ML20238E734
| ML20238E734 | |
| Person / Time | |
|---|---|
| Site: | Palo Verde  |
| Issue date: | 08/31/1984 |
| From: | ABB COMBUSTION ENGINEERING NUCLEAR FUEL (FORMERLY |
| To: | |
| Shared Package | |
| ML20238E420 | List: |
| References | |
| FOIA-87-261 CEN-285-(V)-P, NUDOCS 8709150167 | |
| Download: ML20238E734 (67) | |
Text
{{#Wiki_filter:~ ~
]Q)gGRA
~' ^ *
$U."if.,"75,'.Oi'0%fe.'
1984 An I3 p;q g 9g con No. % &
~'
ClEGIDIt vos,p ggy ggy (yy_p
. l
'~
%WNGRA)ED NONPROPRIEIAU
~
REPORT ON PALO VERDE NUCLEAR GENERATING STATION LOW PRESSURE SAFETY INJECTION PUMPS FAILURE TO START AUGUST 1984 91g7 870910 HOLMESS7-261 PDR 9"i ' POWER mi SYSTEMS TMBUSTION ENGINEERING. INC d
,, , . ; =, , 'p I 5., .
W f' '
'i
. g %
v ,
"['[I,i($'s 1 e 4i 3
LEGAL NOTICE , t THIS REPORT WAS PREPARED AS AN ACCOUNTOF WORK SPONSOAED BY COMBUSTION ENGINEERING, INC. NEITHER COMBUSTION ENGINEERING NOR ANY PERSON ACTING ON ITS BEHALF: A. MAKES ANY WARRANTY OR REPRESENTATION, EXPRESS OR IMPLIED INCLUDING THE WARRANTIES OF FITNESS FOR A PARTICULAR . PURPOSE OR MERCHANTABILITY, WITH REPECT TO THE ACCURACY, ! COMPLETENESS, OR USEFULNESS OF THE INFORMATION CONTAINED IN , THIS REPORT, OR THAT YHE USE OF ANY INFORMATION, APPARATUS, METHOD, OR PROCESS DISCLOSED IN TV,.lS REPORT MAY NOT INFRINGE PRIVATELY OWNED RIGHTS;OR B. ASSUMES ANY LIABILITIES!WITH RESPECT TO THE USE OF, OR FOR DAMAGES RESULTING FROM THE USE OF, ANY INFORMATION, APPA-
%ATUS, METHOD OR PROCESS DISCl,OSED IN THIS REPORT.
J ( l { l f f a i > I
,/ -
; j
\
w l + _ \
- l. :
,%L , . *
- t. -
,- g 4. .
~.
REPORT I ON PALO VERDE' NUCLEAR GENERATING' STATION LOW PRESSURE SAFETY INJECTION PURS FAILURE TO START Prepared by COMBUSTION ENGINEERING, INC. Windsor CT. I
P" i. I i . 3 % ' TABLE OF CONTENTS SECTION . TITLE 1.0 ' INTRODUCTION 2.0 SWe(ARY 3.0 INSPECTIONS AND EXAMINATIONS 4.0 TESTS 20 ANALYSIS 5.0 DESIGN MODIFICATIONS LIST OF FIGURES FIGURE NCE TITLE 11 WF PUNP CROSS SECTION 1-2 PUMP GENERAL ARRANGEMENT 4.2-1 TOROUE PLOT - NORMAL SLCCESSFUL START 4.2-2,3 TORQUE PLOT - NON-NORMAL SUCCESSFUL START 4.2-4,5,6 TORQUE PLOT - UNSUCCESf,FUL START
- 4. 5-1 DEFLECTION MEASUREMENTS 5.1-1 IMPELLER SERRATIONS LIST OF TABLES TABLE NO TITLE 1-1 FUNCTIONAL DESCRIPTION OF LOW PRESSURE SAFETY INJECTION (LPSI) PUW
- 4. 5-1 SKAFT DEFLECTION RESULTS - PALO VERDE
- 4. 5-2 SKAFT DEFLECTION RESULTS - BUFFALO APPENDICES A PU W DESCRIPTION B PALO VERDE LPSI PUMP RELIABILITY C RESPONSES TO NRC QUESTIONS 7/5/84 1
,e ,
1.0' s Inteeduction 1l,1 - , General Description The low-pressure. safety injection (LPSI) pumps are components in the safety injection system of a pressurized water nuclear power plant. They are designed to be on standby ready to operate at flow rates up to runout for 1 year. Periodic testing is performed to demonstrate systen integrity and equipment performance. The pumps are also used for approximately a.4-week period once a year to circulate reactor. coolant to remove ~ decay heat following plant shutdown. See Table 1-1 in Appendix A' for a functional . description of the pump. The pump _ is a vertical, single stage close coupled centrifugal pump with the motor installed on a supporting head which is in turn supported try the stuffing box extension and casing.. The casing is welded into the suction and discharge piping and supported on pedutals. The LPSI pumps are centrifugal, of the vertical bottom suction. tangential side discharge design. The igeller is mounted at the end of an extended motor shaft. These pumps incorporate a pull-out element feature whereby the pug internals can be easily removed.from the casing by lifting the motor. (See Figure 1-1 and 1-2 in Appendix A). 1.2 H1 story - On March 18,1983, LPS! pump B in Unit 1 of the Palo Verde Nuclear Generating Station (PVNGS) was operated for approximately three minutes with the suction valve closed due to operator error. The pump was shut down. An attegt to start 3 hours later was unsuccessful and the motor tripped on time overcurrent. The pump was cleaned and dressed and operated repeatedly until May 16, 1983, when it failed to start again following operation with the suction valve closed due to operator error. The motor was replaced with a Unit 3 LPSI motor on June 19,1983. The pump failed to start after a successful two hour run. -The pug was reworked and successfully started ten times. On July 6,1983, the pump failed to start. On July 23 the lower casing ring was replaced, the_ ige 11er refurbished to the original design dimensions and the clearance between the impeller and casing ring increased to maximum allowed by the technical manual. No other modifications were made. This pump did not fail to start in fifteen subsequent starts. On October 8,1983 LPSI A pug in PVNGS Unit 2 failed to start after 41 successful starts. There is no record of improper valve alignment in these 42 starts. As with the other failure to start the lower casing ring was circumferentially displaced approximately 3/4". Its failure characteristics and cogonent indications matched that of the failures in Unit I (LPSI 18). Although the failures of LPSI 18 were initially believed to be vaique, (i.e., failure to start due to startup with suction valve closed) it was at this point considered that the failures might be comon to all the Palo Verde LPSI peup , assemblies. ' a
i- 1.'3 3 Safety Concerns
~ The Low Pressure Safety Injection (LPSI) pumps are. designed to assist in satisfying the Reactor. Coolant. System (RCS) Heat Removal and Core Heat Removal safety functions by providing for post loss of Coolant Accident .(LOCA) core reflood and Long Term Cooling (LTC). In addition
~ the LPSI pumps are part of th safety grade equipment used to achieve 1 l
' cold shutdown following other events. The fully redundant two train '{
LPSI system is fully capable of satisfying the safety functions given i the failure of one LPSI pug to start. For a large break LOCA, the core reflood requirements established by the plant safety analysis are satisfied with one High pressure Safety-Injection (HPSI) and one LPS! pump delivering to the RCS. .In the unlikely event of both LPSI pugs failing to start, the time to accomplish core reflocd will be extended. When reflood is extended it j 3 is possible that acceptable peak clad tegeratures will be exceeded. A condition which could result in both LPSI pumps failing to start is- J unacceptable. For a small break LOCA the LPSI pumps are used'to provide LTC by entry into shutdown cooling. . This is the general function of achieving cold shutdown. This function can be satisfied given the failure of t,oth a LPSI pugs to start by using one of two redundant Containment Spray Pumps (CSP). 1.4 Tests and Analyses - The cause of the fail-to-start was theorized to be hard contact between the igeller and casing ring. Accordingly, three possible solutions were attempted. The first was to mitigate the ' effects of { the ige 11er wear ring contact by changing the casing ring material 4 and serrating the impeller. The second was to decrease the contact ' force by increasing the gap between the wear ring and the igeller. The third was to decrease the startup deflection of the impeller. The ' first two methods were unsuccessful in preventing the January 29, 1984 failed start. Ultimately, the third was successful.
~
i In order to understand the ige 11er deflect 1on phenomenon, a simultaneous test and analytical program was begun. The test program determined the igeller deflection in the original conditions and also with certain proposed modifications. This data was used as both a , benchmark and as input to the analytical effort. Using the test data plus basic electromagnetic theory an analytical model was developed. This model was capable of reproducing test data but could not predict the effect on impeller motion of structural modifications. l A number of different analyses were performed. These included efforts by Westinghouse Ingersoll-Rand, Combustion En consultants, Failure Analysis Associates (FAA)gineering, and MechanicalBechtel and Technolog, Inc. The models all included the impeller, motor shaft, f rotor and shaft bearings. They all assumed that the motor frame, and 1 motor support system were rigid and did not interact with the motor shaft. The validity of this assugtion has never been proven, although experimental data, i.e., accelerometers mounted on the frame at various elevations, indicate very little contribution to shaft deflection from the motor frame.
]
l
o . t
' Arizona Public Service initially engaged Failure Analysis Associates
'(FAA? 10 investigate the failure to start problem and they perfo rmed the first Dynamic Analysis of the LPSI and CS Pumps. In their model l
' they adjusted the bearing stiffness so as to match bending frequencies { as detemined by an instrumented inpact test (LPSI Pump only). The { LPSI,puq. result seemed to confirm that the failure to start was I excessive deflection of the impeller during startup. The CS pump i result, however indicated that those pugs were also suspect.
\
FAA reconnended .th'at an taproved model should be developed and used in the startup simulation. Thus, although the FAA analysis was useful in indicating the importance of shaft dia:neter to deflection it was not a suitable model for predicting the startup behavior of the CS motor and is not discussed further in this report. e e l 4
2.0 Susanary -
~
The subject of this report is the failure of a LPSI pump motor to j achieve operating speed after the starting circuit was energized. The problan encompassed three of the five LPSI motors that have been
. operated at Palo Verde Units 1 & 2 during plant pre-operational test-ing. The specific motors and the history of the failure is covered'in Section 1.2. The occurrence on more than one motor on more than one occasion required an' evaluation of the possibility that all LPSI pump motors at Palo Verde may be susceptible to this failure mode.
After investigation by Arizona Public Service (APS), the MSSS supplier (C-E),' and the equipment designers (Ingersoll-Rand and Westinghouse), the conclusion uns reached that the original LPSI pump motors are unacceptable for this intended service and would be replaced by new motors which are diplicates of the containment sprcy W. ors. Other C-E plants have a LPSI pump of the same design. but use motor: with a ; stiffer shaft and have experienced no operating problems. The CS ' motors are 800 HP vs. 500 HP for the original LPSI motors. The increased starting load is within the capacity of the emergency diesel gen trator. , The failure to start has been attributed to the design characteristics of the original LPSI motor. The motor shaft and its support structure. are such that, randomly during the starting transient the impeller end of the shaft deflects to the full extent of the available clearance and produces a significant contact friction force. This friction force results in the development.of a rotational mode which prevents assembly beyond approximately[251 rpm. acceleration Operation in thisof mode the rotating] for{6 seconds trips the overcurrent circuit breaker and causes angular displacement of the casing ring. The radial magnetic force which produces the bending of the shaft originates between the motor stator and rotor. The large deflection of the motor shaft cospared to other motors has been demonstrated by
. test. The problem has been resolved by utilizing motors with shaft j
deflections within the available clearance. ' Containment Spray pump (CS) motors are not susceptible to the failure to start. This has been confirmed by evaluating the symptoms of the failure in the LPSI motor and by tests that' demonstrated that these synytoms are not present in the CS motor. Principally, impeller radial deflection is significantly less than the available clearance during the starting transient, and marking on the casing wear ring and impeller after 100 demonstration starts was normal for this pump design. In addition, a statistical analysis of the motor operation within the safeguards system was performed. This analysis justified that a series of 100 consecutive starts was sufficient to. demonstrate L .the necessary reliability of the motor, see Appendix 8. Tests con-ducted by Westinghouse have independently verified that the I containment spray motor shaft deflection is much less than the ! available clearance and is also much less than the original LPSI motor phaft deflection. L- _ - __ _ _ _ _ - _ _ _ -.
