FU to the f aulted steam ge~nerator (at l

same plants both steam generators,) and at some plants the logic will also shut the MSIVs.

At mest units the cooldewn transient wculd be tenninated by the stern line break icgic (MSIV cicsure and/ ar feedwater isciation will terminate the cooldown, with a

=?

delay for S3 drycut in the case cf cnly feedwater isolatien).

It f

is, kcwever, necessary to take credit fer the c:erater terminating er thrcttiing the H?: rum:s at a later time in the even?.

i

]

At Ocenee-1 with no .SIVs er automatic FW isciation, coerater.

I acticn must terminate the.MSLB cverccoling transient by cicsing the l

W valves anc alicwing the steam generator to steam dry.

No j

' credit was assumed i-the CENL escrt fer crerator acticn to limit feecwater flow for tne MSLB accident.

I 4

(3) Main reedwater Svstem as it Would Ocerate for a RFT and MSLS

.t In the calculations of the ORNL report, full main feedwater flow is 1

assumed threughout these events.

'r"th one steam generator flooded and the other steam generator isolated, as in the MSLB, the coerator will prcbably not be able to maintain the turbine driven MFW pumps

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

-. ~.

s e

-6O

.a 4

m u

3 J

.t If '

+

w

+4 10 -

I in a running condition because the prinary steca source has been lost.

The condensate pumps and condensate booster pumps do not have encugh head to naintain full flew for these events.

Multiple f ailures must cccur in the ICS to prevent automatic r nback and trip of the MFW train. Without a sufficient supply of feedwater (condensate in-the hotwell) th: MFW pumps will eventually lese sucticn.

As pointed cut in the Summary and Conclusion section above, the ORNL report does assume more than the actually available f

amount of condensate for the F5T event.

Al' of these reascns why MFW may be lost er reduced ware ignored in the ORNL report analyses, thereby making the everecoiing more severe for the'R?T and MSLB accidents in the rescrt.

In additien, the feedwater temperature was assumed to ramp down to l

the hot well temperature within one minute After MFW pump trip, a conservative assumption. The likelihood of such behavior is extre ely snail since multicle failures cf uricus systems would have to CCcur.

I 3 e TeOOrt acknowIedges that ;1 ant design.odificatiCns have alreaty been made whicn will reduce One' likelihood of excessive

~

feedwater transients at Ocenee-1.

"o attempt was made to determine the effect Cf Onese c0ificatiens On the Olant's susceptibility to such transients, 4.e., no CreCit was given for the decreased expected 7

l

'frecuer:y cf these trarsients resulting # rom the odifications.

i 3.

The Overall Re: ort frc, a Drebability and Risk Assessment (RA) Viewcoint (Prcoability cf Transient and Accicent Secuences) 1 I

(1) Sum ary l

We have concluded that the occurrence frequencies estimated in i

the ORNL report for the types of initiating events analyzed are reasonable and in fact are censervative for the MSL3 and 1

1'1+---

m,-,-w-~,-4,.-._..,v,,,wrm,,,,,._.,,,

.,ww,,,,.c

.rym.,_,.

~

. RFT when cacpared to estimates in use by the NRC staf' f.

Comparison of the ORNL report and NRC estimates is given in the following table. The ORNL report does not appear to distinguish (as done below) between the probability of the initiating event and the resulting cvercooling event probability.

Estimated Frecuency Per Reactor Year ORNL NRC Pressurized Pressurized Initiating 0verc'coling Initiating Overcooling Event Event Resulting Event Event Resulting from Initiating from Initiating Event

• Event

• 3x10-1 6x10-2(5 &W) siO-'(3&W)

RFT 1

Not Stated (CE&E) <2x10- (CE&E) k La rge Not Stated 5x10-6 3x{0-#

3x10-6 MSLS Small 3x10-#

Not Stated 3x10-1x10-5 LOCA Large 1x10-Not Stated 1x10' Not Stated LOCA 3

s10 i(S&W) 3x10 j((E&W) l Ranc're Mct Stated Not Stated Ex10- CESW)

N10-'(CE&g,)

Seco-

• Ecugi o f recuency of initiating event times orc:acility of accitional failures and/or error ;robabilities as discussed in text.

Overcooling transients at pressure in--PWRs result from small break LOCAs, main steam line breaks, or feedwater transients, only if additienal failures, either hardware-or human-related, occur subsecuent to the initiating event.

From a PRA viewpoint, we believe that a more realistic way to analyze an overcooling transient at pressure is to consider it as a secuence of events.

Using event tree methodology, the overcooling transient

O 12 -

sequence is represented by a set of event trees.

Each event tree in the set has event headings, corresponding to a different initiating event, a specific assumption made in the analysis about feedwater fica rate and/or temperature, or a postulated f ailure (hardware-or human-related).

Rigorous determination of th'e estimated frequenqy of cccurrence for each event sequence thus generated wculd involve assigning an estimated cccurrence frecuency to each event and combining them to cbtain the estimated event secuence frecuency.

Such an effort was beyond tne.limithd scope Of this review.

However, within the past year, the NRC staff has made simplified analyses to obtain estimates of the frecuency of everccoling transients wnich are surmarized in the abcve table.

To date, we are not aware of any subse:uent analysis, including the subject ORNL report, that would cause us to alter these estimates.

The integrated NRR/RES task acticn plan being prepared fcr tne technical resciution cf tre pressurized thermal shock issue includes a ripercus

.a analysis c# :ne everccciing transient

.e,--..o......:. - 4.

.J e e. c..s

.s. r.\\ c..

=c

=..,2...w c.-..

2.. ~- e e

<-s

-( 4 j miscussiCn Scecific CcT~ents rc:arcirc. eich of the Classes of initiatin.

events are as #0llcWs:

i

-.ine ecst sericus cressurized overcco,iing 3

anche cecc : vent:

fta, n

event was that at Ranche Secc cn " arch 20, 1978, in which the coolant tercerature crcpped from 550*: to 280 F in about I hour while the system cressure first drocped, then returned te near its criginal value.

Based en tnis excerience, an occurrence frecuency ci : x,10-2 cer f

' reactor year was estimated for a B&W plant to experience an overcooling transient as severe or more severe than the Rancho Seco event (as described in M. A. Taylor's memorandum of October 29, 1950, Ref. 3).

. Since this occurrence and the occurrence of two less severe events, operstors have received special training in transient response.

Babcock & Wilcox plants have added a back-up power supply to the non-nuclear instru-mentation bus, whose failure initiated the three transients above.

The NRR staff examined the impact of the improved power supply and operator training and suggested that

^

these improvements might have reduced the frecuency to 10-3 per reactor year for an overcooling transient as severe as the' Rancho Seco event for B&W plants.

For more severe events, such as the RFT, that might challenge the Ocenee-1 vessel if they were to occur today, the staff

^

estimates that their frecuency.iA 10- per reactor year for B&W plants.

The cperating exoerience cf CE and Westinghcuse olants

~

has also been examined.

There have been no events like tne Rancho Seco transient, but there have been so,e o recu rs o rs.

