ML20236M922

Safety Evaluation Supporting Util Proposal Re Elimination of Postulated Primary Loop Pipe Ruptures from Design Basis

ML20236M922
Person / Time
Site:	Three Mile Island
Issue date:	11/05/1987
From:	Office of Nuclear Reactor Regulation
To:
Shared Package
ML20236M912	List: ML20236M909 ML20236M922
References
NUDOCS 8711130333
Download: ML20236M922 (6)

Text

__

y f[ hg UNITED STATES NUCLEAR REGULATORY COMMISSION

-l WASHINGTON, D. C. 20555

\...../  ;

SAFETY EVALUATION BY 1HE OFFICE OF NUCLEAR REACTOR REGULATION I

ELIMINATION OF POSTULATED PRIMARY LOOP PIPE RUP1URES AS A DESIGN BASIS GPU NUCLE,AR CORPORATIO,N THREE M]LE ISLAND NUCLEAR STATION UNIT 1 l

DOCKET NO. 50-789 BACKGROUND By letter dated Parch "O, 1987, GPU Nuclear Corporation (the licensee) contended that the dynan,ic effects associated wfrth asymmetric loss-of-coolant j accident (LOCA) loads could be eliminated from Three Mile Island Nuclear Station Unit 1 (TPI-1). As permitted by the revised General Design Criterion 4 (GCC-4) of Appendix A,10 CFR 50, the licensee may use fracture nechanics

" leak-before-break" (LBS) techndogy to elininate the dynamic effects of l postulated primary loop pipe ruptures from the design basis of TMI-1. By letter dated August 21, 1987, the licensee. submitted the technical basis for the elimination of primary loop pipe ruptures for TMI-1 in Babcock & Wilcox (B&W) report BAW-1999 (Reference 1) in response to the staff's request by letter dated July 2, 1987. The licensee also referenced B&W Owners Group (B&WOG) reports BAW-1847, Rev. 1 (Reference ?), and BAW-1889P (Reference 3),

which have been reviewed previously by the staff (Reference 4).

)

The revised GDC-4 is based on the development of advanced fracture mechanics technclogy using the LBB concept. On April 11, 1986, a final rule was published (51 FR 12502), effective May 12, 1986, amending GDC-4 of Appendix A, 10 CFR 50. The revised GDC-4 allows the use of analyses to eliminate from the desigr. basis the dynamic effects of,costulated pipe ruptures of primary coolant loop piping in pressurized water reactors. In the "sumary" section of the final rule, it is stated that the new technology reflects an engineering advance which allows simultaneously an increase in safety, reduced worker radiation exposures, and lower construction and maintenance costs.

Implementation pemits the removal of pipe whip restraints and jet impingen,ent barriers as well as other related changes in operating plants, plants under construction, and future plant designs. Containment design, energency core cooling, and environmental qualification requirements are not influenced t'y this modification. In the " supplementary infomation" section of the final rule, it is stated that acceptable technical procedures and criteria are defined in NUREG-1061, Volume 3 (Reference 5).

a711130333 871105 DR ADOCK0500g9

-T-Using the criteria in Reference 5, the staff has reviewed and evaluated tFc licensee's- submittal for compliance with the revised GCC-4 This Safety Evaluation Report provides the staff's findings.

TMI-1 PRIMARY LOOP PIPlHG The TMI-1 primary loop piping consists of T8-inch and 36-inch nonin61 diameter straight sections and elbows. The piping materials in the primary loops are-low alloy ferritic steels (SA-106 Grade C for straight sections and SA-516- )

Grade 70 for elbcws) and wrought stainless steel safe ends (SA-376 TP 316). '

The welding processes used were submerged arc (SAP) and shielded rtetal arc-(SMAU). The material properties were provided in Reference ?.

STAFF E1A,t,t!AT1,0,N,,C,R1,1E,R1 A The staff's criteria for evaluation of compliance with the revised GCC-4 are- -

discussed in Chapter 5.0 of Reference 5 and are as follows: 1 J

(1) The loading conditions should include the static forces and moments I (pressure, deadweight, and themal expansico) due to normal operation, i and the forces and moments associated with the safe shutdown earthquake (SSE). These forces and moments should be located where the highest .l stresses, coincident with the poorest material properties, are induced q for base materials, weldrents, and safe ends, i

(2) For the piping run/ systems under evaluaticn, all pertinent information j which demonstrates that degradation or failure of the piping.resultinc l from stress corrosion cracking, fatigue, or water hamer are not likely, i should be provided. Pelevant operating history should be cited, which includes system operational procedures;. system or compor!er.t r' modification; water chemistry parameters, limits, and controls; and resistance of 1 material to various foms of stress corrosion and perfomance under I cyclic loadings. j 1

(3) The materials data provided should include types of materials and materials specifications used for base metal, weldments, and safe ends; the materials properties including the fracture mechanics parameter "J-integral" (J) resistance -(J-R) curve used in the analyses; and i long-term effects such as themal aging and other limitations to valid  ;

data (e.g., J maximum, and maximum crack growth).