( *- The containment spray pump motor and LPSI pump motor designs have been i examined to determine why one motor is more reliable than the other. The design feature of the CS motor that is most significant is the larger diameter shaft between the motor bearings. Other factors such as different frame size, rotor diameter and length, and rotor to stator air gap may contribute to the shaft deflection but are not separable by analysis. The test measurements of shaft deflection are the result of al.1 contributions and demonstrate that the set effect of l the different design features is minimal deflection of the containment l spray motor shaft at the pump impeller. The bearing rigidity of the containment spray and LPSI pump motors are the same since they use the i same bearings. Shaft stress concentrations such as undercuts at the bearings had little effect on the problem since the no start condition appeared in motors both with small undercuts and with normal undercuts. )' J A motor of equivalent horse power to the Palo Verde LPSI motor was evaluated as an alternative substitute for the PVNGS LPSI motor. This motor, procured from a cancelled System 80 plant, has a larger shaft diameter. Shaft deflection tests showed that the impeller radial motion during the starting transient is similar in magnitude to the CS motor deflection. Tests on this motor design provided additional confidence that the failure to start phenomenon is limited to the ; particular design of the Palo Verde LPSI motors. j
)
i I l
. 1
.. t -
.3.0 Inspections 8 Examinations 1 j
Visual examination of the original pump parts in the failed-to-start f pugs LPSI,18 and LPSI 2A showed galling of the lower casing ring, { which was made of K-Monel. There was also evidence of metal pickup -{
^ from the casing ring on the igeller. Marks on the casing ring {
caused by impeller contact were also observed. They'were at regularly spaced intervalt and provide evidence that a " whirl mode" (nutation) occurred during the failures to start. See Section 4.6.4 for a discussion of the whirl mode. , Subsequently, the upper and lower casing ring material was changed to Nitronic 60 (Section 5.1.1). In addition, the ring to impeller clearance was increased by several mils (Section 5.1.2) and the igeller running surface was serrated (Section 5.1.3). Six pump runs-were made with this arrangement using LPSI Pump 18 and the motor failed to start on the 6th attempt. However, it did start on each of six more attempts without any disassembly or modification. The pump was then disassentried and inspected. Visual examination of the disassembled 1 arts revealed light rubbing had occurred around the entire circumference o the lower wear ring at the location of the lowest impeller land, ere were faint but r ular marks on the ring again believed to be av dence of a whirl mode There was also some metal transfer from the casing ring to the i 11er. Also, as in other failures to start the casing ring had been rotated approximately one quarter to three quarters of an inch. The anti-galling properties of the Nitronic 60 ring material were confirmed after inspection by the pump vendor's metallurgist from its appearance as compared to the K-Monel ring. The next series of tests involved a Palo Verde containment spray pump (800 HP) motor installed on the LPSI 18 pump. After six water runs the punp was disassembled. Light contact had ocurred in the lowest . region of the lower casing ring on approximately three-quarters of the circumference 0=t the Previously observed re 1arly spaced wear marks indicative of the whirl mode were not present The quadrant in the region of the pump discharge appeared untouch . Its appearance was i as expected when pug running contact does occur. The pump was reassembled and 100 more starts and runs of 20 minutes each were made ("100 start test"). The pump was again disassembled. Light contact between the impeller and lower casing ring had occurred at all four impeller serration lands again around the same three-quarters of the casing ring circumference. The contact was light, but more prominent in the quadrant opposite the discharge. There was a very small amount of metal transfer from casing ring to impeller except on the uppermost serration which had none. Again, the appearance of the impeller and casing ring was deemed to be normal by all parties to the test including the pump vendor's metallurgist. The metallurgist i provided photographs of his lab test samples for comparisons. It was J acknowledged that light igeller/ casing ring contact with this close coupled overhung pump design is expected and acceptable. _ _ - _ - - - _ _ _ _ _ . - - - - - --- )
. r . .-
4.0 Tests' and Analysis 4.1 . Torque Measurement Tests 4.1.1 TEST 5ET UP, PROCEDURE, CALIBRATIONS, ETC., BY APS. Torque plots of a number of successful starts and one non-successful start are available. The plots record the dynamic response of the punp motor combination to the motor torque. The plots do not record the total torque developed by the motor, but only that portion transmitted to and absorbed by the pump. 4.1.1.1 Successful Starts
~
The torque plots for successful starts show two different patterns, normal and non nnemal. The normal and non-normal starts are similar except for the occurrence of negative spikes in the non-normal starts. The negative spikes were not expected to occur and were classified as non-normal since they indicate that the impeller is driving the rotor. - 4.1.1.1.1 Successful Starts Normal _ A normal successful start is shown in Figure 4.2-1. The plot shows two patterns. Up to 250 ms the plot is a damped sinusoidal motion with frequency 55 Hz, and amplitude of 580 ft-lb about a mean of 330 ft-1b. . After 250 as the frequency changes to 100 Hz, the amplitude drops to less than 50 ft-1b and the mean torque increases gradually to 620 ft-1b at 670 ms at which time the data. ends. The first pattern would indicate the impeller is torsionally vibrating while the rotor is held tight by the magnetic forces. The second pattern represents the impeller and the rotor both torsionally vibrating. The shift from a one mass to a two mass mode would account for the sharp drop in amplitude. During the first 300 ms the pump only absorbs torque due to rotation inertia. After 300 ms the mean torque starts to increase as' additional torque is absorbed to start pumping the water. 4.1.1.1.2 Successful Start - Non-Normal . Sample non-normal successful starts are shown in Figures 4.2-2 and
-4.2-3. These plots show similar patterns to the normal. starts with the exception of negative spikes. These spikes occur in a random fashion, with no pattern as to number per test or time of occurrence, other than always occurring during the first pattern (as outlined above). These negative spikes indicate that the inpeller is driving the rotor. Figure 4.2-2 shows also that from 70 ms to.90 as no torque is being absorbed by the inpeller. The positive spik'es preceding the negative spikes seem to indicate that something inside je motor is causing it to momentarily slow down. _
l 4.2.1.1.3 Non S'uccessful Start ) Figures 4.2-4, 4.2-5 and 4.2-6 show a non-successful start. These plots show several differences between the successful and non-successful starts.
. The second positive spike reaches 2090 ft-lb, while for the 4 successful starts the maximum positive spikes are no more than 1650 ft-lbs. ..
~
From 50 to 70 ms the mean torque jumps to 1320 ft-lb, and the frequency rises to 150 Hz. Both before and after this period the plot resembles the first pattern of a normal start however, there is no transition to the second pattern. The amplitude damps down-somewhat but such less than for a normal start and there is no change in the frequency. The mean value rapidly increased to and stabilized at full motor torque, + 250 ft-lb. I
}
y y
- a
- 1.
, y .
_=y -
. ._.....p.......,...
-j . . . __T s
.[....,...
. . .!._ . ..... _____.p_ . . . _ . . . . . , . _ . . . . .,:. ._ .
..g......
..._ . . . . . . g *.. . _ _.. -. .. _ . ._ . 9,._.
._.........e........
p _ _ .9_ _ __ . . .. .s.g . o ,
. .. JE. i.
.. . .. .,.(.. . . . . _ . . .. .
)........ . . . .
_ _ . , _. ...._s.__ .
. . . .l.....,_... .. .. ..... .
.. t . . . s . . ..... .....g. .... g .
- s. ._ e.. 1
. 9_
, -. s .! . .. .. . . . . . .CD.. ., :
]
I . . . . . . .
._.. ....f . . . .. _._ _. .
. . . . s... . . . _ .......J. .I . . . . . . . . . . . . 4. _ . .
.. _. . . . . . ~ . g ....+ . . . _
. . - .: -. ! _ - .. l . . - . . _ . - . . .
.. _.. J
_...:. . - . _ _ ..j.. 3 . . . .j . . l . j .. . .. . . . _ . .
. .: . !.....g.__ . _ ...
..- _ 4 ... .. , .
.. 1 ._ . . . . .I _
7.p._... . .........:__ ..__ . .__ . . g
.... . . . . . . -_3,._._.._ _ . _ e. . . . . . 4 . _ _ .
. . _ . . . .. : . . . 4 G _. .
._._._..._ ..... . . .. _ . t./'). . . .
. !. . . _. .[. _ . . .. . t... . . .
.{.
. . . . l . . . .. .. . . . l,..___ ... . . ____.,.___.
. .{
. _ . . L. .. _
. . r... ._ l . ..... __
...)_..._ .
__...J.. a.. . . . _ . .. . . I,. _ _- . ..,. . ...... _ - . . . . _ a _ . .j.. . l ._ . . . . . . . . - . . . . . . _ . _ . . . .t.
.... .. ..;,.. , . .. q...
.!._ _. . - . . . _ . - . . . - . . . . .g .
.. ..s.
...4.. ,
, . _. . ... . . ..g... . _..g.. . . .. .. . . _ :. . . . . .ar . .. .
. , t.J . , .
.q .............g.......
.3..._.-
_.7....
- _.....i f.f l':) . .; _
o . . . ,. . . . . i. .. . _ . .
.J,. ..... ...i,.
_ . 2,H "
,g . .. _
. . _ . . . . . ..g.
. .._...l....'.
__ g g;
,s .. .__ . G.r
.. ._..:.... . . 1. .. . ...: ... . .6... .. _ ; ..
. . . .. ._I. .. ..._ . . p._ . . t .. . . . .. . . . %... . CD n }
..p i g, p.
__ ;. _ a.. _ , _. 4._... . .. g . . .
- i. .. i ,
. e.
. .. . . . . , l
...l N_.
! 1. . . -
. T., ... .
i. l t......
- _u _. . .. _; ._ _ _ .
m.
- s. . _ .
4......
,. ' . . . ..g l ' 3 m J,.. . . . . . .
1 r$"_"* _ ___.._.,__ . m m ...__.1._. . . . . . . . i.e po- _ , . . . . . . r.. . _. . . _ . ....._, W \
. _ . . _ . __ 4
.. j
.s
. _ i.. . . . . . . . . . . . ..._ . _ . . . _ _.
p
.m ..
H. . ._ _ . . a ;
- g. . . - . . . . . *'
m._,e l l '.n_ j_ g. . .
-e -- g
.g . - - g _ 4~ ~_~ 7' " -" ,"~_j 4 ,,, ,.ij
' ' h7g- --
d - t"... ". i.-. . . . . .h._ __J;___S.IlF.Lds3 0 td(0.L '..s . . . .. ,. 2-. .. . ... .s a .. ; . .- ...a.t .. . . . . . . .. - - . .. Figure 4.2-1 9 s' I
t.
. 1 *-
4 _..1,..._..,......_.....
....t...._..g.......l_.._...... . i
.. . , ,... . , . . . . _. .. ..... .. . . . . .., ... .p.
. . . . g. .
... . . .. _ .G
. . g. . . . . . . . [... .. .
-- . . r .
n.
- .. . . . . . . . . . .g.... .. a . _
. . .. . i.. ... .-. _ . . . # . .. . . . . . :
. ... . . . . . ... . . s. .
... e i
t... r .
=__ ..., . . . . . . . . . . .
.p
- s. .. .
- t. . . . . . . . . . . .
. . .g
.....g,
._...._.e.......
g . .. .. . ; ~ .
.... h - _ . . . . . . . . . ,e. ...g. :p.
_. . . . . _ . _ . .. . : ...._.. . .L . . _
.1_.. _
.. g
.. ..,p
,. . . _ . ._ _ . . ;. . .. "} ... . . . _ . . j. .
.q . _ . : . . . . .] _ _ . . ; __ _ _ . . . _ . ,
r
....._t.....- . .
t_ p. . [ _ _ _. ; .._ (~ . . . _ . . . . _..a s .
- _ p.. . . . . . . . .
i .
. . . . . .,a. _ . 4 .. ..
_. .. .: . . .. :.. a. . .. .. .i. 1 .. . . . ._. . __ 4._ .. _ . L . .. .. e .- ..! 4 , .... _e . _t....,_-_-...
.A _..6.
. s. 4... . :r . _
_ . ,__7_._ _ _. p .. _.-.--- _
.__..a. . . _ . . -
o '_
.__.. i . __. ;
i i ng 4
-. . .=. ._ .
_ _ . .= _ . .. . g.. , ( hs. ..._..4._.- . - __ 1 _ _. .. 4
..y ,
_ _ - .- e..
.w a. .---- . -
r..
. i
- s. r
.tn.x. w
--.r, . _ -
L. . . .
;.&. g . ,
.r .
,, p.a n,_ .
- u. . .._ _m_ _q __
e_. y y., . . . i m .
.e
., ._s x
. . . . . - .- a .
.. Cu _-
s. _ __ i
.e
- v. __.
e-(.m.. .
= ___ ' - .
.n_ _
*= . _ _ .a* _s ..t ' "
rv. _. ._. i m_ . _._ __ l g _
..e-$,
tp
, , , -= - r- y- .
_o a.
, T' u
=
g . __ .. 9 . __ .. . -- .s 4 . _ 4 - -_- _ = _ .. _ . L ._. - ' a
- i. . .. .! . .. . 4 - -
- .----) .-
.,4 1
3
.... . . . . . . . _ _ . I.. -.
L. L._._ __ - . .
.. i -
3
. "M. 9 .. . ., .C. D) U
___.. ,7. . _ _ . V _. ..
..CM. ,_t- L 4' .
. . . g ... . . . , . . . . . g. . -. .. .s._. g. g , .p, .m y o .y ._, _ , ...t 3
, _3
. .:.. _.-_ s. :.. - . - - .
s_ qt f-Figure 4.2 j -
, 10 . .
3 '.*
p' .. . . _ .w.. e ...._~m. . . ~ _ . . . . - _ .. . .
' e, .
g s .- \ ( . g l L f im.. M.
.. p.v .. .ss_._..' --
- a. _'
. y 3 . .. .. ..&. . .a
- 3. . . .
..... .s
. . ..... .. . .. . . .. 3
. 3 ,. .$.....
.. . . .,.... p.g
._ . . . .. . . . . . . .. _ .. . _ . . . . . . ,. . c.,
1
..e. . . . . _ . .
. . . .3 .....
.. >.- l..
,. . .. s. . . .. ._
_,m- .
*je
.. , .. . 3
.." .4 . ....i.. ,
.. .. p. ..
1 < . -s e, g . . s . . . . . . . . . . . . . . _ .
. U3.
s _. . ...,. s. _ . . , . . . . o.. . . .
.. ......y.. ..
j ....;.....
.. . . g . . . . ge 6 . . .. . . ;.
. . . ., c. 4 .. . . . . .
l-.
. . _ . . .i. . . ._ _ m;
} ._... ___.,...
,.....g...
. . _g
. O. .