These are events which typically led to secondary steam dumo valves er stean byoass valves sticking cren, but which did not resuit in steam ficws larce enough to produce very severe overcooling transients.

i The most seve're of these transients occurred at Arkansas Nuclear One-2 (a CE plant) on Dec' ember 27, 1978, where a

~

main steam relief valve lif ted and failed to reset, thereby causing the reactor coolant temperature to drop by 107"F in 52 minutes.

This lack of severe overcooling events at CE and W olants plus the greater thermai inertia

A

_ 14 of most W and CE plants, leads the staff to estimate an RFT occurrence frequency cf a factor of 5 lower than for 35W ~ plants,. 2 x 10-5 per reactor year.' Also, the'. staff estimates the frequency of a large steam line break or

.its equivalent to be no greater than about 10~# per reactor year, and fr a pressuri:ed overcooling event resulting from a MSLB severe enough to challence the Oconee-1 vessel if it were to cccur today, the estimate is a factor of 30 lower, 3 x 10-6 per reactor year.

These estimated frequencies are summarized in the above table. There may be a factor of 10 uncertainty associated with these estimates.

5.

Small Break LOCA:

The ORNL. report does.not provide cc plete calculations for the sma-1-1 break LOCA.

However, in a simplified analysis of an overcooling event initiated by a small break LOCA, (i.e., htEeen 2" and 6" equivalent diameter), the NRC staff (Ref. 5) obtained an estinated i

i f re;uency of cccurrence Of 1 x 10~; per reactor year.

.his result was based en an assu e: cccurrence of 3 x 10~ cer reacter year for :ne LOCA event and an operator human errer prcbabili y of 3 x 10-2 (eseratcr f ailure to tnrcttle or ten-inate safety injection punts).

i c.

"ain Scea-ine Ersak:

The table en page 2 4 of the CRNL report gives an estimated frecuency of occurrence for an i

overecoling event resultirg from a MSLS of 5 x 10-6 er reactor year.

Reference 5 contains the results of a simplified analysis by the NRC staff of the probability cf occurrance of an overcooling transient caused by a MSLS ar.d subsequent coerator error.

These results are summari:ed as follows.

l

~

9

For the case of a large MSLB-initiated overcooling transient, the estimated frequency of occurrence was 3 x 10-6 per reactor year.

This result was based on an assumed frequency of occurrence of 1 x 10-4 per reactor year for a large MSLS and an operator HEP of 3 x 10-2 per

. demand (f ailure to teminate feedwater flow to the steam generator (SG) with'the broken line and/or failure to close the main steam isolation valves to that SG).

d.

Runaway Feedwater ~ransient (DJT): The RFT analyzed in the ORNL report assumes multiple failures subsequent to an initiating event.

A better description might be given by the term "cverfeed transient."

Such transients usually arise. frxi other transients which initially e ty the steam generator (s),' such as, in this case, a stuck-ccen bypass valve.

Foliewing this, a less of automatic feedwater centrol or a manual error ccupied with the f ailure of the operator to diasn~cse the situation and take acerccriate corrective action would result in excessive fee $ater being sucalied to the stea generater.

T e NRC sta#f nas perfcr ed a review of Licensee Event Reports (LERs) regarcinc evercooling events.

3ased en this reviea, a frecuency Of cccurrence of cverfeed transients of 3 x 10

• per reactor year was estimated for S&W piants.

~5e ccreesecnding estimate for '.lestingnouse and CE plants w as 6 x 10 7 per reactcr year.

A realistic estimate Of i

ne frequency of occurrence of an RFT must censider the

1. recuency of occurrence cf tne initiating event and further independent failures (e.g., failure of the steam generator high level t'JW pump trip) and/or continued inapcrcpriate cperatcr actions which exacerbate the t ransi ent.

Tne inclusion of nultiple failures, both human and hardware-related, requires analysis of an

1

~.

entire spectrum.of RFTs.

The NRC staff's action plan for resolutien of the pressurized thermal shotk issue includes such a complex analysis. The occurrence frequencies are p

l believed to be low, but quantitative results are not now 4

available.

However, a preliminary estimate is given in i

.the abcve table.

j C.

The Overall Recort fran a Svstems Ccde Viewooint The NRC staff censiders the use cf the IRT ccmputer program to evaluate the response of the primary system to severe overcooling transients Oc be inappropriate, since this class of events is well cu: side the range of the program's capabilities.

IRT is capable of handling mild or intemediate transients which do net result in void femation-in the primary system.

IRT does not adecuately conserve mass and energy.

The followinc critical items demonstrat'e the shortcomings of IRT fer use in analyses of severe overcooling transients:

( '. ) rics Distribution:

I:.T cces ne scive the he entum ecuation. The input data scecifies the crimary flow.

The heat remtval rate is de;endent on ne flew.

For the cases cresented, a natural circulation flow taken from the Ocense FSAR is input.

1 i

-(2)

Voids in the Primary System:

In tne IRT calculations, veic fcreaticn is allowed oniv in the reactor vessel 4

j upper head region. The effect of voiding is not preserly 1

treated in IRT, which is a homogeneous ecuilibrium c al culati on.

or certain cf the cases presented in the 00.NL report, the upcer head region is vcided at about 100 seconds.

Af ter this time, additicnal voids are incorrectly 1

..w

-,._.-,---m

assumed to be hanogeneously mixed thrcughout the primary system. The assumed primary system flows' in cases involving assumptien of single-phase natural circulation flew are therefere ince.rrect.

In fact, it is expected that the voids will collect in the pipes leading to the steam generators and interrupt the circulatien, de-ccupling the se:0ndary system fran the primary system and renoving the heat sink.

The loss of the heat sink will stcp the c ool d o*vn.

(3)

Energy Salance:

For the " runaway feedwater" transient at ICC0 secencs, the discrecancy in the energy balance accunts to 25 (R e f.: 7).

Brockhaven National, Laboratory (BNL) estimates that this correscends te a temperature errcr cf a:prcximately 30*F.

It is not known whether the

~

error is censervative er non-censervative. For the MSL3 cver ccling transient, the discreqancy is negligible.

The NRC staf# asked Les alares Naticnal Labcratcry (L'NL)

p ce #cr-T:JC coce calcula:icns cf M5L3 and "ruraway

.- - a.. - -

z. 3-

...,. z. a =... g :-. - ns<

e.n.n..ee. <n. <.,.

~

. =. e.....,. n.. u Oc these cerf ormed by 3NL with IRT.

The results fra the TR* C cal cul ati c".5 for the M5L3 sr.os nat the terceratures calculated by One t.vc ccdes are in agreement with cne ancther (Ref. 3).

The cressures differ, however, because cf tne cifference in cceling cf the pressurize #.

A ncn-i t

1.s i ecuilibrium mode..is usec in.-. wnl.ie innC uses an ecuilibrium.cdel.

Actual press _ures wculd cr:bibly lie between resuits from the two codes.

TRAC and IRT differ greatly in their modeling cacacilities of varicus phenomena.

For example, two chase flow in the crimary is mcdeled in TRAC and not in IRT (except in the

18

~

2 ucper head and pressuri:er).

An agreemen't between the results calculated by the two codes could be expected to occur only if single liquid phase exists in the primary.

.In summary, IRT is not an appropriate program to use for severe overcooling fransients. The treatr.ent of momentum (flow) and void fonnatien are key elements in the transient b ehavi er. The results presented after 100 seconds are highly suspect., The staff believes the overcooling rates calculated are conservative.

