ML19210A812

Forwards B&W 770924 Final Rept on Examination of CR-3 Decay Heat Pump Shaft Per 770908 Commitment.B&W Recommends Shaft Replacement at TMI-1.Met Ed Studying Feasibility of Metallographic Replication & Hardness Tests

ML19210A812
Person / Time
Site:	Three Mile Island
Issue date:	09/28/1977
From:	Herbein J METROPOLITAN EDISON CO.
To:	Reid R Office of Nuclear Reactor Regulation
Shared Package
ML19210A813	List: ML19210A812 ML19210A814
References
GQL-1317, NUDOCS 7910310701
Download: ML19210A812 (4)

Text

DISTRIBUTION AFTER ISSUANCE OF OPERATING LICENSE gissicN 00yo TN I

' u.s. NuCLsAn masutArony

NR -

UMBEma gg i p.:.C,Popu . 195, l NRC DISTRIBUTION Pon PART 50 DCCKET MATERIAL l FROM: CATE OF CCCUMENT

TO: ,

Metrcp 11 tan Edison CCepany 9/28/77 {

"r. R. '4. Reid Reading, Pa. cAra macasvs3 l J. G. Herbein 10/3/77  !

CNOTCalZIO PmCP INPUT PCmM NUM8Em CP CCPits mECflVf D JusTTan uC=rv mlGIN A L UNC LASSIFI E D

/ ,: vane /f ENCLCsumE /

l :ssCateTICN i Consists of the PM Metallurgical Evaluation

and Failure Analysis Report on the Crystal i Rive r Unit 3 f aile d shaf t .. . . . . . . . .

l l

I (3-P) (1/4")

PLANT NA'.iE: Three Mile Island U 1

{% P+ho%y u

.c h >ill sen+niNNo%be  %.k- o l .%w .

'-L- n L. ,

i b o 4 t.

• Rae ~ R. 1m m.m "

le nI e n ). re c k /

S AFE*"( , FOR ACTION /lNFORMATION / I

! l BRANCH CHIEF: (7) I M /d l l l I t {

I l 1

I i i i is  %

l I [ ] INTERNAL DISTRIBUTION  !

l L 7er, e T i -- > [ i l l l 1NRC PDR l I i IICE (2) '

I IOELD I i iHANAUER , I l t i CHECK t l l I 1 iSTELLO I I I iEISENHUT l {

l lSHA0 { l 6  !

I 'SAER l l l l l SUTLER l l l I GRIMES l l l 1 lJ. COLLINS I l l l l } MORC \ CD I. I ktD. Have l+od vs / chA fas i i t) ' I i i i  !

I l l i I l I l l 1 i I i i l i i i i l , , EXTERNAL DISTRIBUTION CCNTR CL NUMBE R -

1 I LPDR: fVM#/$lku"D . 84 I I i f- '

ITIC #- '

l

] N, P .-

NSIC '

77A 7170 o/ 3 16 CYS ACRS SENT CATECORY 6 l

!  !! ll 7 010s10 7o 4 p l

B ll=~f? U

.w

/ me-,

Aeect 900 HOC #ISS METHOPOLIFAN EDISON COMPANY POST OFFICE Box 542 READING. PENNSYLV ANI A 19603 TELEPHONE 215 - 929G01 September 28, 1977 GQL 1317 O .

^

Director of Nuclear Reacter Regulation .- -

Attn: R. W. Reid, Chief Operating Reactors Branch No. 4 '4 p.J .'D7/

U. S. Nuclear Regulatcry Cc==ission 6 ,'.* 7 Washington, D,C. 20555  %

Dear Sir:

Three Mile Island Nuclear Statien Unit 1 (TMI-1)

Docket No. 50-289 Operating License No. DPR-50 Decav Heat Pump Shafts In accordance with cur September 8,1977 cc=mitment please find attached the B&W Metallurgical Evaluation and Failure Analysis Report on the Crystal River Unit 3 failed shaft. In su==ary, the report indicates that:

1. The crack initiated along a bottem corner of the keyvay. Crack initiation appears to have been due to fatigue.

2. The direction and type (shear mode) of initial crack growth is censistent with stresses frc= torsional loading.

3. The crack =crphology is not censistent with crack initiation or growth by impact overload or stress corrosien cracking.

h. The shaft does not appear lave received a proper H1150 heat treatment , but rather appeL . to have received a heat treatment sc=ewhat like an H1025 heat treatment.