I,....w. ""E y.,. .. ... . . 4 . p ... 6 . .. r . . . . . . . . .
. .1, ._ g._ . .q
_. l. p._ ...
........._._.4...__...__
._g.....
. . .. . f n.
. _ .g
.g.
. . . .. _ ... . . ...........q,..... ,. . . . i.
i 1
* , .. .......g.. .. _< U . ..
s .. . . . . . . . . - . , >_... _ 4, . . . ,
.:rL . 3 .
s .. .... . . . p . . .
.. p _ ... . _ .. _ . . <q_
. .j i
1_
=
_. p. _ Ty ... ._...a.... ._ . _..-j ,... .... .
.l .._4... 4.g.g,
. a: .- _...d-_---p....E..._ >
- r. -_-- b. . . _ - ...-.._t.__
,n
.. g...._. . _
- y. ___.- .- g3 8
f' p.
.9.. . . l .. .... .. ..
._. . _. ' ._ . . a.eJ
=
.4 . . .p.,;.
. .. . . .. . e. . ..
. .. . (_. . .; . .
g .. . e . . . . , . . . . . ..
.i .. ...g.- .
.. ... . . . . . p... . . .
. .A . . . _
. .rw_. __...t._.'...
I.... w.
.y
.b...
. $.= , . . ._ . .
9 . -* r- ,
.-gpp . . . . e.._., ._ . _. -
g - . - . - 3 ~. .
. .. .i . . . . . . . . . . .. - . ..
, ...... e. ..& .N.L
__.2
... . . .'.4 . .
g: ___ .
....__t._..... . . . _. .___. . .
.g,,-
. a_ . . . . .; . , , , , .
_ _ .L._ ~. _ _, , . . _
.. .. . . . _s8 . . . .. ..... . ....I... .. ~.. *
~_-
- .. .. . _ 1
. _i 1, . -
. i. . .. . I ... .: .. % = . .. _
.. a..
. _ 4 . . . - p . . .. . . ..%. -- _ _ _ .
. . , . . .l . . . . . . . . . . . . . . - - . . . . . c . .. 1 OpW2 _. . , c m..... .. .. - _ . . . . . . .
cmi- -
.u... .' .. - ... p _. .. . - . 3. . . . .. gyi 4 .:ingn t...r.- _- . - _ . . - '
a s . n I
. . s.. .s . . . . .i .. . . ;. . . u . . . . i Figure 4.2-8 .
11- *. e
, . . . n-i . c.- g .. . . . . , . . . . p, . .. . .. .. ...,... .. . _ . . m_
. . . . . . . . . - . -= --
. . . . . . . . ~ . . .. .
- g. ... .. ..,... ..
l . ... ... .. .. . . . . _. .. .. . i...._.. . . . .
. . . . e= g T~.
~ ~ ~~~
~
7~~ _- L.~~ 55 ~_ . ._..._ _ _ . .-
* .i P.= . . . .. .
-~ ~
.~~ .:. : :
' ~~
- ~~~ -" . " . ~ '.gC.
y
.. _ _ _ 1
. .a- . . . . . . .._ .
---._ . _ ____.L.. . . . . . . .
.- l~ ~ _ . _ _ ..
t .. T.~. ~ 7 ".-"-~
~ - .
- ~~
i~ ~. - E i [ _
-.-{--.... . .. .z G)
I
... e. -, _ a.
t. g
~
.}. _ _
-Q
( _ ' ' _ -*.' y
.r a- . . _ ....._-....I.
me.- . W --
--- -- W " .
.. -cu...
- i. _ . . .- . _.
p .__, . , , , . , -
-f" " * * " - * * "
, . b - ._ ,
1 .
~
b
. .. . . _ .me
~ g
$,n. e .
. . _ . _ . , . . _ . , ~
h..b A_. _ ~ t . . . . , _ . . . x .? w . . . . . . , . . . .- . . S gr,,. . .. . . . . _ -
. .. .._ ; _.. .. .. y ;..... . . . . .
g ._._.3......~._... _ __.. __. . 4 ' "._- m .- --
. - .. . r - - _- - - - . . . . . _ a-_
_. r - _q .
. __ #_ . - .- .e ,-- F ,
- - - _..t.-.-
.__ _____....._; 4._.-4.._. a-- -
4...-. .
.. , . - _ . . _ . - e-
,. .. . _ . . , . . . ~ . . .
_-- . _ . . -. -e
...m.......__..s.... . . . . -
- _ _ _ . . 4 . o
- a t Oct . ..... .. . . o __ .... dW .. ~
- i. ... .. ... .-. _. .... _ ..__
N. pori .~~..~ .L.T. ._ Oz.rci
.f- -
. gn-m_30y n g ~.I . . - . . . . . .- ,,. . . .
..._._......_-.........._.....--.*.._.--.s.*.s---_.
y . .
' .w . .
Figure 4.2 4 . m , ,. - I 1
. g j
\
. j 1 ,. . .
i r_ .
. - _ . _t...___
%. . t.
B _. -. , + ..sn. e
. m.__.:
_=
...= 4 -.
=. .
_e.4 -
.__. . -. . - . . + . _
q .. . . . . .. %__:.. . _ _ _ __.
-W ,
_-- . _ _ _ _ _ i .*- mmum
. g _j
~
--+ -
. . _ =
__ __:.1
- a. ' .- .
-_.= _
- - _ _ . ,_ .:. .___ . _w _ i
_.__ i,. _w__a
._ . . 4
]
_ _ 4
.__._._p_.._
_.__4._.___ __._. . ____a =. - a i.... .
- - - . . .s . -_ _ _ _. = - _ . __ 4
__. s . .. . T ___ ___ .__m r-- r>_-.__
. ..... g . . . . -
, _ _ 9 ,
r __
,m. .
.3 n __ . _ . . _ _
.__.e__... .._
. u .-
=
- j. .__
. _ y.4 _.. .. .. . ;
a, ..__ .._ .
.u... . _ _ . . ._. i ..
... _ .L. ...... ... . . . . . _. . . . .
.r __
s ._. _ _ _ _ _ _ . . . _ . _
. __. 4 . . . . . _ _ . _ ,
in . _ _. _ __
,s ..
-r ..
. h. - -! - . _ - _ - - * .. _..*
. . _ _ .s..
t
- s. .
1 , _. . . . . . . . . . . . . . T .-,.;. . _ . -
. 4. . .. . .
.g. - . _ _.._.: _ . . .
u
- 1.. .1. .. s. .. . . . . . . . _ _ _ . . . t r . .. . ,
.. o .. o p tr ..
'r ~ C99 o W._.
c c.i . er.E.i . l 9
. ; . g ... _ .. , sn -
r.a - ! - -- . - . 4 .._ -. - . . . __ 8 a . .. .
- Figure 4.2 5 ,
l 13 L . t
P s 3 . 7
.....g.. .
.. .. a..__- _ _.
; 4
,, . . _ . .. . __. _. _ -p_ _ _ _ . .. - _ - _;g ,4 r .
.f, , ._
..m . .
. . I..
.. . w ,k
. f*t &
.6 . . .y... .r p .. =
.g . - )
. . _ . , .. : .p.: .: .r -
"'~ ~%g
;, ,4 p
- ~ ~
.. i _.,.. =w -
.J u.- _ . .., l n r ...,
;=.
g
,.3 . - .-._
. gL_._']
_.y3
, . . .- . A.(w ,s e
_=. , i _ _. ..=. .. .
=- I a,,'llgj _ - * -
, -. . .. 1
., -e
.g . .
w
~--^'^'M
_gg
. S1L.
g .p ,- _. O_M"sE. =
}_ gia n .ya
_. . ~ @ g . . .j
. . . e.
. . ;. _~ -- . , .
.p.
.e .:'r*
. . . - n,
. t,
._ )
.- . .. -- _ _ . . _. _ _ . _ p .._.y
=
l..- _ _ . . . _.g ;
...g.
_. _glll-lpur, e.
.g ,.
.g [g.;
s ,
.{.- . _
F.4- c _.- . , _..
=
_'._'s y ecj ij -
. ..aa . . _ . .. -
lh l L~ _ . . = . _ ._ _j
.. y _
o . ._ _ .
- g. . . . -. .__- _ ..
.Ot =
. . .l
- y. , , ~ ~ .g.
c , r. - .. e,85,
,- ._ . g .-
g..... o _ _. . m_ __y. x
. ss5 % a s . _.
. a,_.;
g_
- g. .
lk h q
'W J '#
4 a J . . ._., .. .
-gi ty) d II'm* . -- * ....a
. .. _.s -
.. - ,_ . . . =.=
ly g 4 .,,,,,. g&-r_
- Q . = _*
. a . . . _ . _ . -
g _ _j F. _G k] W,. .,s - 7 "' . __ 2 CF4h= .. _ ,c. _.. m - m, , y=- " - _ _ n_- = -
<y _ m _-- - . _ . . . . . . _ _ . _ . _..
_ . - _ =
- =m_ .
w~ =-
=
_ marn-- -
= - -
e__ -
.___.4 w !!gsaigna 1 M_ _
o_.
-, m- ,-;
..__..._a._.
5..% - ._.-.C..E.S I o ..,,. ..t.. o C_Crf.f- = . . . _ . ,. .F.F o opp _
"iftti 4.,9fmh B . 01 _. r- '
s
.. ._ ,i .
r -
. 4 Figure 4.2-6 .
-14
- l i
4.2 Dynamic Analyses By Westinghouse (W ) and C-E _ l t The C-E model was based on data supplied by FAA, Ingersoll-Rand and Westinghouse. The bearings were modelled as gagped elastic supports. The forcing function, supplied by W , F = F SIN wt was applied at i five points on rotor. The negative magnetic spring, F = -169,000 j lb/in was applied at the same nodes as the forcing function. The ! model used a consistent mass matrix with_ mass added to represent the impeller and the' rotor. The model was used to calculate the natural frequencies, both lateral and . torsional, and displacement time histories due to start up forces. The effect of the whirl was investigated by applying a displacement time history, D = .016 SIN (12E x 2to)t to the impeller while simultaneously applying the start up force. - The mass matrix and stiffness mass were generated by the STRUDL DYMAL conputer code. The DAGS computer code, a C-E proprietary code certified with the NRC, was then used to add the non-linear elements to the model and calculate the (ynamic time history response. C-E hap developed a forcing function of the fonn[F = Fg sin wt + Fn sin wGwhich generates displacement time histories which closely match the test data for the Palo Verde LPSI pump. Although the C-E developed forcing function closely matches the test data for the Palo Verde LPSI pump, this function does not accurately predict the displacements for other motors (i.e., the containment spraymotor). 4.3 Analysis by Mechanical Technology Incorporated (MTI) - C-E engaged MT! as a consultant to assist in the analysis of the LPSI failure to start problem The MT! analysis used, as a basis, the model developed by FAA. ey did include a magnetic field with a ~
'~ strength of -350,000 lb/ n, ediich is more than twice the field strength supplied by W. The effect of this field was to reduce the static stiffness of the rotor shaft and bearing from 413,793 lb/in as
. calculated by FAA to 74,906 lb/in. MTI assumed an unbalanced constant side load of 2,000 lbs., resulting in a static deflection of .0198 in. at the impeller and .0267 in, at the rotor midpoint. Using the same -350,000 lb/in. in the C5 model resulted in a drop from FAA's 576,923 lb/in. stiffness to 263,158 lb/in. for MTI. Using the 2000 lb. side load resulted in deflections of .0035 in. at the impeller and
.0076 in, at rotor midpoint. This led MTI to conclude that use of the CS motors was an acceptable solution. The FAA analysis, using a field l
strength of zero, reached the opposite conclusion. It can definitely be stated that the field strength is not zero, but its value during the first .05 sec. of start up is not known nor readily defined. e
, l 1
IThe* MTI analysis, although a static analysis, also demonstrated that manipulation of the magnetic forcing function could ptoduce agreement between analytical and experimental results. - 4.4 Anal,ysis by Ingersoll-Rand (I-R) Ingersoll-Rand employed two separate analytical approaches to evaluate rotor critical speeds and deflections, Initially, the lateral critical speed was evaluated using the I-R program CRTSPDI. Shaft mass was dstributed along its length, based upon diameters. The I-R models did not include any stiffening effect from the magnetic field in the ptor air gjp. Bearing stiffnesses were varied in the range of
.5 x 10 to 1 x 10 lbs./in. With a weight of 142 pounds at the igeller, the first resonance of the shaft was 49.0 Hz for the 2-bearing system. A supplemental analysis was made considering a 3-bearing system, with the lower impeller wear ring surface acting as a bearing. For this case, the first critical increased to 74.2 Hz.
To incorporate bearing runners (sleeves) into the model, and to evaluate externally applied forces in the rotor core area and at the was employed. igeller, the Bearings werefinite element assumed conputer to have a stiffness program, of 5 x 10ANSYg,lbs/in. Four (4) model configurations were studied. The resonant frequencies were as follows: CONFIG. DESCRIPTION {c,1 A Same as described for CRTSPD1 45.9 Hz. B Bearing runners modeled 47.2 Hz.
~
C Impeller Mass removed (i.e., bare 88.9 Hz. shaft extension)
, D Same as B, but with seal sleeve and -47.9 Hz.