Hewever, it is not possible to cuantify the encunt of censervatism.

D.

The Overall Recort from a Fracture Mechanics Viewcoint (1) Codes The use of the OCA-1 computer code to calcelate the stress intensity (using assumed flaw =crchology) is accectable.

Nevertheless, both

be CRNL arc the NRC ::aff agree :na: it is imper: ant to modify the cece at the earlies: ;cssible date c include One tem:erature,

cepencence of material parameters (such as the elastic medulus, coefficients of thermal expansivity and conductivity), rather than using average values.

Also the ;cssibii.t.

cf crack arrest in materials of high :cughness (relatively high temceratures, ucce -

shelf energy levels) has ne been a: dressed.

Since acvanced elastic-l slastic fracture rechanics concepts are required in treating this matter, the lack of a solution in the ORNL report is not surprising.

I However, accurate calculation of the crack arrest behavice will 1

hinge on the treatment of the~ high temperature, high toughness i

aspects.

In general, the fracture rechanics calculations in the ORNL report were performed in a manner with which the NRC staff agrees, but

~

19 -

refinements to the OCA-1 computer code have been suggested for sane time and improvements could be made today.

The OCA-1 code can provide results which can be judged accurate; the results will be censervative when censervative inputs (lewer bound K RT 7,

ND7 per R.G.1.g9, etc.) are used as was done by ORNL.

(2) Warm Prestressing Warm prestressing (WPS) is the ten applied to a phenomenen which can limit the extent of total crack advance during same evercooling transients. WPS has been demonstrated in the laboratory with small soecimens and in a large thick-walled cylindar during an unpressurized thermal shock experiment.

In general, the NRC staff believe's that considerable cautien must be usec if any credit is taken for the effe' cts cf wann prestress in analyses of the pressuri:ed thennal shock problem.

The draft ORNL report dces show results bcth with and withcut WPS (Tab'e S.6) altneug9 the carrents in Table E.7 indicate that WPS is effective cnly #:r a large-break LOCA. Tnere are sc many detailed variatiens in :ne ccstulatec a:cident s:enarics, involving turnir; pur:s en er eff anc tac?ing several water sources, that the time variations in K. are :uite unce-tain.

Oriv altn a rather srecth chance in K.

a relative to the tougnness, K.

(which aisc varies with time) can

C, the benefits of WPS be assured with ccnfidence.

E.

'Fiuence Uncertaintv The ORNL rescr: states :na: the uncertainty in fluence estimates can be as great as :50%.

However, centrary to the ORNL estimate of

50%, the staff believes that current fluence estimates can be made

d:hin :20% orovided that ene uses:

(a) a well calibrated and cenchr.arked trans;crt code and (b) measured values of the neutron flux and its distribution.

t.

s 20 -

The staff has a threefold progra.i for code calibratien and benchmarking using:

(a) the results of the PCA experiment, (b) "the surveillance capsule results frm Maine Yankee and Fort Calhoun, and (c) the surveillance results fem ANO-1.

Consequently, the staff ex::ects future fluence calculatiens to be used in longer term resolutien of this issue, to be within :20% instead cf the ORNL re::crt's stated uncertainty cf :50%.

t P

7

(

P

References 1.

Ecard Notif7 cation - Thermal Shock to PWR Reacter Pressure Vessels

( : y..:,.. n c ),

.v..a v o,

:1..

v-2.

Thermal Shock to PWR Reactor, April 28, 1981, Memorandum for Marcid Denten from D. Eisenhut.

3.

Insights en Overccoling Transients in Plants with the B&W NSSS, Octo:er 29, 1950, Me.mcrandum for S. Fabic from M. Taylor.

rressuri:ey i nerma i :n ock, Commi s si en a:er, e.-y-c.1-c.c, May,,

r

. r..

1981.

5.

Pre:uency Of Excessive Cecidown Events Challenging Vessel Integrity, Ashek C. Thadani to Gus Lainas, Memerandum-dated 'oril 21, 1981.

5.

Summar.v Of Meeting witn the Babcock & Wilccx, Westinghouse, and Cor. ustien Engineering Owners Grou s en July 28, 29, and 30,1951,

- c. e. =... '. " =. i.v,.- -... =. r..

- -.a..ce."..'.-=.. --=.....': s " e s-k. ^. r =..=. r m

. - =. c e... =. ". =..e s =. l.e

,' : : ' ",, '. - k. =.. N.. ', 1 1 '. ^.. =...=. '..., P '..'., s i, " " - " s. '. ',

.c..:.

e i.v.-.

.e.

e.,. 4..,

"v. ass an.

4 s

n.....

r.

3...:....... =.. r...

m.

Er.er;y Non-C nservatis-in Overcooling Calcua:icns with IRT."

3.

~?ers:nai communication :: N. Iurer, RES/NRC, from LANL, about n-....r 3.:,

3 0. c 7..

s...

9.

January 5,1981 le :er f rom G. D. Whitman to Warren S. Ha:elten.

e DUKE POWER COMPANY CHARLOTTE,N.C.28242 A. C. TH I ES (7 o 4) 373 4249 St.c= Vict Pet.ssoCNT P=couctione u.o Tau.suessicas October 20, 1981 Mr. Robert F1. 3ernero, Director Division of Risk Analysis Office of Nuclear Regulatory Research D. S. Nuclear Regulatory Cc=ission Washington, D. C. 20555 subj ect: ORNI. Kvaluation of the hreat to ?WR Vessel s

Integr ry ?csed by Pressuriced ne=al Shock Pressure; d

Draft Inter'= Report lear Mr. Bernero:

Duke ?cwer Cc piny. appreciates the opportuni:7 :o previde ce=ents en the subj ect docu=en:. As you are aware, Duke has previded certain specific technical infocatien iegarding the Ocenee Unit i reacter vessel in an effort to assist the TRC in the ec=pletion of this evaluation. Duke engineers have reviewed the subj ect docu ent and censider that 9e av=1"'** ~' --"ede eie-can deficie-dae da ^e area of the= al-hydraulics ceaditdens and recre-d#d

~~e application cf these transients sents unrealistic trarisient cenditiens.

n to the Ocenee i specific material prcperties results in -de'a W --

--4

-a=a-i-C a== calcul ed veeeel l'f a 3-a, Our =cre salient cence=s are in the icileving, paragraphs with additional details provided in the attached.

Ss.

he evaluatic6 of the reacter vessel the=al shock issue is extre=ely ccepley and recuit 2s a thercug'n understanding of several highly technical disciplines.

Areng the tec'~nical areas invclved are ins:ru entatien and centrels, syste=s analysis, reac:cr vessel caterials, ncn-destructive exa=ination techniques, linear elastic fractufe rechanics, and prebabilistic risk assessment.

In crder to de a reaningful' evaluation, these technical areas need to interact in a eccrdinated canner; the results of cne area cannot be input inte subse-quent analyses without a thercugh understanding cf the basis cf the input.

This docu=ent ices not indicate that any coordinated effort was at:c=pted by the various 3rganicatiens involved to assure that the results previded were realistic.

In f 6c,1the dec=eeit tends to != ply tha: the individ;al tasks were perfomed iCdependently of each other with the end result being a totally de s4einted dec=ce that is not suitable for understanding and ce=unica: ng the real cersrective e: the issue.