(4) A through-wall flaw should be postulated at the highest stressed

l locations determined from criterion (1) above. The size of the flaw should be large enough so that the leakage is assured of detection with at least a factor.of 10 using the minimum installed leak detection capability when the pipe is subjected to normal operational loads.  !

(5) It should be demonstrated that the postulated leakage flaw is stable under nornal plus SSE loads for long periods of time; that is, crack iI

l growth, if any, is minimal during an earthquake. The margin, in terms of I applied loads, should be at least 1.4 and should be determined by a flaw j stability analysis, i.e., that the. leakage-size flaw will.not experience i unstable crack growth even if larger loads (larger than design loads) are.

applied. This analysis shculd demonstrate that crack growth is' stable ]1 and the firal flaw size is limited, such that a double-ended pipe-break l will not cccur. j (6) The flaw size shculd be determined by comparing the leakage-size flaw.toL the critical-size flaw. Under nomal plus SSE Icads, it should be  !

demonstrated that there is a margin of at least P between the leakage-size flaw and the critical-size flaw to account for.the uncertainties inherent in the analyses and leakage detection capability, j A limit-load analysis may suffice for this purpose; however, ar. i elastic-plastic fracture nechanics (tearing instability) analysis is preferable .

l STAFFEV,AL,UATJph The staff has evaluated the information presented in References 1. through 3 for conpliance with the revised GDC-4 Furthermore, the staff performed independent flaw stability computations using inforn.ation provided by the .

licensee in Referu.ces 1 through 3. The staff's independent calculations were- l based on an elastic-plastic fracture nechanics analysis developed by the staff '

(Reference 6). The analysis was implen+nted in a computer progran' that p(erfoms Referencea 7).

tearing nodulus analysis and accounts for incremental crack growth l Although the staff disagreed with certain assumptions used by the licensee in the flaw stability analyses, the staff finds the TMI-1 primary locp piping'in <

como11ance with the revised GDC-4 based on the results of the-staff's independent calculations. The following paragraphs'in this section present the staff's evaluation. i l (1) Nortral operating loads, including pressure, deadweight', and thernial l"

exparsion, were used to detern.ine leak rate and leakage-size flaws. . The flew stability analyses performed to assess margins against pipe rupture at postulated faulted load conditions are based on normal plus SSE loads.

Leak-before-break evaluations were perforned for the limiting location in g each size of piping.

(2) For B h* facilities, there is no history of cracking failure in reactor coolant system (RCS) primary loop piping. The RCS primary loop has an cperating history which den.onstrates its inherent stability. This includes a low susceptibility to cracking failure from the effects of corrosion (e g., intergranular stress corrosion cracking), water.hamer,- 1 or fatigue (low and high cycle). .This operating history totals over 52 reactor-years spanning 13 years 'of operation, i

l 1

1 4

(3) The material tensile and fracture toughness properties were provided in 3 Reference 3. The material fracture toughness was described by "J " and l "J ," based on the original " deformation theory J" and the recentlh ,

inkroduced"modifiedJ",respectively. The licensee used Jy in the flaw l stability evaluation. Fokever, in a recent meeting held at the David Taylor Naval Ship Research and reveloptrent Centet at Anrapclis, Maryland, on August 5 and 6,1987, sone recently obtained material J, data were l discussed. The meeting was attended by rnen.bers of the elaktic-plastic .

fracture nechanics testing er.d analysis connunity, including  !

representatives frorr PE. The recon,n,endation of the meeting was that further research and standardization of the J pparaneter are required.

Thus, the staff considers it inappropriate for the licensee to apply the t material Jn in the current LBP. evaluation. To be conservative, the staff selected the lower-bound weld toughness JE (Peference 3, Specinen No.