I ige 11er stiffness effects included The ANSYS Models were also used to determine the magnitude of magnetic force which had to be applied at the rotor core to cause the - deflections observed at the pump end of the shaft on field tests. For configuration A, an 8200 pound static force was necessary to get a lower impeller ring radial deflection of .026 inches, as measured on initial air run tests. For configuration B, the magnetic force required was found to be 10200 pounds. With configuration C, a 200 pound pull applied at the igeller mounting position (with no magnetic pull applied) yielded a .0044 inch deflection at the shaft end, as cowared to a .0045 inch deflection measured in the field. Several additional analytical studies were conducted to evaluate the effects of proposed changes to the motors. The scope and results of these studies are summarized below.
~
L - - - _ - - - - - - - -- - - - - - - - - - - - - - - - - - - - - - - - - - -
,. I i
) -
!1. 'The influence of shaft stiffness was checked by using a 6 inch '
shaft ' diameter in the rotor core area. Deflection at the i impeller ring location, under magnetic loading a~pplied at the rotor core, would be reduced by a factor of about 30% from the l deflection calculated for the original 5-3/8 inch shaft of the i i Valo Verde LPSI motor. , l
- 2. Shaft critical speeds (N ) were compared usiig the I-R program _
CRTSPDI. Stiffnesses foh the bearings were made equal to those 6 used in the' W -R&D analytigal model,1 x 10 lbs/in for the thrust. bearGgs and 4 x 10 lbs/in for. the radial bearing. The first and second critical speeds were as follows: 1
~
' SHAFT DIA. g $$2 5 3/8. inc. 49.1 Hz. 73.0 Hz.
6 i n. 49.9 Hz. 65.6 Hz.
- 3. Application of magnetic forces at the motor core, as a dynamic load with 0.27, daging, yielded deflections at the igeller which were twice those of the static load case.
~
4.5 Summary of Tests N 4.5.1 Tests at the Palo Verde Site 3o 4.5.1.1 Tests on the LPSI Pug Motor I
" The tests conducted at the Palo Verde site on the LPSI motor consisted of the following: shaft deflection in air without impeller, shaft deflection in air with impeller, and shaft deflection in water with impeller. -
The cause of the failure was believed to be the impeller striking the lower casing ring with sufficient force to prevent the pump from reaching operating speed which caused the motor to trip on time overcurrent. It was desirable therefore to evaluate the shaft deflection for the starting condition. Figure 4.5-1 shows the location of the deflections measured in the shaft deflection tests.
- The shaft deflections at the upper hearing, lower bearing and just -
above the stuffing box were measured on both the air tests and water test. In addition, the deflection of the shaft at the impeller was measured in the air tests. This deflection could not be measured during the water tests because the impeller could not be instrumented to measure deflection when the pum was pumping water through the system.
. j i . DEFLECTION MEASUREMENTS
' ~
7 Upper Shaft End
///// }
'f t
- I. wo . C)^
- 1 .
( I
) i h a
----ll
%j i
h h)9 0'- Y g / /, I Mid Shaft s Q'
//3 U .I,l
~
~ - :p 'K ll tV - m" -
Lower Shaft Erid j
'8 Impeller
/ tl-
. _ ~d i _
t Y l Figure 4.5 1
In addition, some pump and motor modifications were made during the testing in an effort to eliminate the cause of a failure to start. Modifications 1, 2 -3 and 4 below will be retained in the final resolution and are discussed.nore fully in Section 5.0. De following modificati'ons were ande during the tests: 1.- Use of Mitronic 60 Casing Ring De casing' rings were changed from K-Monel to Nitronic 60. The K. mnel ring had a fairly high tendency to g:11 when-it rubbed against the CA6tm mater 131 of the impeller. The Nitronic 60 has a lower galling tendency as well as a lower coefficient of friction than the K-Monel.
- 2. Increase Ispellar to Casing Ring Clearance ne clearar,e was increased to reduce the chance of contact i between the ispeller and casing ring.
l l 3. Serrate the Impeller j The impeller was serrated to reduce the impact on pump head of increasing clearances and to reduce the area of contact with the , casing ring if contact should occur. l i
- 4. Motor Doweling In an effort to ensure closer centralization of the impeller within the casing ring, the tolerance on the alignment of the motor to support head was reduced from 5 a.11 to 2 mil and the {'
l interface doweled to maintain this tolerance during reassembles.
- 5. htor Lower Bearing Preloaded In an effort to reduce shaft deflections the lower motor bearing design was changed to include a slight preload. This redsced bearing internal clearances which resulted in slightly lower radial shaft deflections. -
4.5.1.2. Tests on the Containment spray Motor The same shaft deflection tests that were performed on the LPSI motor were also performed on the Containment Spray (CS) motor. A *100 start" test was run with the CS motor mounted on the LPSI pump after the shaft deflection tests. This test consisted of 100 consecutive starts with 20 minutes of running between each start. The 100 start test was run to demonstrate pump reliability. The 100 start criterion was the result of an evaluation of pump and system reliability by statistical methods. Appendix B provides justification for the use of the 100 start criterion. . l Modification 5 in 4.5.1.1 above was not necessary for the CS motor because of its limited contribution. 4.5.1.3 Results of Tests at the Palo Verde Site -- Table 4.5-1 shows the results of the testing done at the site. Thes mil diametral or01.5 to 15.5Jmil radial. Coworiginal r,ing t clearances specified i in the
~
arison with data shows that the startup deflections of the LPSI Pump motor shaft in air with impeller would exceed ring clearances. As previously - stated it was not possible to obtain impeller deflections with the pug actually puging water though the system. An attempt was made to ' i estimate the deflection using a ratio from the air tests of the L deflection of the shaft above the stuffing box to the deflection of the impeller. This ratio could then be applied to the deflection of ,
. the shaft above the stuffing box from the water tests. When this was done however, the deflection calculated during a failed start indicated no contact with the casing ring. Since this was known to be -
false, this approach was abandoned. It is expected however, that the " g--- absolute value of the igeller deflection in water will be less than the absolute value measured in air, since the water will provide more resistance to igeller movement than air. L{ y If the absolute value of the deflection measured in air is less than the wear ring clearance, it is reasonable to expect that the failure to start phenomenon will not be present. It can be seen from Table 4.5-1 that the CS motor thaft deflections are considerably less than the LPSI motor. It should be noted that both the upper and lower CS motor bearings are the same as the ,
, original LPSI motor bearings. Preloaded bearings were not used.A
; further slight reduction in deflection would probably be obtained
?
using a preloaded bearing. After the deflection tests were run on the CS motor, the 100 start test was successfully completed to demonstrate the reliability of the CS motor on the LPSI pump. . 4.5.2 Data Reduction and Analysis of Test Data a Visicorder traces from typical LPSI test runs for both failure to start and normal start cases have been analyzed. The evaluation by measurement of visicorder traces, established the validity of the hypothesized failure mode [(whirling of igeller as it rolls around the inside diameter of casing. ring"lSeeSection4.5.4), and also established the pump rotatien atD!5 r , maximum during a ,, typict) failurt to start occurrence 81.e. approximately 2 revolutions in G seconds)] LPSI pump performe.ra curing starttp tests have been analyzed using an i analog multi.:.bnnel magnetic tape recording dtvice. This device had seven input ffenals; two L%ent Displacement Transducers (LDT's) at right angics r.a the top motor twaring shaft extension, two LDT's above j the lower mctce bearing, hL $0* angles, two LDT's blow the motor hearing at 97 angles, ana finally, one magnetic or optical pickup to
' determine shaft rotation angle and instantaneous pump snaft rgn.
k _x____ m.- _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _ . _ . _ . _ - _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ . ___ __
1
.,j These signals were recorded simultaneously on tape. These tapes were
, converted to digital form and stored as a set of computer files.
These files were accessed to examine in detail some specific LPSI failure to start runs. One specific evaluation was made to determine
;, if the ben' ding of the pump shaft occurred in one plane or in multiple
' planes & ring the startup time period. The latter hypothesis was confirmed. The objective of this effort was to deduce the forces causing this deflection and hence evaluate the alternative pump motor shaft deflections when subjected to the same force vector spectrum. <
This would have been used for motor redesign or for screening alter- )
,.' native pump motors for use at Palo Verde. The results from measured !
deflections in air tests however are more readily usable to screen o alternative motors. This latter approach was finally used. l 4.5.3 Testing Done at Westinghouse in Buffalo u i
, Two motors were obtained from a cancelled System 80 plant for possible i use on the Palo Verde LPSI Pung. The shaft diameter at the rotor on these motors is 6 in. which is the same as the Palo Verde CS motors and other motors which have operated as designed in other C-E plants. I w One of these motors was tested at Westinghouse in Buffalo to measure t shaft deflections. In-addition, there was a CS motor it Buffalo for
[:i repair and two Palo Verde LPSI motors were returned frt.m the site. ] P Deflection tests were run on all 3 different motor models. The !
;- results are summarized in Table 4.6 2. As can be seen, the )
P deflections for the alternative LPSI motor are comparable to the Palo Verde CS motor and much less than the Palo Verde LPSI motor. . ti As with.the CS motor the alternative LPSI motors do not have pre-1 loaded bottom bearings. The results of these tests corroborate the
]
'2 tests done at Falo Verde on the LPSI and CS motors. They also show the alternative LPSI motors are acceptable replacements for the Palo d Verde LPSI motors.
'l
}
9 l 1 j l l l
i t rl . 7 0 6 f v ii x a am M a 5 1
'4 1
2 h n, 's i r e e . h nl ol p t m ie tp u f
._ P o
. cm ei I 3 9 d t . 4 S n n0 e/
g v 4 2 2 P L e DW A 1 1 r i m e t w f r o l r e E e h l t D l R e t E p a V n a o 3 t d
- O i . 2 6 3 e L t x g r A c a 9 2 4 n u P e M 2 2 1 i s
- s S n u a e
1 T.R el ' di ' d m d.4E S r,m e e t e n r R l r i a E a LN li ea 5 3 2 t r BO p g
. 9 b e AI nn v 6 0 9 o l l
TT C Ii A 2 2 r e E
~ -
t o p L l o i m F a m E i t D d S u a C o " T F r n h t A s e o i H s g r S g n a n w n i o r i r n i i r a w t 'a a e o c e B h e n B s l i d f l e ' s e s a d n d n n a ) o o i g o 2 i r i l t e t i r e r e c l c O P t o e l e N l f e p l f r r ( e m e o o D I D t t r o o o M M t . . . o 1 2 3 p p M m m u u p P P m : u s I I P e S S t P P S o L L C N
?Y
t rl . 1 0 2 f
. ii x a an M a 5 2
4 8 h s
, n, ir e e h t
nl ol ie f tp o cm d ei l . 8 5 4 . n f0 g r e e/ v 2 3 5 e r DW A 2 l l e e w p o i m l h e I S t P L t O a L n . 9 3 5 h A o x t d F i a 4 8 9 i e r F t M 3 w u U c s B e n a
- - l u e S f r T el m 2L dl e e r a
- U r e r
- 5. ES e , w a 4R l r r
li s EN ea . 4 6 0 r e LO p g o l BI AT Ii mn A v 9 2 6 7 t o l e TC m o E . m L l e i F . a e E - i r t D ~ d h u a t o T r h t F l A e r l i w H a S a r r n o i w f a o h s n s t i _ s s e s n t n o o i r i r r t e t c o c l t r o e l e o o t o l e p l M t M f f e o e m I M D I D I S P S S L C P . . . L 1 2 3 e d e e d r r t a . e e n V V : - r S o o e E
- t T l
P a l P a l A O N
,g
p 4.5.4 Conclusions Based on Test and Analysis Results
,t The direct cause of the failure to start was theorized to be hard
'!! contact between the impeller and the wear ring resulting in the tape 11er becoming locked in a whirl mode. A whirl mode is defined as an epicyclic motion of the impeller within the orbit defined by the wear ring diameter. This theory has been confirmed by detailed
' examination of the test data for the failed start. This data indicates that during the fafled start a similar condition exists.
The motor accelerates to 100 RPM in less than .05 see when the whirl
' mode was initiated. The motor speed drops very rapidly to 25 RPM and
- then remains constant for the remainder of the test, approximately 6 sec. The ispeller whirl frequency is 130 HZ. Based on the geometry
' of the impeller and wear ring and a 25 RPM shaft frequency, a whirl frequency of 165 HZ would represent a pure rolling whirl. Therefore, during the failed start the impeller both rolls and slides around the wear ring. This action creates the r. characteristic wear pattern on the wear ring which is the principle physical evidence that a whirl phenomenon has taken place. Analysis indicates that the contact force between the wear ring and the impeller due to centrifugal force is large enough to both maintain the whirl mode and to have a friction component large enough to balance the motor torque. While turning at 25 RPM the motor is producing less than 6 H.F. For the 6 sec until the test is terminated less than 25 BTU would have to be absorbed.
This amount of enery would increase the temperature of the inpeller by less than 2*F. With the failed start requiring such a small amount , of energy to be absorbed, the lack of any significant physical damage
, beyond casing ring displacement is not surprising. -
It was hoped that as the evaluation program developed, the nature of the lateral startup force would be determined, either from the test data or from basic theory or a combination of both. A study of the available literature confirrtd that electro-magnetic forces in a motor during start are not well understood. None of the analyses were able to produce a verified model, one that could reproduce test data Lng predict har different motors would behave. The main structural l difference between the three motors is that in the rotor region the j Palo Verde LPS! motor has a 5 3/8 in. diameter shaft and the other two < have 6 in diameter shafts. In bending this makes the small shaft 2/3 - I as stiff. Electrically the CS motor is 800 HP and the other two are f 500 W. Using a ratio of actor HP to shaft stiffness the alternative i LPSI motor would be rated best and the LPSI and CS motors would be ' essentially the same. The test data shows the CS and the alternative LPSI motors to be essentially the same and_th ANPP LPSI motor hyl greater deflection by a factor of at least(2.h with impeller and ,4.2) without impeller. Differences of this magnitude cannot be explained using the present models. This indicates that there must be something very different about the electro-magnetic force in the Palo Verde LPSI motor and/or something very different in how the shaft responds to L that force. Until that difference is understood any attengt to j redesign the Palo Verde LPSI motor could not be analytically < supported. An enpirical solution was therefore chosen, namely 1 substituting an existing motor of proven design.