,A One of the principal.cechanists centributing to the occurrence of pressurized

.s reactor vessel fracture is the creation of certain unique te perature-pressure ti=e histeries at the reactor vessel, the calculatica of which weuld require insights into plant design features, syste failures and effects, plant per-fc=ance constraints, and transient behavior.

~he fracture rechanics analyses e= bodied in this decurent are, in =ost part, based en arbitrary, artificial vf D bNbI M D

_J t

(

Mr. Rober't M. Bernero, Director October 20, 1981 Page 2 ther al-hydraulic accident conditions 'and not ger=ane to the real plant situation, especially for the Oconee reactors. The =ajor deficiencies in the thernal-hydraulic analyses are identified in the attach =ent to this letter. Portions of the OFSL report have also very apprcpriately discussed the li=itations and deficiencies in the the=al-hydraulic analyses. Yet fracture =echanics calculatiens were dene for these extraneous and irrelevant accident conditiens.

The subject docu=ent is inconsistent within itself, which can cause signifi, -

cant interpretation dile==as.

The report was originally intended to be an evaluation of the 35W NSSS design and its susceptibility to pressuriced ther=al shock. Ecwever, the docu=ent contains state =ents which =ake it unclear as-to whether er not'the intended purpose was achieved as noted by the following.

In Chapter 1.0, it is stated that although Ocenee 1 was selected for the initial

study,

~ ' "...ther=al-hydraulic behavior needs to be further evaluated as recc== ended later in this report and because there are special

~ ~

centrol syste=s provisions in Oconee-f limiting transients,

=cre analysis needs to be done befere their results are applied to Ocenee-1 or generaliced to other plants."

is is further elabcrated upon by the fellcwing frc= Chapter 5.0:

n "All the current si=ulations possess li=itations which give concern fer the realis: of the ther=al-hydraulic predictions. These li=itatiers are, in part, inherent in the codes and also result frc= =edeling deficiencies and questionable input assu=ptiens..."

And yet the folicwing state =ents occur in Chapter 8.0 without qualificaticn:

"A su==a y of results for the five overecoling accidents analyced is presented in Tables S.6 and 8.7.

Table 8.7 indicates the tctal nu=ber of EF?Ys that a S&W-ty'pe reacter can operate before the overecoling transients censidered would likely result in vessel failure."

and also,

"...the inclusien of cladding in the analysis will also result in s= aller threshold fluences.

Thus, in this respect the results in Table 8.6 and 8.7 are sc=ewhat opti=istic."

• ie consider these latter two state =ents as =isleadh.ng and inappropriate cen-sidering the significant li=itatiens of the study.

a

4 O

Mr. Robert M. 3ernero, Director October 20, 1981 Page 3 An additienal cene.ern is that the subject doce=ent does not sufficiently address significant progra=s currently in progress that address the areas cf vessel =aterial properties that are supported' not only by Duke Power, but also by other utilities that cwn plants with the B&W NSSS design. This is particularly surprising because by le*ter dated h'ay 12, 1981, J. Matti=ce, SMUD, on behalf of the 3&W Deners Group, sub=itteo a letter report to the Staff outlining such progra=s that had been ce=pleted and those still under-Ev failine to recocnire the other encetae etudies en this issue, the way.

- recore ine11pe we-

<-<e

-s, v, -

--ma'esg, e

---,M na.

This is incorrect._

It should be noted that certain branches within the NRC Staff are aware of these progra=s.

The evaluation of the reactor vessel fluence aspect, the interpretation of the Oconee reacter vessel =aterial para =eters, and the fracture =echanics calculatiens contained in this report have also several li=itations.

It is apparent that the chapter en fracture =echanics calculations contains several pessi=istic.presu=ptiens and opinions based on unsubstantiated data and limited infer =ation.

A technical report of this nature should be based on an obj ective analysis.

Further, the docu=ent f ails to address two i:portant ite=s which are associated with this issue.

One is the enhanced inservice ena=inatien of the reactor vessel beltline regica velds in crder to achieve a higher confidence level in selecticn cf initial flav size. As the NRC Staff is avere, such an enhanced exa=ination was perfer=ed en the Ocenee 1 reacter vessel during the current allows cutage, using an ultrasenic technique with a stand-off distance that detection of near-surface flaws. Not culy were all results within ASMI code alicwable, but also they were s= aller than those sizes critical to the ther=al shock issue.

All indicatiens were considered to be pre-service induced rather than service induced. The seccnd ite= is ther=al annealling, which is briefly

=entioned, and then cnly in a positive sense. While the technique used in centrolled tenditions =ay see= prc=ising, extensive verk and effert will be

)

required to perfect a technique suitable for use on an irradiated ?WR reacter vessel.

It is misleading to state that such a technique-is currently practical, I

particularly unen solely based en a personal ce==unicatien and preliminary laboratory results.

l As in the case of =any othcr severe accidents, reactor vessel ther=al shock cannot be envisioned to be forgiving to all bounding and overly censervative In order to obtain =caningful conclusiens of the severity of as s u=p tiens.

the proble=, it is necessary to analyze systematically accident conditions by censidering relevant initiating events, =echanistic syste= failures, and credible operator actions and by utilizing phenc=enolegical codels and =etheds that realistic syste= boundary conditiens and plant perfor=ance take into acccunt Duke has recogni:cd that the reactor vessel ther=al shock issue constraints.

i l

D I

-m

Mr. Robert M. Bernero, Director October 20, 1981 Page 4 is a very inportant issue that requires careful study and ti=ely. resolution, and the way tb approach the issue is by =eans of a cogent and systenatic analysis of relevant accidents and by consideration of plant specific features both in regard to syste= capabilities and vessel para =eters.

Duke has been fervently working on such an effort, and it is our hope that when this work is ce=pleted, the necessary perspective on this catter will be obtained.

In su==ary, the report in its present for= is not suitable for understanding and co==unicating the real perspective of the issue.

In fact, it could unduly distract attention fro = the orderly efforts now being pursued on the resolution of the issue. Accordingly, we ask that the report be =odified significantly taking into account cur ce==ents.or be withdrawn frc= general release.

Very truly yours,

/ / W A. C. Thies RLG/php Attach =ent ec:

Mr. R. C. Kryter instru=entation and Controls Division Cak Ridge National Laboratory P. O. Ecx I Cak Ridge, Tennessee 37830 4

i A

DUKE POWER COMPA.Y N

Detailed Cen=ents on OL'1, Draft Interi= Report Evaluatien of the Threat to FWR Vessel Integrity Posed by.?ressuri cd Ther=al Shock Events Chanter 1.0 page 1-2, 3rd paragraph:

We agree with the statement about the need to perfor= " realistic systems analyses to determine appropriate input temperature and pressure transients for the vessel integrity studies, and [to evaluate accurately} the =echanical integrity of the pressure vessel" through plan specific studies. Ecuever, the analyses conducted thus far fall short of this goal, as recognized on page 1-3:

"...because ther=al-hydraulic behavior needs to be further evaluated as recc== ended later in this report and because there are special control system provisiens in Oconee.1 li=iting transients, = ore analysis needs to be done before

~

their results are applied to Ocenee 1 or generalized to other plants.,,

Chapter 2.0

~ - -

A clear and consistent' definition cf the " runaway f eedwater transient" is necessary.