ET631) fron the weld toughneps data in Re.erence 3 for the staff's  ;

independent evaluation. {

I Furthermore,thelicenseeusedanaverageDIaterialstress-straib relaticoship for the flaw stability evaluation. The staff considers it l j

more conservative tc use a lower-bound traterial stress-strain i relationship for the flaw stability evaluation. Since the material at the limiting location is SA-106 Grade C, the staff used the stress-strain data for this material in Table 5 of Reference 3 and reduced the stress values by IST in the staff's independent evaluaticn. The 15% reduction j was deterniined by the staff to be sufficient to provide. a lower-bound i estimate based on the reported data. The staff compared the repot ted yield and ultin. ate strengths with the corresponding ASME Section III Ccde minirrun. values and found that the Code n.inirnun. values were approximately I5T less than the reported values.  :

(4) TMI-1 has a RCS pressure boundary leak detection system which is consistent with the guidelines of Regulatory Guide 1.45 such that a leakage of one gallon per minute (gpm) can be detected. The licensee used the average stress-strain relationship in the leak rate evaluation. l The calculated leak rate through the postulated flaw is large relative to the staff's required sensitivity of the plant's leak detection systems; I the margin is at least a factor of 10 on leakage and is consistent with l the guidelir.es of Reference 5.  ;

i l (5) In the flew stability analyses, the liter.see performed separate i evaluations usir.

properties ard (g ii) (i) thethe base weld netal metal stress-strain stress-strain and and toughress toughness properties. The staff considers it more conservative to use the combination of the base netal stress-strain properties anc the weld netal toughness properties for flaw stability analyses. Thus, in the staff'c l independent evaluation, the lower-bound base rhetal stress-strain relationship and the lower-bound weld metal toughness Jn properties, as ,

discussed in item (3) above were used. Since the fract0re mechanics l driving force " applied J" calculations were based on Jp , it was )

l t

I

_--_._m_

consistent to use the material toughness il . Furthermore, the licersee's.

calculaticnsdidnotaccountforcrackgrohth. The staff's independent' l calculations accounted for increniental. crack growth. ,l J

From the results of the staff's independent flaw stability analyses, the I margin in ternts of applied loads exceeds 1.4 ar.d is consistent with the j guidelines of Reference 5.  !

(6).Similar tc item (5) above, the margin between the leakage-size flaw ard ,

the critical-size flaw was evaluated. Front the results of the ~ staff's ]

independent flew stability ar.alyses, the n:argin in terms of flaw size j exceeds P and is consistent with.the guidelines of Peference 5.

l STAFF _C,C1CLpSI,0p,5 The staff has reviewed the inforn$ tion subnitted by the licensee and has l perforned independent flaw stability computations. Althou'gh the staff .

l disagreed with certain assumptions used by the licensee in.the flaw stability-analyses, the' staff concludes that the TMI-1 primary loop piping complies .with the revised 00C-4 according to the criterie in NUREG-1061, Volume 3 (Reference

5) based on the results of the staff's independent calculations. Thus. the probability cr likelihood of large pipe' breaks. occurring in the. primary.

coolant system loops of-TMI-1 is'sufficiently low such that cynamic effects associated with pcstulated pipe breaks need not be.a design basis, i l

l l

I  !

1 l

)

l l

]

e -

-c-REFE_RENCES

-(l) B&V Feport BAW-1009, TMI-1 Nuclear Power Plant Leak-Before-Break Evaluation.of Margins Against Full Ereak for RCS Primary Piping,' /pril 19P7.

(2) R&W Owners Group Repcrt RAW-1847, Rev. 1 "The.B&W Owners Group-Leak-Before-Break Evaluation of Fargins Against Full Break for RCS Primary Piping of B&W Designed NSS", Septen:ber 1985. .

(3) B&W Cwners Group Report BAW-1889P, " Piping Vaterial Properties .for B&WOG Leak-Before-Break Analysis of RCS Main Piping", October 1985, B&WOG Proprietary.

(4) letter froni D.' M. Crutchfield of NRC ~to L. C. Oakes of B&W Owners Group, ,

dated December 12, 1985. t l

l (5) f:UREG-1061, Volume 3 " Report of the U. S. Nuclear Regulatory Commission:

Piping Review Conmiittee, Evaluation of Potential for Pipe Breaks",

November 1984 (6) NUREG/CR-4572, "NRC Leak-Before-Break (LBB,MRC) Analysis. Method for .;

i Circumferer.tially Through-Pall Cracked Pipes Under Axial Plus Bending 1 l Loads", Pay 19P6. l i j (7) " Fracture Mechanics Handbook and URCPIPE-Computer Code", Degraded Piping-Program - Phase II, Contract No. NRC-04-84-103,: Battelle's Colurrbus 1 l

Divisict:, (In Progress).

I Dated: MV 6 51987 l Principal Contributor:

S. Lee l

1

.j l

l

^

L__ _ _ _ _ _ _ . -