-24 }'
6 s
? .
i<
.U . <
.} The conclusion has been reached that the failure of the pump to start ,
is dependent upon whether or not the igeller strikes the wear ring in
.just the right way to initiate the whirl mode. The large impeller
- deflections which cause the problem, are not random. Every test of the Palo Verde LPSI motor, whether with or without impeller has produced large startup deflections. If the deflections are less than the clearance the impeller does not strike the wear ring with sufficient force 'to initate the whirl mode. A satisfactory motor is ,
selected therefore by using shaft deflection as the acceptance
- criterion.Cogarison of the ispeller deflections of the alternative LPSI, Palo Verde LPSI, and Palo Verde CS motors demonstrates that either the alternative LPSI or Palo Verde CS motor satisfies this criterion.
h' e e e e 1
~.
t~ . ,.
- h 5.0 Design Modifications l 1 L 5.1 Modifications to Motor and Pump to Minimize Casing Ring Contact i }' I
- As noted.in Section 4.5.1.1 four of the five modifications made during
, the test program will be retained in the final resolution of the
[ t problem. A detailed discussion of these modifications follows.
~
" 5.1.1 Nitronic 60 Cas'ing Ring ,
, By observation of pump failures and close examination of pump
' components, it was determined that the igeller[,as w maintaining constant contact with the casing ring rather than sliding on the
. casing ring This is the " whirl mode" discussed in Section 4.5.4. It I was detemi]ned that the CA6MM material of the impeller and K-Monel material of the casing ring were very close in hardness properties l which would make them relatively " sticky." A change in the casing
, ring materia 1from K-Monel to Nitronic 60 was made.
Nitronic 60 is formulated to be an anti-galling material. It contains 3.50 to 4.50 percent silicon which reduces the galling tendency.
. Published galling resistance data by Armco (Nitronic 60 supplier) and 3
data from I-R tests indicate that Mitronic 60 is cogatible with the
~
CA6NM material of the impeller. Nitronic 60 compares with 304
~
stainless steel for corrosion resistance and is non-corroding in bortted water. Another prime characteristic for casing ring material is its ability to retain its position within the suction ring in a
; design thermal transient. The coefficient of expansion for Nitronic 60 between 75 and 400*F is 8.8 - 9.2 x 10-5 in/in'F. The casing ring is mounted in the suction ring which is made of 304 stainless
- steel having a coefficient of expansion of 9.4 x 10-5 in/in'F. The i (ynamic coefficient of friction for Nitronic 60 is 0.42 which is en improvement over the 0.56 obtained with K-Monel. It is concluded that
,Nitronic 60 is an acceptable replacement material for casing rings.
5.1.2 Impeller to Casing Ring Clearance f he original technical manual limits for these diametral clearances is [p.023 to 0.030 inches. The diametral clearance was increased to[0.029 I to 0.03$1nches. This is a direct way of reducing potential contact between the ige 11er and casing ring. The increase in clearance can also reduce pump performance. Opening the clearance along with serrating the igeller, discussed in 5.1.3, below, was expected to decrease the developed head by two to seven feet. Investigation has shown however that the ECCS analysis is still valid. This performance l has been confirmed by testing the modified pumps during full flow ! performance testing. l
. i V
i .
n lf* t: 4 e I. F- 5.1. 3 Serrated I w eller s' h! Normally it is reconnended that the hardness of the impeller material be substantially (i.e. 50-150 SHN) different from that of the casing L ring materi'a1. Evadlation of available materials dictated that
. increased clearances and/or serrated inpe11ers be used. A serrated l' i impeller will provide a reduced surface area of contact with the -
. casing ring should contact occur. Figure 5.1-1 is a sketch of the serrated impeller. The selection of this design provides for vortexing in the area of each serration which will oppose the bypass flow (discharge to suction flow between impeller and casing ring), i The net effect is to regain some of the pump performance lost .in
. increasing the clearance. These effects were combined in the evaluation discussed in 5.1.2, above.
5.1.4 Motor Doweling The assembly tolerances allowed in the pu'np/ motor stack-up were
. reviewed in an effort to ensure closer centralization of the impeller within the casing ring. One of the areas of greatest allowed tolerance was in the alignment of motor to support head. This tolerance has been reduced from 0.005 inches total indicated reading to 0.002 inches. In addition, to maintain this close alignment the support head to motor interface has been doweled. Since the final pump assembly cannot directly measure centralization of the impeller -
within the casing ring, it has been inferred. by measuring centralization of impeller within the stuffing box extension. The centralization of the casing ring to pump bowl interface with the stuffing box extension has been verified.
.t 5.2 Differences in design and justification of use of Containment Spray vs. PVNGS LPSI Motor
' The general concensus of all concerned as to the cause of the failure to start of the LPSI motor is excessive shaft deficction. This exces-sive deflection under the starting forces results in contact between inpeller and wear ring. The motor then trips out on time overcurrent.
The chief difference between these two motors relative to the failure - to start problem is the larger shaft diameter at the rotor of the CS motor. The Palo Verde LPSI motor has a shaft diameter at the rotor of 5-3/8". The CS motor shaft diameter is 6". Both motors have been tested to determine shaft deflections in air both with and without the LPSI punp ige 11er attached. The CS motor showed significantly lower shaft deflection than the LPSI motor. See Tables 4.5-1 and 4.5-2. In addition, the CS motor was started successfully 100 times witn 20 minutes of running after each start. Subsequent inspection of the , wear ring showed only slight contact opposite the discharge on the lower ring. An I-R metallurgist stated that this was normal for pumps operating successfully in the field. All interested parties concurred with this assessment. - l The CS motors have been operated during system tests and checkout at the Palo Verde plant and have experienced no failures to start. j
s .. g
- IMPELLER SERRAT10N5- ,
m e, ' " ' L, ] , J9 ., eresen .
- ~.
p - 1 l- - j t- -
- . 1 4
s
. :- 's -
t. 1 i l
' l
. /
- f. Y I
ri . (/ ..
~
l t
. l A ,
\
*. .- /
u
,/
- i /
u . *
/ l
- .- i g
J 1 j
- Figure 5.1i I _ _ _ _ _ _ _ _ _ _ ._.
C , k
'I
- e. . ..
1
~.
b l n . i -
},
I ; 5 I . 1 t i I. i i
\
APPENDIX A i r e e b 9 k h i f
+i rs
- O' .
TABLE 1-1 y a ,.
.L FUNCTIONAL DESCRIPTION ,
('. FOR t. f 8X20 WDF. LOW PRESSURE SAFETY INJECTION /LPSI) PUMPS
- WITH . ORIGINAL MOTOR ~ "~ ' '
.I ' ~
ingersoll-Mand Co. f , PUMP MANUFACTURER
'
- Centrifugal
. TYPE MODEL ,
EX20 WDF ! SE RIAL NOS. Unit 1: 0976- 36/37
- Unit 2: 0876-40/41 Unit 3: 0876-44/45
?l LPSI Pump SERVICE LlQUID PUMPED 40 Dog. - 400 Dag. F.,(8 orated) Reactor Water SPECIFIC GRAVITY OF LIQUID PUMPED 1.0 - 0 3 I
. PUMP FLOW 4300 GPM (Rated) l i PUMP SPEED 1790 RPM (Rated) t TOTAL HEAD f 335' (Rated)
SUCTION PRESSURE 450 P.S.I.G. (Maximum)
-, NPSHR
- 18.0' (Rated)
NPSHA 20.0' (Rated) EFFICIENCY . 82.0% (Rated) PUMP BHP (9 SPECIFIC GRAVITY 1.0) 444 (Rated) I PUMP BHP (9 SPECIFIC GRAVITY 0.9) 400 (Rated) PUMP ROTATION CCW - Hi HYDROTEST PRESSURE (CASING, STUFFING BOX EXTENSTION, AND GLAND RING ONLY) 1140 PSIG
- Westinghouse MOTOR MANUFACTURER 5,
' 500 (Namepiste) r MOTOR HP 1.0 SERVICE FACTOR t' . ,I -Al-l t1
7' J2 , ,. . APPROXIMATE WElGHTS ,
~
PULL-OUT ELEMENT MINUS POTOR (STUFFING 80X EXTENSION, E ROTOR AND SUPPORTING HEAD) 2407 lbs. 1
~
. A40 TOR ,
, 4500 lbs.
.' TOTAL WElGHT OF PUMPING ELEMENT 8907 lbs.
t t BALANCE OF PUMP (CASI'NG, SUCTION RING, DIF7USOR) 2135 lbs. l TOTAL WEIGHT COMPLETE ASSEMBLY 9042 lbs. (Dry) 1000 lbs. (Flooded) s .
+
9 e
- 9 S
f. s
's:
e a b L - r. I i
- . .A2- -
t-
,y .,+
~ .
rg _ P ... -
.L ... ,. >.
g 7 .
-f 4
* .~< 'O m <
.N Qu <eo .
Ne m <g e .a e.
- 9. e- 96 . g aN O r g
; .- eem mme 2_ f,
~
i .. I 4
- .f a
j
,b % ?
- s \ -
A / N
- s. / . .
,', =
s,
/.
m-l' ) A
)j n
'y N '
e-N,,,1 /
'M
. ~ /
g .
- . r. . .-
m. f .,(' - - _ _ _ q
' a l .
% ${ ,
N
// / ,I
/
N
. y
~
\ - -
-N / .' / 1
) '
i
"( o l
> j 3 ,
.I ,
I .dl
-A3- Finure 1-1 .
Page 1 of 3
u .. b - i F PARTS DESCRIPTION 1. I CASING (1)
', (a) Suction and discharge nozzles prepared for butt welding to the
; system piping.
- . Suction ring fit for the suction ring (269).
! Gasket (36'3A), and stuffing box extension (264) fits.
Pressure bolting to attach the stuffing box extension (264) and
, supporting head (172) to the casing (1).
SUCTIONRING(269) The suction ring is bolted to the pump casing (1) and acts to direct the flow of liquid from the casing suction nozzle toward the pump impeller (3). The suction ring contains mounting fits for: (a) A renewable wearing ring (6). Gasket (3638) (qb) C) Diffusor (56) which converts fluid velocity to pressure energy. Rotation of the diffusor during pump operation, is prevented by a drive loc pin (26A13) which projects from the suction ring (269) and enters a mating hole in the diffusor. IMPELLER (3) The impeller (3) is a closed type, one piece construction, single suction. It is hydraulically balanced and also dynamically balanced. Inge11er balance holes and close. impeller - to - stuffing box casing ring (6)/ bushing (88) clearances limit the effective pressure on the , stuffing box and balance axial thrust. STUFFING BOX EXTENSION (264) The stuffing box extension (264) is gasketed and bolted to the casing (1), and contains the mechanical seal (429) which seals the liquid from the atmosphere at the shaft area. It also contains a renewable wearing ring (6) and bushing (88) which aid in reducing the pressure at the area of the mechanical seal. SUPPORTING HEAD (172) The supporting head (172) is bolted to the stuffing box extension (264) and casing (1). The extended shaft motor drive is mounted on
; thesupportinghead(172).
l6 j ( l i e
. Figure 1-1 )*
Page 2 of 3 j
-A4-
{
L e-
-V r . ,L 4- .-
lb<
- I CASINGRINGS(6)- -
l0 Two renewable casing rings (6) are furnished to maintain running !I clearan.ces between the inpeller (3) and the stationary mating fit. l4 One ring is located in the suction ring (269) and the other is located !~ ! inth_estuffingboxextension(264). Both are thermally shrunk in
~~
r place and further secured by two lock screws. ECHANICA1. SEAL (429)
! A mechanical seal (Durametallic Type PTO) is employed to prevent fluid in the pung from leaking to the atmosphere.
The mechanical seal embodie:s the radial or "end-face" principle - that of sealing in a plane perpendicular to the shaft. It is designed to give long life, reduced maintenance, and automatic self adjustment. 4 i 1 Figure 1 1 Page 3 of 3
-AS.
~
4 LT e.
=
t . [f CASINGRINGS_(6) . I Two renewable casing rings (6) are furnished to maintain running
?
clearances, between the impeller (3) and the stationary mating fit. One ring is located in the suction ring (269) and the other is located _ in the stuffing box extension (264). Both are thermally shrunk in
- - place and further secured by two lock screws.
ECHANICAL SEAL '(429) A mechanical seal (Durametallic Type PTO) is employed to prevent fluid in the punp from leaking to the atmosphere.
- The mechanical seal embodies the radial or "end-face" principle - that
, of sealing in a plane perpendicular to the shaft. It is designed to give long life, reduced maintenance, and automatic self adjustment.
. i
..s I
i Figure 1-1 l
)
Page 3 of 3 !
-AS-l
a . n
!L , ,.
L CASING RINGS (6). .l. Two renewable casing rings (6) are furnished to maintain running i-clearances between the impeller (3) and the stationary mating fit. One ring is located in the suction ring (269) and the other is located ! .. in the stuffing box extension (264). Both are thermally shrunk in
, place and further secured by two lock screws.