5 The questionable heat treatment and the ncn-filleted keyvay are the metallurgical factors that led to the failure of the Crystal River III Decay Heat Pump Shaft.

Met-Ed, GPUSC, and MPR have evaluated the B&W Metallurgical Report; pu p performance data accu =ulated by both Worthingtc.. and Met-Ed ; the results of the pump design review being perfor=ed by MPR, S&W and Met-Ed; Armco material data corre1r. ting fatigue properties of 17 h PH with h1h0 alloy steel; and the material documentation verificatien that was performed.

Key points of th evaluation are cutlined belev:

1 .JI ^G ')/

ll } 792-7 ?'O O/3

Rr W. I Sid, Chief September 28, 1977 GRL 1317

1. Worthington assured Met-Ed in a meeting on Septe=ber 20, 1977 that thousands of this type of pump have operated satisfactorily with no reported cases of shaft failures. Worthington stated that all of these pump shafts have had the sharp cornered keyways sbnilar to those believed to be in the TMI-l pump shaft. Dese were fabricated fram hlh0 alloy steel. Worthingten has advised us that 161 pumps are in service that are the same size as the CR-3 and TMI-1 decay heat pumps. This broad base of experience gives us confidence based on operating history that the TMI-l Decay Heat Pump design is of proven reliability with a large margin of safety.

2. The shaft material of the CR-3 and TMI-1 pumps is 17-4 PH specified in the H1150 heat treated condition. The B&W Metallurgical Evaluation and Failure Analysis indicates that the CR-3 failed shaft properties were somewhat similar to the H1025 heat treated condition. Armco fatigue strength data shcws that 17 h PH material in the H1025 heat treated condition has a higher longitudinal and transverse, s=coth or notched, fatigue strength than the h1h0 alloy steel used by Worthington for their standard ec==ercial pu=ps of the same design.

We expect that all H1150 fatigue properties would improve over these of the H1025 heat treated material documented for smooth bar longitudinal specimens. Due to the fact that 17 h PH material in both the H1025 and H1150 heat treated conditions is superior in fatigue properties to the h1h0 alloy steel used by Worthington for their standard cc==ercial pump of the same design, we have concluded that the excellent cperating experience with the h1h0 alloy steel shaft is applicable to the THI-l Decay Heat Pu=p Shafts . 3 The TMI-l pump shafts were fabricated from a different heat of material about a year before the CR-3 shafts. We have confidence that the TMI-l shafts were heat treated correctly.

h. Load and stress analyses performed by MPR and Worthington indicate that applied cyclic stresses on the shaft are considerably less than vculd be expected to cause fatigue failure. While there is some uncertainty as to the loads that are experienced under lov flow conditions, the credible upper bound stresses are vell belev the fatigue endurance limit, for 17 h PH material heat treated to H10?5 or higher.

5 The TMI-l pu=ps have already operated satisfactorily abcut 3 to h times that length of time which caused failure of the CR-3 pump. Ultrasonic inspections have confirmed that there are no transverse defects greater than 24 mils deep in the TMI-l shafts. The vibration tests of the TMI-1 pumps shev that they are quiet, smooth running pu=ps with very 1cv levels of vibratien. Based on these vibration and UT inspections we concluded that these pumps are not subjected to high loads. B&W has recommended decay heat pump shaft replacement at TMI-1, because the TMI-l Decsy Heat Pump Shafts probably have sharp corner keyvays; docu=entation is not available to substantiate the IMI pump shaft heat treatmentl; and E&W's metallurgical analysis concludes that the CR-3 failure was due to i=prcper shaft heat treatment interaction in ec=binaticn with the sharp corner keyvaf. 1 Records prior to 1971 were destroyed in accordance with the previcus Worthington practice of 5 year record retention. IIO av/ )

, Rs W. Beid, Chief September 28, 1977 GQL 1317 In view of B&W's shaft replacement recc==endation, Met-Ed is currently investigating the feasibility of metallographic replication and hardness tests to verify the heat treated condition of the TMI shafts. We are also considering the advisability of shaft replacement during the spring 1978 refueling. It is Met-Ed's present intent to acec=plish the above tests and analyze the data within the next 6 weeks. Sincerely,

                                          /

J. G. Herbein Vice President JGE:WEP :tas

Attachment:

33W Metallurgical Evaluation and Failure Analysis Report rau^ ") }n is['6

              .}}