The thermal-hydraulic analyses utilized in this reper ~ censider this transient to censist cf an unmitigated main feedwater overfeed transient event with a concurrent f ailure of the turbine bypass valve systen fc11 cuing a reac:or trip transient.

Ecuever, the prcbability discussien of Section 3.1 apparently visualizes this accident as a more general secondary syste= upse:

cenditien which includes stea= generater overfeed ::ansients, stes: generator pressure centrol malfunctions, and events involving failures in feedwater flev centrol and SG pressure control functions.

Chanter 3.0 page 3-1:

In order to obtain the real perspective of the safety significance of this probic=, one needs to consider the probability of occurrence of a break in the reactor vessel at the 'ectree: ' location and of sufficient sice to ec=-

prc=ise adequate core cooling capability as a result of crack initiatien and prcpagation.

This probability is cc= posed of several (pessibly independent) probabilities, including (1) the probability that a break large enough wculd occur given that the fracture mechanics calculatiens predict a through-wall crack propaga:icn, (2) the probability : bat a through-wall crack propagation uculd occur given the specific pressure-temperature condition (this probability

~

is dependeat on the probability that flaws of certain unique size and orienta-tion capable of through-wall propagation exist at the location of minimum

=aterial strength), and (3) the probability.that the potential transient'.

events pro 3uce the pressure-te=perature conditions necessary for unarrested crack propagation.

page 3-1, Table and 3rd paragraph:

No basis is provided for the assigned probability of a runaway feedwater transient (RFT). The value provided is arbitrary and is not based on any review of operating experience or quantitative assessment of probability of RFT that causes severe overcooling conditions. The RFT characterized by a frequency of occurrence of 1/Ry represents a general secondary system upset

. condition of an evercooling nature and not the accident treated in the sub-sequent sections of the report.

1 j

page 3-3, 1st paragraph:

)

The IF?Y results provided in this paragraph are not valid due to the inherent

~

errers and~ limitations of the ther=al-hydraulic conditions utilized. Further-

= ore as discussed in detail later, the assumed fluence rate per IFPY is i

inaccurate.

l page 3-3, 2nd paragraph:

i The basis of this state =ent is not apparent. Figure 5-4 shows predicted I

temperature response for all transients including RFT and MSL3 (IRT). Within j

600 secs for RFT and 250 secs for the MSL3, primary coolant temperature" fs predicted to be belew 200 F.

This figure would tend to indicate that the 0

j predicted tc=perature is well below 212 F during cost of the transient rather than well above 212 F at the ti=e of predicted failure.

i Chanter L.0 page 4-1, 2nd paragraph:

\\

l The last sentence is incorrect.

The neutren pcuer signal obtained frc= the j

R?S can codify main feedwater demand if its mis =atch with the ICS reactor I'

de=and level exceeds a set tolerance oniv if the conditiens in the stea l

cenerator cernit, i.e., 3TU-limits, high and low S.G. level limits override.

Loep A and 3 steam generator feedwater demands are reduced to zero in 15-20 seconds following reactor trip due to the combined actions of cross limits i

i and ETU limits. Tripping of RC pu=ps due to HpI actuation also requires that

-he operater verify the reactor has tripped. When the reactor trips, the i

ICS controls feedwater flew as described above.

page 4-1, 3rd paragraph:

The section is entitled " Reactor Protection Syste=."

The integrated control syste= (ICS) discussed in the paragraph is not part of the RPS and should be separated out.

e

o A reactor trip will not only occur upon turbine trip, but also will occur on loss of main feedvater.

The RPS low pressure trip is 1500 psi.

page 4-2, 3rd paragraph:

The Low Pressure Injection Syste= is incorrectly described.

Only two LPI pumps are started automatically. The third pu=p can be manually started and aligned to either A or 3 train.

Although the core flood tanks (accumulators) are centioned in Sections 4.4.2, 4.4.3, there is no description in the syste= description paragraph.

page 4-2, 6th paragraph:

The L?! System is actuated when the pri=ary system falls below 500 psi.

Substantial flow, however, by this system, could occur only when the syste=

pressure falls below 200 psi.

page 4-2, 7th paragraph:

While the discussion of the main feedwater control is ' fairly accurate, the discussion fails to include any mention of turbine bypass valve control and tend to limit the. potential cnly briefly discusses features of the ICS that for an overcooling event.

~~

Fur:her, there is no mention of the Emergency Feedwater Syste= and its centrols and instrumentation, which, in fact, are totally independent of the ICS.

page 4-5, 2nd paragraph:

The main feedwater pu=ps are supplied as written.

The sentence is incorrect water f rc= the condenser het well through three cendensate booster pump and three hoeve 11 pu=ps.

=akeup to the hetwell, not directly to the feedwater pumps.

page 4-5, 4th paragraph:

3 gallons, the Although the maximum inventory fro = all sources is 295 x 10 Although conden-142,000 gallons in the hotwell.

actual usable inventory is sate makeup to the hetwell can be achieved from the UST and CST, the =aximu=

is 192,000 ccndensate available for an uncontrolled main feedwater flow event gallons (or for 9 minutes at full flow rate).

page 4-5, 5th paragraph:

The total f eedwater demand vill run back The first setpoint is incorrect.

a =axi=um rate of 20% per minute to track generated =egawatts followidg a that demand.

reactor trip if the conditions in the sten = cenerator will permit at If the steam generators cannot accept a 20% per minute runback, the ETU limits g

i

-4 i

will reduce the de=and to whatever value is appropriate.

4-l The second se: point is partially described correctly; the following should be added.

The feedwater valves.will. transfer to e=ergency level control.

which compares the actual level in the steam generator to. a 50% level setpoint.

This circuitry will either open or close the startup valve as appropriate with the pu=ps controlling on D/P..

1 In addition to the listed trips, each main feedwater pu=p will trip on low suctica pressure or on overspeed.

page 4-6:

An-attempt is made to represent-functionally the =ain feedvater portion of I

the ICS. This figure should be redrawn to represent = ore accurately the control'syste=. As a =ini=u= the level li= iter should be =oved above the j

controller. and-another controller added to control the startup valve on loss j

of all RC pt=ps.

s i

page 4-7:

1 For single control failures occurring below the =anual control points also, the high level trip of the =ain feedwater pu=ps will be available to =itigate the events.

~

No discussion of the availability of'instru=entation and controls is presented.

A description of the present syste= was provided to ORNL (copy of July _23,, 1981 letter of 'a'1111a= 0. Parker, Jr. to NRC) and yet no =ention is cade of the L

cultiple instru=entation available to che operator.

The first sentence o1. page 4-7 should be changed as follows: This review divided the main feedwater portion of the ICS into three general areas, as shown in Figure a-3.

page 4-7, 2nd paragraph:

J In the seccnd sentence, =anual control is recuired following ICS failure.

Sections 4.5.4 through 4.9 use the ter= excessive feedwater on nu=erous

~

j occasions with no at:c=p: to define the a=ount of excess. Someone who does j'

not know the sys:e= =ay not understand the differences and in fact could interpret excessive feedwater to =ean the hypothetical runaway feedwater i

transient.