ECHANICAL SEAL l429)
. A mechanical seal (Durametallic Type PTO) is employed to prevent . fluid in the pump from leaking to the atmosphere. . The mechanical seal embodies the radial or "end-face" principle - that of sealing in a plane perpendicular to the shaft. It is designed to give long life, reduced maintenance, and automatic self adjustment.
e l l l l, i r f j Figure 1-1
; Page 3 of 3
-AS- 1 i l
\
f,; _ LOW PRESSU RE i - - SAFETY INJECT 10M PUMP I-i - -
~
RADIAL BEARJNG
. ~TN i -
~
Y////5 kg'mm/5 l l wo om RADIAL AND THRUST BEARING ,N I l 1( )
.&W t a h
a
. s
. - i
. i, .
y 4 s / /1 r///4 <
, ~~-i; C
$ s 1 .
55 l ;
- 1 A -
r L Y /7 CASIHri RINGS ii 1
'A
/
. J. .
1 ' I Figure 1-2 -
-A6- ..
a
g g . o,
- r. *.
(f-muss E g i ~. g
- k,. )
P e 9 e, b 4 0 i P e 4
?
5 ' 4 APPENDIX B O 1 i i 1 l L___.__._ __ _
e s
'PALO VERDE j LPSI PtN RELIABILITY l
i
~i i
Prepared by the ~ Reliability systems Group Nuclear Power Systems Division I Combustion Engineering. Inc. l Windsor, Connecticut
. 1 i
l-i l l.
-B1-
}}L t::g L:r-
~'
TABLE OF CONTENTS 1'[; ..
- t. .-
$ECTION- TITLE Lfst of Figures List of Tables 1.0- PURPOSE 2.0
' SYSTEM RELIABIL!iY REQUIREMENTS
- . 2.1 LPSI Systam 2.2 Shutdown Cooling System
[ <, 3.0 .. g . SYSTEN RELIABILITY MDDELS 3.1 LPSI System Reliability Model 3.1.1 LPSI System Description 3.1.2 Assumptions 3.2 Shutdown Cooling System Reliability Model 3.2.1 Shutdown Cooling System Description 3.2.2 Assumptions 4.0 LPSI PUW RELIABILITY REQUIREMENTS
,5. 0 CONCLUSIONS O
e I
.t: .~
f. i. t-
.c t ,- - ; LIST'0F FIGURES i l t
FIGURE NO. . TITLE 3.1.1 PVNGS Safety Injection System - (Injection Mode) 3.2-1 PVNGS Shutdown Cooling System (LPSI and CS Pump Operation) 4.0-1 LPSI System Unreliability versus LPSI Pump Unreliability 4.0-2 Shutdown Cooling System Unreliability versus LPSI Pump Unreliability LIST 0F TABLES TABLE NO. TITLE i 2.1-1 LPSI Unreliability l I l
-B3-
[ ., 4 pi '
- 1.d
- PURPOSE 1
!' The purpose of the work presented in this report is to determine the number of successful LPSI pump start cyles that demonstrate an g acceptably low LPSI pump unreliability.
h 1' 2.0 SYSTEM RELIABILITY REQUIREMENTS-p A survey was performed of reported Probability Risk Analysis (PRA) L results to determine a range of system reliability values which are I typical of operating plants. This range of sysen reliability values I i
- is intended to determine an " acceptable" value for Palo Verde LPSI and Shutdown Cooling Systems.
q
'2.1 (PSI ' Systa,m, Seven PRAs were reviewed to determine the calculated unreliability of
- 1. PSI. Results of the su "' r*Pr'5'"t'd '" Tabi e 2.1 . The highest value (4.7 x 10~S") 'an'd the lowest value (2.5 x 10- )
define the range of acceptability. j 2.2 Shutdown Cooling System The Shutdown Cooling System (or a Residual Heat Removal System) has not been evaluated in PRA studies in general and therefore a range of i typical shutdown cooling system unreliability is not readily deter-mined. Shutdown Cooling System failures have not been found to be ; dominant contributors to risk. An alternative method of determining ! the number of LPSI pump start-up tests that are required to demon-strate adequate shutdown cooling system reliability was employed. This method involved performing a parametric study of the LPSI pump failure rate and the shutdown cooling system unreliability. The parametric study was used to evaluate the influence of LPSI unreli-ability on shutdown system unreliability and to determine the point at which LPSI pump unreliability so dominates shutdewn cooling ) unreliability'as to be unacceptable. The nuu6er of LPSI pump startup j tests that are required to demonstrate the associated LPSI pump i unreliability can then be determined. The shutdown cooling system ! unreliability was evaluated by performing a fault tree analysis as 1 outlined in Section 3.0. ! 3.0 SYSTEM RELIABILITY MODELS
)
Fault tree analysis was employed to determine the reliability of the PVNGS LPSI and Shutdown Cooling Systems. The methodology employed in { this analysis is consistent with and is described in CEN-239 i Supplement 3. The fault tree models were used to determine PVNGS LPSI l and Shutdown Cooling System reliability for various LPSI pump failure rates. ' A system description and list of the assumptions made in I p'erforming the fault tree analyses are provided. The results of the ,l
- l. fault tree analyses and probabilistic evaluation are used as input to 2
determine the LPSI pump reliability requirements and are discussed in Section 4.0. [ _f
pr '
?
i K . s. . t,.. i
]
t' . .; TABLE 2.1-1 ,
~~
LPSI UNRELIABILITY_ PRA UNRELIABIt.ITY REFERENCE anector Safety 4.7 x 10~3 Wash-1400 Study Page !! - 137 l I
. German Risk 1.7 x 10-3 The Federal Minister of Study Research and Technology.
August 15, 1979
.{
3
! Ifen Safety. 2.5 x 10~4 Connonwealth Edison i Study . Study. Page 1.5 12 -
Seguoyah 1.9 x 10~3 NUREG/CR-1659/1 of 4 RSSMAP Page B.7 - 10 j econee Unit 3 2.4 x 10~3 NUREG/CR-1659/2 of 4 RSSMAP Page 8.6 - 14 C4 Generte 1.0 x 10~3 CEN-153 PWR B 9 ) 1 . 7 e
h f 3.1 (.' PSI Sistem Reliability Model l 3.1.1 System Description A schematic,of the PVNGS LPSI System (Injection Mode) is presented in Fi gure 3.1-1. The injection mode of operation is initiated upon receipt of a Safety Injection Actuation Signal (SIAS). A SI AS is r produced upon any two coincident low pressurizer pressure (1700 psia) or high containment pressure signals. The SIAS may also be initiated manually in the control room. Upon a SIAS, the LPSI pugs automatically start and the LPSI header isolation valves open. During injection mode, the miniman fiev lines downstream of each pump are kept open to prevent possible dead head operation. The pumps take , suction from the Refueling Water Tank (RWT) and discharge through four LPSI header isolation valves via two redundant LPSI headers. The safety injection water then f1cus to the reactor vessel through a safety injection nozzle on each of the four RCS cold leg pipes. If offsite power (normal AC) is unavailable, the ESF buses are connected to the diesel generators and safeguard loads (the LPSI system) are then started in a preprogrammed time sequence. When the RWT level drops to its predetermined low level at the end of the safety injection phase, a recirculation actuation signal (RAS) is generated. The RAS signal stops the LPSI pumps and ends LPSI System operation. 3.1.2 Assumptions The LPSI System fault tree analysis is based on the methodology egloyed in CEN-239, Supplement 3. Cogonent failure data used in the probabilistic evaluation of the fault tree was also derived from CEN-239. The following additional assugtions were made in performing the
, fault tree analysis for Failure to Delivery Sufficient LPSI Flow:
- 1. System failure is defined as the inability to deliver sufficient LPSI flow to the reactor core. Sufficient LPSI flow is defined as one LPSI pum flow to two RCS loops.
required to deliver the flow from one pump ). (Two flowpaths are
- 2. Isolation of the pug mini-flow lines could result in dead head operation and damage to the pumps.
- 3. The only operator action considered was manual backup of SIAS from the control room.
- 4. It is assumed that components on Train A receive SIAS-A and cogonents on Train B receive SIAS-B.
- 5. To facilitate tree construction and evaluation, it was assumed that cogonents on Train A are aligned to Train A electrical supply buses and components on Train B are aligned to Train B electrical supply buses.
l
-B6-
.+
- i. ? . !. != i, .
it it it it it il o n .. n n n +. N ~~ - Z Z 2 2 Z 11 ;
, I
.l .l -
.l .I N-- Ii 1 I 1 -
1 u i Z u l Z u I Z u I Z Z , 1 ( C C C
^
l
- Z fZ fx fZ FZ Z =
@i 8
@MII @2 @X I @M 1 @M!I @
ll@&@?_~!np I k w _, g8 egl
. I i
I 1 a(I 1 C Ei 2 I
@i :I 8 .1
@i :I 1
5 X t, 3 E '
' " I E -
E @K @Kg g j @K Z y e
. 2 I
*: g- _ _
O !! O.!I TT*d'iId!! . I II e . r - __ p . .____.
= -~
w 4
0 W-g. gc Dh , , . . , . b< 6.. Maintenance error resulting in the. inadvertent isolation of a LPSI pump has been' included in the analysis. -
- 7. , Palo Verde's periodic test' and maintenance program is assumed to J prec'lude wear-cut. type failures.
Si Palo Verde's pung modifications are assumed to be adequate to b prevent th.e failure mode.
. 3.2 Shutdown Cooling System Reliability Model
~
.3.2.1 forstem Description A schematic of the PVNGS Shutdown Cooling System (SCS) is presented in Figure 3.2-1. ' The SCS is a forced circulation heat removal . loop designed to transfer reactor coolant sensible heat and core decay heat from the RCS to the Essential Cooling Water System (ECWS). The SCS
^
consists of two independent trains, each utilizing one LPSI and one Containment Spray (CS) pump to recirculate reactor coolant through a
'shutdowncoolingheatexchanger(SDCHE). The SCS is manually put into operation after reactor coolant temperature and pressure have been reduced to'approximately 350*F and 400 psia respectively. Shutdown
; cooling is initiated using.only the LPSI pumps with the CS pumps providing additional shutdown cooling flow when reactor coolant temperature decreases below 200*F. Shutdown cooling is then continued using the CS and LPSI pumps in parallel until refueling temprature is attained.
l 3.2.2 Assumptions The SCS fault tree analysis is also based on CEM-239 Supplement 3. The fo11aving assumptions were made in performing the fault tree analysis for Failure to' Provide Sufficient Shutdown Cooling Flow:
- 1. System failure.is defined as the. inability to deliver sufficient shutdown cooling capability to the reactor core. Sufficient shutdown cooling capability is provided by one LPSI and one CS pump, the corresponding 50CHE and 2 RCS flow lines. Sufficient cooling is also provided by operation of two LPS! or Two CS 1
pumps, however both SDCHE's and four RCS flow lines are assumed required for the mode of operation. 2.. The only operator action considered was manual alignment of the shutdown cooling system trains. It was assumed that the operator recognized the need for shutdown cooling and attempted to start 1 all pumps.
- 3. It was assumed that components on Train A receive power from i
4 . electrical distribution system Train A and components on Train B receive power from electrical distribution system Train B.
- 4. Failure of the RWT isolation valves to close does not inpact
< shutdown cooling operation.
-88
- n1h _ __ __ __
;i! !
e 4 . -
*~
=
P
** n n .
nT n :
~ -
=
- n. n7 n . ' ,s .
eW.2..
. g i
~' .
;. ~ s T
n ._
~
Im a- ,. na e M ~ n. T . . E
. i .g
- w. .
i , . n I
, .l .g, E
- l. -
m "1 ~ n .
= 9 *" kM ._ =:::
y
%::.Q. r* ***
s .,3 n
- : a ._ :
e : . *
= " ._ :
. ,** R -
i l
- Xl _7 gxW _e O- " _ -
9 3-
=
p Q
^ n U..
W :~ .*
~.;~
" .o 9.-
i- ? - ( . M) E TN _ - SO
- - . YI
- 1. ST A
K X GR O
- e 1 t
fE iP LO
- 0== 200P 3 CM j l U H NP H EW
'l O e R OS U DC GT H n .
I UD e f e F HN SA (}'c~-X g f
. M M .
rGg~"s, g J= rs=== SI GS 0 fP VL p
~~- x P
(
]l -
- .9 -
r ( m " :=, h,~ , 0tSu e w
- w R =- .
jl
- e. - - - .
. H@ .
e w
. N. N N .
a.
- .w ,
8 l
- o. I -
iu 9::==. .ll _ :: s: .
=
o e
= : ..w km N"=-
.m :
.; W .
e A =: ===c
=
=i ce.
c
- m .
=
=
- ,o, t
jl .'
i s ' Mi n ' ~19 y . g, , i
, r .g ' ':
- d ' ,. , <*,'t *
- i \ u
.. Dl, M ,t '3 k ~
; m ,a x
' ; yh . t s , .
, o '
N
+
n g , M&. ' k Palo Verde's periCic test and maintenance progras.'is assumed to I ,s._, } K+: preclude a waarout type failures. ;3 A i
-?3 ' uhf N
- q. p. Palo Verde's pump modifications are assumed ,to im adequate to E '? ) spr,e,v9dt thi failure mode.
