ML19308D831

Forwards Engineering Branch Decay Heat Removal Pump Shaft Failure Evaluation Rept. Corrective Measures Acceptable

ML19308D831
Person / Time
Site:	Crystal River
Issue date:	09/01/1978
From:	Noonan V Office of Nuclear Reactor Regulation
To:	Grimes B Office of Nuclear Reactor Regulation
Shared Package
ML19308D832	List: ML19308D831 ML19308D833
References
NUDOCS 8003191017
Download: ML19308D831 (3)

Text

m ~

.<n.. :.w-W. y. -,+:n.;.r - ew+,p

. w m.; ;r,

. a ~ z vw.

-. ; x n mw, v,nm m-

..s

.r

.w... -,,.

c, E. y...w y.s e.., e, e. m m a.3..c w

e : :.,...

.it.%,e

.o. cm,

L.;. f." w %., X.. e -

w

e..u

..++r*-

.m m

. - ~,

1.

.s

- e,i *,,' h @

9. **., i bl..^ w us,7 f..

j... s e.--*-a-..

m

2w,-s;

{

",;* ': - ~~ q+ e+

r a

a.. '- -

m'

~ G.- w ^ '.

2 -- ~

_.v-W.W" r

~*~ < :;w~s? rc x:&,, m&+ a.vm

'.'.2.'.

n tw e

. <;'.T

. h :.

..).~.p

.- Q<

~ :m-

?. ;."~

~~

']

, - w.,,,(,,

. ~ ~..-.

~.

' '.:,y

• 5 86 s

f

' ~

=

MEMORANDLH FOR:

B. K. Grimes. Assistant Director for Engineering and Projects Civision of Operating Reactors FRCH:

Y. S. Noonan, Chief, Engineering Branch Division of Operating Reactors

SUBJECT:

CRYSTAL RIVER UNIT 3 - REVIEW OF DECAY HEAT REMOVAL PUMP SHAFT FAILURE Plant Name: Crystal River U' nit 3 '

NSSS Vendor: Babcock and Wilcox Decket Number:

50-202 Operating Re ctors Erench and Projact Mar.agrr-GRB-4, CL Nelsen Engineering and Projects Branch Involved:

Er.gineering F, ranch Description of Task: Revicu of Decay Heat Removal Purg Shaf t Failure Review Status: Conplete The Engincarirg Brar.ch, Divisicr. of 1perating Mctorc, has nyiewad 6ta sutaitted by F1 rida Pcuer Ccrpcratian en ?he Ctcay Feat &cul Pungs rip-la, and 15-13, r:;arding the i? rep chaft failure an.i action, taken to r:~.cy the unreliable perforrance of these pumps.

The data was sulr.itted by correspcndence dated July 7,1978. M nt 10,1c73. /.ugust 30,1970 and August 31, ic73 in response to cur reqmc: of Junc 7, I E.

!n find the teatures Men to establisi, operaMi tt

f mi reliabil'ty 6f tha Oscay l'e3t R c;1al Fucps. DHP-14 and DMP-1B are acceptable.

Attached is our E,aMtion Report.

Vincent S.- Noonan, Chief

~

Engineering Branch n

g -

Division. of Operating Reactors e.-

1 s 2.

~

m

Enclosure:

As stated 4^

y_b

~

Contact:

J. Zudans 3

2.-

aq jeggessner j&

,'a."w~n$n

. _ y

.(_;

icdiY?5tel1o

~~ '[

[d,, -=

.~

hY

' ^

i '-

C D.'Eisenhut'

~~'~I

/

7

.~W

.R. Reid l!)

~ :

ES:DOlRlfu

_ u ' - \\$ k,&

~

@: DORM ___

m

~

9nnm ames*.

kfsb vneonan W~ M

. ~,... ~

1.

s

.. m m

.~

,/78 oars >

o./. 7A

'0/

f i.J :.

'E' h

.h.