This should be clarified in future reports.

The lastsentence should be changed as follows:

It should be noted that with-1 ou: the steam generator high level trip for the feedwater pumps, failure of a,startup level signal to a " low" condition can result in an overfeed of one stea= generator.

page 4-3:

i In Table 4-1, it should be noted that several indicated failures cause over-

)

feed to only one stea= generator.

^

.i EGM.

e

,~,-,,,,---mm--,,--,*yy.,,em,,,~mm,.

,,,,-,,,,,,,-,nw

...,,,m_,-%,.

.,~~_,,-,,_m.,-,,,,,-mm-,y-,,,.-,,,,m-,m.,,,,,_,,--,

The Oconee-1 event sequences referred to were sub=itted to the Staff in July 1981 as part of the Abnor=al Transient Operating Guideline Progra=.

These are currently under review by the Staff.

page 4-10, Sectica 4.7:

~

Overcooling transients are alerted to the operator by nu=erous alar =s and are easily recogni ed by decreasing te=peratures and the cause identified by stea= generator conditions.

The continuance of =ain feedwater at 100% flow rate requires =ultiple ICS failures and failures o.f other flow li=itiig functions or deliberate operator action to open feedvater valves to both st.ea= generators and to disable certain trip functions.

Even then, the condition can per. st only for a short duration because it is self-li=iting (due to high SG pressure conditions or due to

' rapidly di=inishing inventory).

page 4-11, 2nd and 5th paragraphs:

Based on a detailed review of the IRT and TRAC calculations, we believe that to characterire the= as being "approxi=ately bounding" is overly opti=istic.

Chacter 5.0 Of the four Oconee events, only two events can be considered as representative initiating events of the general secondary'systd= upset condition category of

~

events of interest in reactor vessel overcooling. These two: events are the 1/4/74 switchyard isolatien event of Oconee 2 and the Nove=ber 10, 197 9-loss cf ICS pcwer event of.0conee 3.

In the Oconee 2 transient the overeocling was caused by excessive stea= lead ce=bined with a high initial design pre-scribed stea= generator level, which has subsequently been reduced. For-the Ocenee 3 event also, the major contributor was excessive stea= load (auxiliary stea: dravdewn and partially open turbine bypass valve) with sene minor con-tribution frc= overfeeding one stea: generator.

In both cases the pri=ary syste= cooldown was limited to 420CF, and even if the operator had f ailed to take acticn the transient would have progressed only to a modest overcooling event and not of the severity calculated to occur in the present analyses.

The third Oconee event (June 13, 1975 event in Ocenee 3) involved a stuck-open PORV, and' the actual thermal-hydraulic transient behavior was =ildene than the calculated small break LOCA transient'.

The fourth event involved a te=porary undercooling in Oconee 1 on Dece=ber 14, 1978.

During 'this type of an event, the pri=ary syste= undergees a rapid but finite cooling of the primary side when nor=al cooling is reestablished. The primary syste= cooldown is limited to 520 - 540 F and as such is not different from typical reactor trip events

, as far as overcooling events of interest for reactor vessel integrity are concerned.

It is worthwhile to examine the operating history of the Oconee reactors with regard to the occurrence of the "RFT" event, which is characterized by the failure of the =ain feedwater flow control system to run back feedwater flow after a reactor trip, followed by the failure of the SG high level trip of the MFWP's and concurrent stuck-open failure of the T3V Syste=, and not con-

=

sidering any operator actions. The three Oconee units co=bined have now accu =ulated 23 reactor years of cperation, during which time 136 reactor trip events have' occurred.

Our review of these reactor trip events indicates that in all cases the feedwater was run back, either prc=ptly or with acceptable delay, after the reactor trip and did not represent a perpetual full flow,

s condi:1cn. Further: ore, the SG high level. trip of the =ain feedwater pu=ps have been challenged nine ti=es'as a result of moderate overfeed conditions due to slew feedwater runback or during loss of ICS power events.

In all' cases successful trip of the syste= occurred as designed. With regard to the turbine bypass valve syste=, we have had no instances in which all the turbine bypass valves stuck open.

Although we have had a few instances involving excessive stea= icads and/or partia1' failure of the TSV Syste=, these events produced only =edest overecoling of the pri=ary syste=.

In all cases successful and ti=ely operator action has been found to occur. Additionally, it should be pointed cut that design changes have been =ade and operating procedures have been written to prevent / reduce the probability of stea= generator overfeeds (RFT).

The present response to all three of the overfeeds listed in Appendix 3 would be a trip of the =ain feedwater pu=ps which would au:c atically initiate auxiliary feedwater.

Auxiliary Feedwater would =aintain stea= generator level at 25" (240" if all the RC pu=ps trip), thereby preventing both stea: generator drycut and overfill.

Operator confusion would not result on loss of ICS power since adecuate backup instru=entation and controls and e=ergency procedures are available.

page 5-7, 1st paragraph:

~

Fluid =ixing be veen the epi and celd leg is of mini =al i=portance during evarecoling transients.

The te=perature on the downce=er RV is affected.pri-

=arily by the temperature of the fluid a: the veld location of interes and the fluid flow rate which govern :he hea: transfer coefficient.

page 5-7, 5th paragraph:

Flev distributien is importan: in deter =ining the rate of heat re=cval frc=

the RV wall and thus the :e:perature gradient in the wall. The assu=ptien of an arbitrary flow affects not only the ther:al-hydraulic calculatica but also the heat transfer frc= the wall.

page 5-7:

Additienal deficiencies.:bi the ther=:1-hydraulic predictions beycnd these identified in Section 5.3.2 are evident and are as follow:

A.

It is inappropriate to use the IRT code for any external and released applicatiens since the code is still under development. This is evident by the fact that the code does not have a =c= ente = equation and therefore all the flow rates,in the analyses are non-nechanistic. In additica, the code has not been videly used 'in the industry and its capabilities have not been de=cnstrated.

3.

The justificatien fer using IRT :o simulate a 36W configuration has not been established.

Once a code has been verified (this has not. been ec:pleted

-m

-e

O for IRT), the nodalization of the syste= being modeled must be gaalified by co=parison with ?ata fro = the syste=.being =odeled. There-is no indica-tion that this has been done using IRT on a B&W plant.

An exa=ple of this is the apparent failure to consider reactor vessel upper head circulation flow in the analyses and, also, the failure to consider the feedwater inj ection location and the pre-heating in the stea= generator.

C.

There is no indication that the analysts had the necessary inti= ate familiar-ity. vith the Oconee plant to set up a realistic and appropriate set of boundary conditions for a si=ulation.

Overcooling transients are strongly affected by boundary conditiens. Without a realistic set of conditions, the transient response will not represent the true response, and the results are essentially meaningless. A lot of simulation experience in terms of plant system fa=iliarity and knowledge of code capability and limitations are essential.

So=e exa=ples of boundary condition errors are:

a.

Incorrect feedwater flow.

b.

Incorrect turbine bypa s setpoint and capacity.

c.

Incorrect HPI actuation setpoint and flow versus pressu?o.

d.

Feedwater enthalpy versus integrated flex delivered.

e.

There are no secondary stea= relief valves or atmospheric du=pf.

f.