F ;%. ' i o1
%{3 60Y'LI,PS'lPUMPRD.IABILITYREQUIREMLNTOA aq g V Q e T3 s(
;w ;
["b t W ! Thi resets ed the ..f$. . . ult tree,analyns for the LPSI an6 Shutdown 'i - J.) 8/ dQrmling Systist arsipown fra Ff; pares 4.0-1 cadd 4.0 2 respectively. t', j ' T Y , The results are precented as p;1ot of the unreliabilityrof the LPSI t-
% stem,(cr Shutdown Cetning $ stem respectively) as a function of LPSI "' -
m
.g p
a ' ' gwp unr?1iability. The LPR puuy uprelf abi ty for failure to stait on demarewas varied from IWY/ demand tojo p/ demand. The erves , generatid afe used to relate the LPSI $ stem and R$ unreliebf11ty to - /
- s LPSI pimp i, unreliability characterf stic. '
i
, p >
s
.s i
. s It haa been proposed that'10W LPSI pump tests be conducted to verify .
'{ pump reliability if 100 tarts, are run with n's failurers observed.
,i then the testing will have demonstrated #Lnomins1 LPSI pump \ ; , ,j unreli stlity (probability of fail to start) that can be detsrmined.
-(- from thetchi-square distribution. The nominal unreliability at 50% '
i s confidenjeis;cogutedas -
'} \y . 22 .50.'2n + 2 1 2D- where t !
.s ,
3 = number of obser;ved failures, 0 failures, 1
, n '
.. i
- m '
D = total number of tests,'100, tests ,_
! .c The' tabulated v'alus of the X2 distribution'is ' i .e
- 'x 2.50,2 .= s1.39EI3
- u. '
The LDSI, pug uitreliabiltty is then (
. U = 1.39 = 7.0 it 10-I per domind -
NNO f
'7 tort Figure g.U.1, it can be /.eer. teat given a U'S! pump unreliability of [.0 :t-10~ , thesnreliability of the LPSI system would be 1.7 x lo This valte it..substantially-lower (i.ec, better) than thi.
e typicalivalues of LPSI system unreliability discussed in Section 2.le h , B , From Figure j.0-2, it caibe seen that given a LPSI pump unreliability [fe7 iy'
- of 7.0 x 10~ , the u would be 5.3'x 10"4.nWaWty As discussed).f the Ruh in Section 2.2,cWng a range Sysm of a
4 " s typical unreliability values rau1A not be determined for the Shutden Coolin2 System. Instead, a ptrawiterie evaluation was done to - ( ,. determine the degree to which LPS1 pug unreliability tunds to
'daminate Shutdown bling System unrelitbility. As can be seen in Figure 4.0 2, syssa unreliability would be relatively insensitive to further improvement (or testing) of the LPSt pumps.
qs
.s ,
y A e
.a10 ,
&sa
~
~ .
J g' k ,, df t. _ _ _ _ _ _ . _ _ _ . . . _ _ _ __._________-_U
o *. a. 0-,2 - FIGURE ,4.0-1
- s. . .._ .
.+...
. . . = .. . .. ._._ . . . .... . . 1 O,
y
- _ - - -
j i 0
- LOI 3TM LM UNMt;L A At3 AL A i1 - - ~ '
e .
/-
-~
6 . vs. LPSI Pt.MP-UNRELIABILITY - -
/
- , . . . t
- f
/
. s
- ~'~ ~~~~
. . . . f.
9 t I. _-. .. ile 1 .
- >. t10-3 -
I .
~
- I . .
/ .
/ _
_ . . . . . _ _ _ . . . . . . _i . .. s / ' ~' ' g-- __ _ _ _ _
.. _. _ _ _ _... .. ._ l
~
-8 .i
~3 '
E, . .. ._. . I E
.F , , . ! m.
3=
-M l . .
ca310 4 '
/ ... . .... .
, ~~~
- / -
i 5 j g. .. a
- 3. .
i '~ -
, ...m .
m P. h 2 I .. *-
- g. ..
p - i.. .... .. . . . . , _ _ _ . . . ., . . _ . . _
. .s. . _... __q._....
.._- _.a..._. 1 . ..s.
; _ .... . ._p.
;_ _ . _ _ .. p .
_..w_...F.._.._......
. . a ._. _ t_
e
.e;
__. . . _ . \. 10. 1 _0'3 _ _ ~ _ _ - 10:2 3o-1
h ' '#.I FIGURE l.0-2 .
- o. ,,
. . . ..gm..s.m . . .. . . ~.. ._. . ._ n --.- _
. u .s.. ._. . _ _ . . . . . . - - - - - - - -
_ .SHUTDOWt-dXi!N.G_ S_YS_TD_r(JN.RELI A. B IT.I.T._T.... . ___ . - ---
. -- vs--LPSP PUMP UNREl.! ABILITY- - _ . ._
g g _ .. .. . . ..
. , . .. . . _ . . . .. _ . . , 4....
l ._ .. . _ _ . ...... .. 1.
. . .. ___ _ . .. . _ ..i
.t.
i. __ _. ..i t . .. . s_.. . . _ . . - _ . _ . - - . _ . . . _ _ . . _ . . . . _ s . . . , . ._
- p. . . .
s_. . t . .. . . . a.s
- a . .
. e . . .
Q , d ...... E f i . . . m. m 10 . . . _ . . . . . . _ _ .
.E
.t__
i_. . . _ . . . _. . . i..._ . __,__
. . . . .l.
i...._....... .. .
.s. _ . _ _ _ _ . . _. _ _ . . _ -.
. t... . _. _._.
. ....-... _..i..._..
- _i
_. .l
, 1
. . _ _ . , --...p.___.w.._.. . . _ . ._... ! , ,.
.. ._. ._ ... . I. . . _ . . . . . . _ _ . .. .
.____o..___. _. . . . . . . _ _ . . . . _ .. i _. . , _ _ _ . _ - _ .
. .,.-._........s_.__....____
. i
__ a..._....__....
. i.._.___=., _ . _ _ _ . _ . . _ _
_. . . _ . _ . . _ . . I 4 _ i J r _ _ . .
,_ _ l
=. . _ . __..,
_....___.t.__... a .. _ _ _ I i h
.. l .
3 .. I J- , .. ,. '
} f
, T
[' . . 1 4. _ l Mcn>* . 10-4-- ._- - 0.c. n_ _~ _ __ ._
1 5.0 ' ~ CONCLUSIONS C-E concurs that with the performance of 100 LPSI pump start-up tests.
' with no failures, the LPSI pumps will have demonstrated sufficient
' reliability that the reliability of the LPSI function will be adequate and the Shuttiown Cooling function unreliability will not be excessively dominated by LPSI pump unreliability.
) e i
-B13-l
mn-r- - - - -, - - - - - - l-
, as g 8
J 4 e 9 APPENDIX C i
i
. l
,m ...
NRC Questions from 7/5/84 Meeting. - 1 Consnent 1: The applicant should provide a detailed failure analysis which ! addresses all the problems associated with the Palo Verde safety I system pumps. Response: This report covers the failure to start problem on the LPSI . Punes. The ' history of the problem, safety concerns, design modifications and test and analysis results are all addressed in the report. i Comment 2: The applicant should address the effects of using different ' and/or larger horsepower motors than those originally provided by the pump manufacturer. These effects should include pump performance and reliability. Response: See Section 5.1 for a description of pump modifications. The same impeller is being used. The Sarger horsepower motor will therefore have no impact on pump performance. As far as reliability is concerned the replacement motor is identical to { the motors used on the Palo Verde containment spray pumps. No I operating problems have been experienced with these motors. In { addition the motors 3. ave been tested on the LpSI pump. Details ' of the tests are included in Sections 4.5.1 and 4.5.3 of the report. Comment 3: The applicant should provide evidence that the safety system pungs can reliably meet the required performance flow and pressure head with any required design changes. Response: The modified punps have been tested in the system during full flow performance testing on Palo Verde Unit 1. The perfonnance met all requirements. Comment 4: The applicant should provide a statement of the pump manufacturer position regarding any design changes including changes in motor horsepower and type, i Response: The pump manufacturer, Ingersoll Rand (I-R) has been intimately ' involved with the problem from the outset. The design changes were sede at their reconsnendation and with their approval. The replacement motor is being supplied by Westinghouse, who manufactured the original motor, with the approval of and through the auspices of I-R. Comment 5: 1he applicant should address the effect of single pump failure on the systen reliability. Response: This is addressed in Section 1.3. It is also, however, part of the safety analysis. These punps are part of redundant systems. Please refer to the Safety Analysis Report.
-C1-
f- ~ l [ . .. . o i.- - l Conment 6: The applicant should provide a complete and detailed accounting
.of. any design changes.
Response:- These changes are discussed in detail in Section 5.0 of the report.- , coment 7: The applicant should provide details of test plans to demonstrate
; resolution of the safety system pump problems including testing
~
of pumps with the use of different and/or larger horsepower motors. I Response: Please see Responses to Coments 2 and 3 above and Section 4.0 of the report. I i l
-C2-
f
. ..* C.E Power Systems , Tel 2o3/6881911 Corn:sstion E ngeneering. lac. Yelem 99297 Attachment to DER 82-42 i 1000 P'espect H It Roac Wenesor. Connecticut U6095
/71%2*7
(' ,~ G2 POWER ' OL m SYSTEMS ,j,1,0 ~" 1. September 3, 1982 :
~ V-CE-17050 SEP19'82 '
Mr. W. J. Stubblefield w n sw.s -- b += Bechtel Power Corporation # 5-E,*"' P.O. Box 49 "** Palo Verde, Arizona 85343 I
' c% l
Subject:
Arizona Nuclear Power Project 2 w o= i Reactor Coolant Pump Pressure 1,, ' " Tap Replacement Unit 2 - Information us Required by Bechtel , O',"*' -
Reference:
DA -- (A) CE letter V-CE-17031 dated 9/3/82 ~~ Cik, Subject Unit 1 RCP Pressure Tap Replacement ]T,7, - 99 wt
Enclosures:
(1) C-E Manufacturing Order No. 9403834 _
"- 't l
(2) C-E Manufacturing Order No. 9230163 --
Dear Mr. Stubblefield:
(- In response to the request of the field for the subject pressure tap replacement work for Unit 2 Enclosure (1) invokes the revised pump specification which moves the boundary. The revised specification (in paragraphs 2.3.8 and 4.1.15) forwarded as Enclosure (1) to the referenced letter, moves the boundary of the pressure tap from the former field stainless socket weld to the end of the inconel tube of the pump suction and discharge nozzles. The replacement inconel nozzle coupling design has been approved by C-E Engineering and formal supporting documentation will be forwarded for the design. CE-Avery Field Engineering Service has been authorized by the Enclosure (2) manufacturing order to replace the Unit 2 suction and discharge nozzle pressure taps under their Section III code authorization in the field as you have requested. The above responds to ths field requested information by Bechtel to move th'e code boundary of responsibility on the l}CP Nc,tzles fcr Unit 2. Should you have questions, please advise. . ler 'w yours, ,
> 7 . -z.
;og . Fer sn ~
Project Manager CF/VSK/mg
~
Enclosure cc: Messrs: E. E. Van Brunt, Jr.w/e J. M. Dilk G. C. Andognini w/e
\ '- W. L, MacDonald G. A. Butterworth J. Vorees S. N. Mager w/e W. H. Wilson S. W. Shock u c- ,--t. ..<-
w/e g/ _ _
3 0 {0 o or -*.. & O .0 .O O O .O c o l:9 o o 'y C MBUsMN ENGINEWNG,1NC {
~'~ ~ ~ # * *
- __..yPO,h/ER 2 1000 PRO $PECT HILL ROAD
]
gf ] c0 US 4 geJ0: gggmNG sN , '
,O l r i'[
TEL. 203 EB81911/TWX 710 4471870/ TELEX 099 297 ! 7c, -JMsPPsNC T ACS, PACKING LIST $
.)
, *
- j
_ FW_ /,fN il j A90 SMI7PsNG -PAP (MS MUST INCLUDE '*6
~.5.LC"4 ~;1 G3 THE APPLeCAstE C-f CONTRACT N O., ^
{ C3MP NENT CODE NO AND ORDEM NO. ,'" vtwv2 8
*ND 686 Coot l To: o .
bhs uet.O ) ra N.I.. ~ ~ ~ ~ h n63o .. L .
~
g o 94,,,gyg,_. g_3,~ ~ teres emota Awo stLitas sus 08L3tak Amt susJtet to tangostwG ST Ut3 COCl4UNicAT80NS 70. , Pum:masta Aw."ca ars assert. . VEaM6' 20th of FOLLOWING MONTH suvan a c cr. ne. cowsmacy wo. a saci. snoesomno. " ~ ~ - 14373 l P .c .., . .re.-. c . . . u -
, ,,, , y , a , ,, , , y,,-
?1 EM tiUANT87 Y COMPONENT CODE D ESC RsPTsON gygwy pnggg 31-14-54-0200 . REACTOR COOLANT PUMPS
~
4 The following specification revision is invoked on this order. This revision supersedes all previous revisions: ( . . . . . CE Project Specification for Reactor Coolant Pumps for Arizona Nuclear Power Project No.14273-PE-480 Rev. 03. Five (5) copies of the above specification are enclosed.
~
)
,1NVOICE 00 NOT ~ **""**'"**** i .