U* S. 40V4 NEM Mye O{m s',1979 = 8864a4[,-[p;,-( g t ggA?, [;k_p s s

D

(.

v

~-

~

4-CRYSTAL RIVER UNIT 3 FLORIDA POWER CORPORATION DECAY HEAT REMOVAL PUMP SHAFT FAILURE EVALUATION REPORT ENGINEERING BRANCH On May 8,1978 a report of a recent shaft failure in the decay heat removal pump OHP-1A was received.

Due to the severity of the failure, the second occun enceof this type on this pump, the staff felt that the reliability and operability pf this type of pump was in question.

The staff requested that steps, as outlined in our letters of June 7 and August 3,1978, be taken to establish the required reliability and operability of these pumps.

Outlined below are the actions and results of the efforts of Florida Power Corporation regarding pumps DHP-1A and 1B.

Pump DHP-1A The second DHP-1A pump shaft failure was primarily attributed t.o casing-to-impell er rubbing.

Subsequently pump 1 A was rebuilt with a new shaft and the impeller hub was shaved back to assure proper clearance between the casing and the impeller. A special test was perfomed on this pump te establish that the rubbing which existed is no longer present.

Frequency data was obtained utilizing cccelerometers installed externally on the pump, connected piping and bearing housing.

The acquired data was receive: for frequency content and noise source identification and then compared to other data sets to identify comparable performance and trends in perfornance for the lA pump.

The data sets compared included two sets from pump 1 A and two sets from pump 18.

Raw data reduction consisted of plotting the overall noise levels, tabulating the magnitude versus frequency for selected accelerometer locationsand overlaying Power Spectral density Plots for direct spectra comparison.

Results from the comparative data taken indicate 'that the A pump is running as well as the B pump.

Also the overall noise leveh ca the A pump casing showed a decreasing trend.

This trend was shown from data in g's RMS cs the A pump accumulated run-in-time. The levels recorded after the run-in period indicate a very favorable pump operating condition as compared to industry standards.

Based on this data a reasonable con-clusion is that there is no non-lubricated metal-to-metal rub as ind,1cated by extremely low noise levels at the various frequencies, no high frequency (1-2 KHZ) spiking and low broad band noise levels.

In addition, post-test inspection showed no metal wear on an rotating members of this pump.

The lack of significant unidentified frequency peaks, the acceptable noise levels both broadband and at individual peaks and the visual inspection which showed no metal-to-metal contact in the pump assembly indicate that the A pump is running with sufficient clearances to preclude the type of failure experienced twice before.

1

-~

gv, g=

2 Further results of the hydraulic characteristic tests for pump A indicate that in all cases, the flow versus developed head were within the acceptable limits for this pump as specified by Section XI of the ASME Code.

These te::ts were run with pump flow at rates required during accident conditions.

Pumo DHP-1B At the conclusion of A&S pump confirmatory tests pump 1B was removed from line for disassembly and inspection.

Initial inspection indicated minor contact between the impeller and the casing and impeller and casing wear rings.

The contact marks were oc the front shroud of the impeller at the extreme outer diameter. The casing mark was opposite the discharge at the mounting pedestal foot. All the marks are dark in appearance compared to the machined areas on the pump casing. There is no evidence of metal removal from either the impeller or casing. The appearance of these contact areas indicate that marks are old, possibly made during manufacturer's shop testing. The licensee has comitted to contact the manufacturer to ascertain the manufacturing history for this pump.

The scratch marks on the wear rings were irregular in location and did not propogate around the impeller wear ring area.

Each mark had a start 7 '

"' point, which is typical of some debris beingtrapped in the annul-

..c inspection by liquid penetrant showed no indications.

The ii..sction covered the whole shaft and concentrated in the keyway area near where previous failures occurred.

Hub material was removed from the B pump impeller to furtterassure adequate clearances as on the A pump.

After evaluation of the data and of the shaft inspection results and the satisfactory performance of the B pump in conjmcticn with the A pump qualification tests aspreviouslydescribed, it is the staff's opinion that no significant degradation has occurred on this shaft which would cause failure similar to that of the A pump and that the pump is reliable.

Conclusion he methods and criteria used by FPC in the testing and repair of DHP-1A and 1B are acceptable to the staff.

In addition, the staff has determined that the testing results demonstrate the reliability of DHP -1 A and 18.

e e