O=ission of control syste= responses or additional assumed failures that are not identified, e.g., high SG 1evel trip of both =rin feedwater pu=ps.

1 D.

It is very misleading to label a particular gnalysis without explicitly identifying the f ailure assumpticas =ade in the analysis. As,an exa=ple, the IRT analysis labeled, " Turbine Trip", is actually a turbine trip with a f ailure of the =ain feedwater to run back, with a f ailure of the high SG 1evel trip of the =ain feedvater pu=ps, with a failure of the turbine i

bypass valves on both steam generators, assuming a rapid decrease in feed-water tenperature, and assuming a failure of the operator to terminate the overcooling or perfor= any other =itigative action. The assu=ptions which determine the transient respense should not be lost in the genetstion of plots of results, and neither should the linitations of the code utilited.

[

Chaoter 6.0 page 6-1, Section 6.1, last paragraph:

Although the variations in Table 6.1 do occur in source parameters, they are not necessarily uncertainties in the calculation of fluence. Many cf these ite=s are accounted for in the caltulational procedure. For exa=ple, cycle and cycle-to-cycle cors power distributions are averaged over the capsule irradiation period with the use of PDQ generated power distrflution data at selected time intervals during fuel cycles.

The basis of these stata=ents is experience in the analysis of 12 capsules fro 8 3&W reactors.

~

s page 6-1,.6-2, Section 6 2.1, 1st paragraph:

Although this procedure was used to calculate fluence frc= Ocenee l' capsules OCIF and OCIE, an i= proved proced;ure 1s presently being used which incorporates the capsule sec=etry and P scattering cross sections directly into the r-G 3

reactor =edel, thereby eliminating the need for corrective f actors.

page 6-3, Tabl,e 6.2:

An i=portant step was o=itted, that of normalization of calculated flux to flux derived f c= =easured dosi=eter activities.

This table should read:

j.

Calculate capsule flux (E>l MEV) by =ultiplying the value frc= the :-G

=edel ti=es the F /21 and capsule perturbation factors and ti=es an axial 3

shape factor based on the axial power shape in a peripheral fuel assembly.

k.

Obtain a nor=alization factor frc= the ratio of flux (E>.1 MIV) derived-frem desi=eter reactions to calculated flux (E>l >2N) in the capsule.

1.

Perfer= an axial 2-D, P, :-: calculation.

1

=.

Correct flux values frc= the :-G :odel with the F ! 1, capsule pereurbation, 3

axial shape, and nor=alization factors.

n.

For veld locations, displacement factors frc= the :-: =cdel and r-G :odel are applied to the vessel flux (or fluence).

page 6-5, 6-6, Section 6.2.3, last paragraph:

The spread in normalizing facters is =isleading with respect to calculational u:.cer:sinties because only fissien reaction data frc= the CCIE capsule vere used :c calculate fluence.

Data frc= CCIF were disecunted because of suspected errers ir activity measurements.

This was the first capsule analy:ed at 36W and such large discrepancies have not been observed in any subsequent capsule analysis.

page 6-6, See: ion 6.3, first paragraph:

The uncertain:y evaluation in BAW-1485 was primarily based on conservative esti=ates with relatively little experience. Thus values of i 30% for predicted beltline region fluence and i 50% f or certain weld locations, were reported.

Since then, 3&W has participc:ed in the Blind Test, a calculational benchmark spcasored bv the Light Water 7,eactor Pressure Vessel Desimetry Improvc=cnt progra= (S~RC funded) and the OCIF fluence calculation has been checked by ancther phase of the,WRPVDIP.

(R. L. Si= ens at HEDL did the analysis.)- The Blind Test indicatec tha: the B&W transport calcula:1cnal procedure vould produce a fast flux (E>l MIV) that deviated < 5 fro: a normalized capsule location to vessel surface and T/4 locations. The HEDL calculation of capsule fluence was 50 greater than the 35W calculated value.

In addition, analyses of 12 capsules frc= 3 25W reactors have consistently shown E/C values within 1 10" for fission 9

e

~.-

3ased on these developments, recent estinates of fluence t.ucer-reactions.

the capsule, i 15% in the vessel for time periods ccrres-tainties are 1 10% at pending to capsule irradiation periods, and i 18% for predicted fluence in Co= parable values _for vessels in reactors without capsules are It =ust be e=bhasized that these are conservatively estimated _

the future.

and + 21%.

+ IS:

values in the absence of a detailed uncertainty analysis.*

The i 50% value reported in 3AW-1485 for veld locations was intenced to indica displacement the added uncertainty (above 130%) of-using axial and azimuthal A displace-Apparently, this was =isunderstood in the ORNL analysis.

f actor of.89 (as is used for the critical weld location in the Op5L factors.

i analysis) cannot be in error = ore than -12% when conpared to the beltline reg on

=ent When statistically combined with the vessel uncertainty of + 30%,

fluence.

this would result in a + 32% uncertainty.

pages 6-6, 6-8, Section 6.4-The i= plication that there has been no verification of the 35W calculational In fact, 3&W has procedure fer the de' termination of fluence is incorrect.successfully E= prove =ent Progra= to bench = ark both the calculational procedure and the desi=eter =easure=ent technique.

page 6-7, Table 6.5:

Data in this table apparently are based on ex[rapolation of the. fast averaged ever cycles 1 and 2.should be based on a predictive procedure described in Ter Oconee 1, the use cf :his procedure is particularly important her.c.use of tion (in tine) d tion a conversion to an 18-month fuel cycle in cycle 6 with a corresponding re uc in ex-core fast flux of approxi=ately 30%.

Outside of Inside of RV Wall T/4 3T/4 RV Wall 2.20E+18 1.22E+18 2.36E+17 1.053+17 2.67E+1S 1.t.SE+18 3.47E'17 1.27E+17 3.26E+18 1.51E+18 4.23E+17 1.55E+17 3.66E+18 2.03E+18 4.75E+17 1.74E+17 3 asis is assu=ption that relative ef f ect of 18-nonth cycle in ASO-1 will be Predictive data are available for Oconee 1 through cycle 7 sa=e in Oconee 1.

the calculatiens have not been made.

but Chapter 7.0_

page 7-1, Section 7.1, Table 7.1:

Detailed descriptions of all data used and certification that such data areAlso, err been provided.

appropriate for those analysis have not for input date has not been provided.

a n-.%,..

l '

page 7-3, first paragraph, next to last sentence:

The basis of the statenent that uncertainty is not large in the parameters 1

included in ASMI Section III is not provided.

j j

iirst paragraph, last sentence:

This sentence conflicts with the_ previous state =ent.

Data should be to support-this position: Explanation should be given as to how such data relate to the i

K;g data which are used to evaluate vessel integrity. Also, data for the uncertainty in the deter =1 nation of RTNDT should te provided.

second paragraph, first two sentences:

These two sentences appear to be in conflict. They should be clarified and supported with actual data to substantiate the opinion expressed in this para-graph.

second paragraph, last sentance:

The reference to support this statement should be provided.

third paragraph, first sentence:

The reference to support this statenent should be provided.

third paragraph, last sentence:

This statement does not recognize the Ocenee Unit 1andthe3&WOwnersG$I oup p.esearch Prcgram which is in p cgress and is generating this data.

page 7-4, second paragraph:

This paragraph appears te express an eninien which should be based en sound data.