_ TAX , INVOICE RE Gl5TRATION NO. '
. TAX / ,
EffiONAL INF ORfAAllON. # r a sa, m es.ca.c out Vendor: CE Avery o f 8. C R A Y l D #e ] A A C , h . euS E R E GWs5 n.n.6UB** OaKak C, m L A A w i mD m 61 Y (op i gj{'l16QQ lapp ~ Unit i2 cts. *m, W .h. 41)E . .. ,, . . v . ... _ . . . . . o, 14373 ^ ww k , $2.59 o4
- - , . . . , u , . . . . m ,, . .,o, co.<
- .~ . . . . o o . . . . . .. . . o. n 7h.mfk. . -
,. , m ., . t .e 1 :.wI A. . m . . . . . . -m,. .. c . au
3741 ;o' ;O L <- O G II .G C o . lJ (~ .L. 4 oi - A
'~O' (p c
. . .. . , ; -. . . . . ,,;,.,ugu,, , na,,,,,, 3,g, .
i ki- :1000 PROSPECT MILL MOAD lCf.v.,15173N.! "$oa ca"" **"5 TEL.203 8881911/TWK 7104471870/ TELEX 099 297 2'9,5E"*T iff_2 L-L ' L SMePPsesG Tacs, - P ACsCING LISTS , "
- p l,AW) SMIPPING PAPE ms MUST I**C LU DE
- M'S .l J'A. D.WJ TNE APPLIC A9L1 C.E CONT R ACT . NO.,
C3MP3NENT COJE NO., AND OADER NO. , vnu.a2 c E :I ..._.. .. U ENEEo"Us'nE'**
~'"**' *" '* *~* *** *' '" ***'."
' TO: - 'o-N -- C. O l'"/ 0 3 o N.O.923oi63 L . . . . . .
. 5
- . CO,.s.uuN Civio io.
fr "* 20th et FOLLO'.VING MONTM PWYca O'a:f!" 'an"'t'!r a ccT. ac. CONTaacT No. ***"'4 sect '""."*s"wo*r "o's*s e 14373-
- R.,
P.ea.rsenT smarreno enamots ...,v,. 4 TEM AUANTaTY COMPONENT COOC DES *RIPT80N EVENT PasCC - - 31-14-54-0200' REACTOR C001. ANT PUMPS
~~
1 Install inconel couplings on suction and discharge nozzle pressure taps in accordance with:
- a. CE Specification for Standard Plant for Reactor Coolant
' . Punps No. SYS80-PE-480,' Rev. 02. .
~
.. b. CE Project Specification for Reaptor Coolant Pumps No.
14273-PE-480, Rev. 03. ,
- c. ; - ASME B&PV Code, Section III, Subsection NB,1974 Edition.
- d. CE-KSB ' Instruction Manual No. 8109-101-500, Rev. O.
e.' CE Avery Quality Assurance Field Manual.
- f. CE-KSB drawing C-8000-101-2017. Rev. 01, Wall Static
. Pressure Nozzle - Suction Safe End,
- g. CE-KSB drawing C-8090-101-2018, Rev. 01, Wall. Static Pressure Nozzle - Discharge Safe End.
,,yo,c , ,,oo ,,, . co--ooin caos
_ tax __ .suvoice hicisTaATsow wo.
- TAE raen. men. tass ous ADalt ONAL asef 0AM Aisses.
Vandor: CE-Avery , p ~ w W 5269 o4:. fyg mE .su p. m e 1 E O f pe L a Au t CO84T R ACT *e A se L , ANPP Unit f2 _
/4.a;
*.w CMAas5E Tes 6- E CT . **
- O T M L A A wT s=O mat . (Cl C M88 % e4 &C% ES18Ma T E V a LWE AbpF
- E O o. t enee ress4. DAT L . sasPv s Y &. G w AbsT Y A n t w a6 A *=I E
- I CONT. AD'd88e. w . .. hW.:Ca. $, 0 ,.I *fw ;
$*L - " Tcl'203/6BB 1911! ' h ) jR Cd Pskt Systsms(
' Comousipn i' Enginjenng in: Tetra 99297 . -
.1000 Prsspect Hill Reic
+
y Wenesst Connecucut 06095 - .Attechnent to DER 82-42~ ! Mr"g"; POWER - . mg 'i q;ms SYSTEMS ..
~ August 12. 1982 " -b'
,r. V-CE-16870 E 6'M ' ' ~
, , , p ... , d 0
- Mr. G. C. Andognini -
T, ,, .,".1"5
; Arizona Public Service. Company. i m P. 0. Box . 21666. -- Sta. '4015 ' L*lJ Phoenix, Arizona 85036 -
g m
Subject:
Arizona' Nuclear . Power Project " Units 1 and 2 Reactor Coolant = i s. Pump Pressure Nozzle ' Sensitization
.- . . . . . . . . . - - . . . -. . "f
Reference:
'(A)- C-E Letter V-CE-16760 dated July 19, 1982 Subject -~ Unit 1 Reactor Coolant Pump 2B et .3 *=
Pressure Nozzle Leak 10CFR50.55(e) Deportability 1 __ Y.'w".
Enclosures:
(1) C-E-KSB Sketch SC-8000-0098 Rev. 00, Alteration r,c_ wie of Discharge Nozzle Pressure Tap
-(2) C-E-KSB Sketch SC.-8000-0099 Rev. 00, Alteration t%.,g &g , . of Suction Nozzle Pressure Tap 3 %
Dear Mr. Andognini:
Enclosure (1) to ' Reference (A) described in general terms the corrective action-necessary to; remove the sensitized 316 stainless steel RCP pressure taps and to address the sensitization of the inconel Heat Affected Zone (HAZ). The purpose'
'of this. letter is to delineate this action in detail.
PVNGS #1
. A. Discharoe Pressure Tao 2B -
- 1. Perform thorough visual examination of pressure. tap from inside the primary pipe prior to fuei load and report results. '
- 2. Perform dye penetrant examination of J-weld inside primary pipe to include the Heat Affected Zone (HAZ) at the same time.
B. .Discharae Pressure Taos 1A, IB and 2A f
"n 1. Remove the 316 stainless steel portion of the nozzle, the weld and as much ,
as 1/4" of the inconel portion as possible. Return the removed portion to C-E since it is intended that the HAZ of the inconel be inspected at C-E Windsor. However, there is concern that access to the replacement fillet weld shown on Enclosure (1) might be overly restrictive. An optimum cutting elevation should be selected and its affect on the overall length of the replacement nozzle evaluated versus the field connected pressure l pioing. For Unit 1, replacement pressure nozzles have been fabricated to the Enclosure (1) dimensions. 1
' l t'
\
-Mr. G. C. Andognini ' -
A August 12. 1982 Pdge 2 V-CE-16B70
)
~.
- 2. Conduct a hydro on the completed field welds:"
C. All Suetien Taps
- 1. Enclosure (2) depicts the removal of the stainless steel portion at about -
the centerline of the weld. This location of cut was assumed to be ,, optimum for two reasons: ,, __
- a. There would be sufficient access to make the fillet weld.
- b. Dressing the stub OD so that it fit into the 1.0001.p0 soc'ket W u'id be a reasonable task. -
If dressing the OD of the shop stainless steel-inconel weld is easier than anticipated, the location of cut could be raised slightly. For Unit 1, i' there replacement suction pressure nozzles have been fabricated to the Enclosure (2) dimensions.
- 2. Hy'dro test the ' completed field welds.
PVNGS #2 A. All Discharoe Pressure Taos ( 1. Same steps as B above including returning th'e cut-off nozzles to C-E Windsor. B. Al[ Suction Pressure Taos
- 1. Same steps as C above.
The replacement pressure nozzles have not been fabricated for Unit 2. There are four '(4) similiarly constructed pressure nozzles in each of the hot legs near the steam generators. They are approximately 44 inches from the centerline of the steam generator to hot leg field welds. We feel it is prudent for you to , perform a re-review of the post weld. heat treat procedure for these welds on Units 1 and 2, to determine if these' pressure nozzles have been subjected to sensitization temperatures during post weld heat treatment. Please infonn C-E of the results of this review. . C-E has been proceeding in the interest of maintaining safety and schedule to provide a resolution to this problem. We must, at this time, request that a , Field Material Requisition be issued to cover our expenditure of man-hours and material costs. Your attention to this matter is requested. . Y 1 ours,
.- s C. Ferg .
Proje & 'hnager CF/ DEW /VSK/eh r ,
.- .- Mr. G. C.'Aniognini Page 3 .
Aucust 12, 1982
. V-CE-16870
(- . .. . . cc: Messrs: E. E. Ypn Brunt, Jr. - w/e J. Yorees - w/e .
. W. H. Wilson _
W. G. Bingham - w/e - R. H. Holm . J. W. Dilk G. A. Butterworth S. N. Mager . ., D. B. Amerine - w/e ' W. L. MacDonald - - J. M. Allen - W/e O e ( . .
** a e see- *oe .e . em , .
e
- 9 0
9 e _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ ________________________^_____________________________..______________.___________________
)1'!!
,j
])m - n . D' r.
.s W ,. e" r
o'
.i
~
_6 V L N
- o. *** :. t:
. /!= p. 5 c
v I
%s' tW o ~
7
, O ,c a. =
d a* Y . , ,
,os
. ~?
4 _ ~ o' R N ,e
,o s
K .- c" A0 ,e mA-
-,sm,w n
m
-- sd . s- eac
- vo 1
w N T Ws.
. 1
_r .:- o . g
.6 -
u c'- " I M M. A , e nPSc M i ! r. . o l R
) .. wY"-
I = l . O . 0 Ea1 s. _.ls f R E F RN ~12 n 7* w t
. _m*S'-_. e nO-o P
Q I R F t afO e/ W m_N- .
? .
e
,_ * . F sG l
wC p*- s.
~ xP .
cN
- ,_ 1" .. .
-Y
- ~
. c7 Y .~
- :.~
c
~,9- .
- A - i
, .o ,' . -
. t -
- e
-(
~
' - - r .
~
_ .N e V
. g <
r . - 7 'N a ~ N N\ a
~
..gs. ,.
N
\\ . .. .
\N
- x N ! i , . . ;y .
. . l' M
3 [
\s b ., .
Y ._ _N N,
\N N N
N N N 2 K N P m-
\N N .=U+r.
U ' p. N N
. 7- *
\x M Mn k= N N N M.
m~. N 1 x g . N N mta ud
=
j N n.Y ~ N f wKC r,ag' v b /o E g ,. .
. N cP 4
&bN m*.AN
,._dyor XC
. y g
,, ;_ , r .
t . 4, 7 . u
, / .
t A .
. gv VE . .
/ s . .
/
s +^ a e -
, / rE
/ i N* _ . yc n e-u r x y -
u C z [ A e t -
- ' o , .
c y~n $/ e
~ us' m s n
.-r
/a ~
o y,t; -y
/
.rM y/
aJ. . u. t s .- ' .. u _. "'
/ e N,
/ , _
- ~
a . s , c/ r l
- v l_ -
2 t r F
. u .
C - .
. o r.
. , \'
- i. . . . .!
- t. . .>. .t ..,. k;[
.I [,
i :i n ' je 4
. .g , ,: g3 , ' ' *e z$"
{1i\i!i ! . i
m x- .. a.o sw, u s p o B cms.: - O wc o w t e ue wao W C-Y O N u5 .
,v e
c n D ( m e.n e pusw 7 o .cor eg4 ne o
.S .
R T A AM I q; 4 N
~m r y en.e p
I R 3 MOF
. e -
. t oss
., t,ce tmmccem t
a S e, s . UN E I . ee
- r. f
.%z 'a.
i
.~:
. .de/
._f rye e
- t. p y e.axc a
uo ' R R P OF
,.~
~n .
'/ ,
/a nh.
/ _
4 ,
. s% ,. .
/
/ . X _. $_ S f_"
** ' : m, s'!
/
/
- t
. t x <, .
x.// / / , 0 E
~ /)
o 0
~
/
0 I m"* 17 - 4 1 N r/ 1 4:. 0
/ 0
/ ,
/ ae 0
0
.W,,I
, ,,, . A. .
-( V ,
..s,
,~4,
//
//
7 . ,/
. />
j.. P c
/n x
g,w.m
~
.K yn . ~
~- ~ g N W _i nu e m w'
_*np r, . a
.m ,
%y Gc_./ ,' ur z 7 c '
sa r w/ +E7m/4,R g u g o s . T m 72w ,e s R P
~
4m ./
' (
m - .
- ...e 4 _ < . : ! .:
-e HB TECHNICAL CONSULTANTS ON l
Date G. Bricenbaugh 1723 Harntiton Avenue-Sutte K \ Richard B. Hubbard San Jose. California 95125 Gregory C. u,nor 4 May 1987 m ne m 26S2716 i Mr. Donnie H. Grimsley, Director FREEDOM 0F INFORMAT10N Division of Rules and Records ACT REQUEST Office of Administration g_f p M/ U.S. Nuclear Regulatory Commission Washington, D.C. 20555 6 (( ;
Dear Mr. Grimsley:
i Pursuant to the Freedom of Information Act, please make available at the Commission's Washington, D.C., Public Document Room single copies of records as follows: A. All documents discussing investigations conducted by or for the Arizona Nuclear Power Project (ANPP) or the NRC regarding the identification and resolution of problems with the Low Pressure Safety Injection (LPSI) ; pumps at Palo Verde Units 1, 2, and/or 3. B. All documents discussing investigations conducted by or for ANPP or the NRC regarding the identification'and resolution of problems with the Reactor Coolant System (RCS) pumps at Palo Verde Units 1, 2, and/or 3. These should include inspection reports, other technical reports, memos, ANPP submittals, letters and other correspondence between ANPP and the NRC, and other documentation, regardless of origin, regarding these issues. . l If there are any questions concerning this request, please contact me directly at (408) 266-2716. Your prompt response to this request will be appreciated. Sinc se ,
/
M r g$ory 7 A. Holmes Associate Consultant \ - l L.L'.L
< w W in :y M Qp i 7f .)f;).
_-}}