Since the state =ent is made that Regulatcry Guide 1.99 is "not exces-sively conservative" for Ocenee 1 veld metals, the data should be either pre-sented er referenced so that a better definitien of " excessively censervative" can be better understeed.

second paragraph, last sentence:

Over 35 surveillance capsules have been removed frem pcuer reactors and the data support the conservatism of Regulatory Guide 1.99.

As for irradiation pregra=s at test reactors, most of these are co=pleted and the data are available.

page 7-5, third paragraph:

The statement regarding reduction in upper-shelf energy of weld cladding is

=isleading.

No mentien is =ade of the fact that these fluences are well above that expected at EOL of any operating PWR.

At the fluence levels predicted, the cladding is expected to lessen the degree of crack propagation.

~

page 7-5', Reference 3:

This infor=ation should be included in the analysis and should not be stated as a reference since it represents a significant reduction in EOL fluence.

Chapter 8.0 page S-2, first paragraph:

It is stated that if the te=perature of a =ajor portion of the coolant in the pri=ary syste= is above 212cF, the opening due to crack propagatien =ay be excessive and core cooling nor =aintained.

It is interesting to ncte that RFT and P.5L3 ::ansient show bulk te=perature decreasing to below 200 F.

Even with the considerable errors in the transients analyses provided, it could be postulated that the copious a= cunts of LP injection available would be =cre than sufficient to =aintain the core cool, in =uch the sa=e way as it is predicted to occur during a postulated L3LOCA.

page S-2, second paragraph:

It should be ncted, again, that only the vessel =aterial properties are approxi-

=ately representative of Oconee 1.

The accidents analyzed are not a: all

. represe=tative of Oconee 1 cr any other plant with E&W NSSS.

fcurth paragraph:

For evercooling events, little if any vent valve flow vill occur because_there is =ini=al dif f erential pressure between the core cutlet and inlet. The thermal analysis is dependent en the downce=er temperature and the flew cend'.:icas.

It is not apparent what flew conditions were assu=ed in the thereal-hydraulic calculatiens and thus what was assumed in thermal analysis of the RV wall.

page S-3, secend paragraph:

The CRNL analysis ignores axial gradient in fluence. Axial gradient of fluence is utilized in the 3&W analyses.

The assu=ption of a pre-existent icng sharp crack is unrealistically conserva-tive.

This is particularly true for Oconee 1 which recently underwent a 100%

j exa=ination of beltline region welds with no cracks indicated.

page 8-4, third - fifth paragraphs:

No basis is provided to support the assu=ption : hat the fluid-fil= heat-transfer coefficient is 1000 Etu/hr-ft2. F.

1: is stated that this correspends to full-flev conditicas, but 1; is not stated what flow conditiens were ac:ually assumed

~

in the transient calculatiens.

It is incensistent to asst =c one mode of system operation during the transient calculation and a heat transfer coefficient based on a different = ode of operation. This is particularly important in that severe overcooling transients =ay interrupt RCS flow and thus reduce heat ::ansfer, and with RC pumps assu=ed running, a finite c=ount of heat is in fac: added to :he RCS.

~ -

c page 8-6, Table 8.1:

The CRNL analysis should have utili:ed actual weld parameters inasmuch as these data were provided by Duk,e.,

The,RTUDT is that of the base metal rather than the weld and the che=istry is that. or a hypothetical weld metal.

page 8-13:

s It is inappropriate to perform the fracture mechanics analyses of the IRT steam line break of RFT cases with all their known deficiencies and atypical-ities.

The analysis, discussion and results for these two cases should be deleted.

page 8-16, Section 8.4, last paragraph:

The fluence rate of.046 x 1019 2

n/cm /EFPY is appropriate for the critical weld location (lower long weld at inner surface of vessel) for the first 4 EFPY (through cycle 5).

month fuel cycle, a better value for >4EFPY exposure is.033 x 10.gThereafter,becauseofth n/cm the critical weld location.

Also, as noted previously, the 150% uncertainty does not apply to welds close to the beltline region.

Based on the initial analysis at the time EAW-1485 was written, this value would be i 32 ; more recent 'esti=ates-indicate about i 20%. These are esti=ates because a detailed uncertainty analysis has not been perfor=ed.

page 3-17, Table 8.6:

The column with WPS and without W?S appears to have the line ites designatiens reversed. W75 should provide additional tine to fracture in all cases except L3LOCA.

Table 8.7 appears to have this relationship correctly ident1fied.

page S-19, Table 8.7:

The revised values of fluence /EFPY to account for the 18-nonth fuel cycle will lengthen the threshold time calculation fcr times >EFPY.

a Threshold Time (EFPY_I)

Co==ents and Qualifications 26 V

26 WF3 not assumed effective (44EFP if it were).

Through-wall crack predicted.

19 2

b3ased on fluence accu =ulation rate value of 0.046 x 10 n/cm /EFPY for the 19 n/c=2/EFPY thereafter.

This value may have an initial 4EFPY and 0.033 x 10 uncertainty of as =uch as 132%.

_13-t Chapter 9.0 page 9-1,

-2, Mitigative Measures:

In view of 'the inherent weaknes's ' contained in the report, it-is considered that identifying the need for any changes is pre =ature.

As a point of clarification, increasing the 3WST temperature would have essentially no effects on reducing the severity of an overcooling trancient.

Because a 53LOCA was not performed, it is not clear what basis there is for stating that an increased 3WST temperature would reduce the degree of over-cooling caused by actuation of F.PI.

In fact, with vent valve flow and plant specific analyses, the ther=al shock concern for S3LOCA is =inimized.

. page 9-2, sixth paragraph:

The practicality of in-place annealling is overstated. While the principle-

=ay have been de=enstrated under centrolled laboratory conditions, extensive work is nacessary to i=ple=ent in-place annealling on an operating reactor-vesse?.

Extensive evaluations are necessary to de=enstrate the acceptability of annealling at te=peratures of 7500-850 F if a plant and its support syste=s were designed to lesser te=peratures, page 9-2, Sectied 9.2.2, General Changes:

An additional change that is presently occurring in 3&W plants that will sig-nificantly lower the fluence en the reacter vessel wall, is the result of geing to le-=enth LEP relcad cycles.

Once-burned fuel is leaded en thE periphery of the core thus lowering the peripheral fuel asse=bly's pcwer and the corres-pending leakage flux cr fluence to the vessel. The results of this are noted in the references of Chapter 7, 2eference 8.

Chanter 10.0 page 10-1, Section 10.1:

See previous cc==ents en fluence analysis.

Section, Concluding Re= arks:

All re= arks are negative in context and are leading to uncertainty cnd lack of confidence in the final results.

Yet, state =ent is =ade that, "Nenetheless, for all their shortcomings, the analyses at hand are the best presently avail-able on a nenpreprietary basis, and... [= erit] a great deal more study using refined techniques."

page 10-2, Section 10.2:

The.i= plication here is that fluence calculations in general have uncertainties in the range of + 30 - + 50%. The uncertainty in the Ocence 1 fluence calcula-tiens is =uch snaller as discussed in the ec==ents on Chapter 8.0.

cm a