ML20238C133
| ML20238C133 | |
| Person / Time | |
|---|---|
| Site: | Nine Mile Point  |
| Issue date: | 10/27/1986 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20238A944 | List:
|
| References | |
| FOIA-87-438 NUDOCS 8709090670 | |
| Download: ML20238C133 (7) | |
Text
- _ _ _ _ - _ - _ _ _ _ _ . _ _ .
1
- I y[p *scwog UNITED STATES y cc g NUCLEAR REGULATORY COMMISSION h j wAssiNcTow. o. c. 20sss st ,% a
'%d l
SAFETY EVALUATION BY THE OFFICE OF NUCLEAR REACTOR REGULATION CONCERNING MAIN STEAM ISOLATION VALVE LEAKAGE PROBLEM NIAGARA MOHAWK POWER CORPORATION NINE MILE POINT UNIT 2 DOCKET NO.: 50-410
1.0 INTRODUCTION
C Two problems were rece,'tly uncovered concerning the Main Steam isolation Valves (MSIV). These problems can be categorized into defective actuators which affected valve closure time and latch disengagement, and leakage rates which exceeded the 6 SCFH Technical Specification limit.
The excess leakage was attributed by the applicant to failure of the tungsten carbide coating on the ball due to high localized stresses, and subsequent scratching of the stellite seat surfaces by the released tungsten carbide particles.
The resolution of the actuator problem will be addressed in an SER to be issued prior to criticality.
The MSIV laakage concern and its interim resolution will be covered in this SER. The first issue to be addressed is the use of one MSIV in each main steam line, with a modified seat spring configuration and recoated tungsten carbide ball, to ensure secondary containment integrity to permit fuel loading and pre-criticality startup testing. The MSIVs will be deactivated in the closed position and " Type C" leak tested to verify leak tight integrity.
The second issue is the appropriateness of the use of all eight MSIVs, with modified seat spring configuration and recoated tungsten carbide balls, to ensure primary containment integrity and pemit criticality and power operation through initial fuel cycle. The applicant has provided information in letters dated October 8,10,17 and 20,1986 concerning the MSIV leakage problem, its resolution and a commitment to perform a modification verification test program during the first fuel cycle. As part of this program, the applicant has comitted to perform a prototype test program to confim the operability of the modified Nine Mile Point Unit 2 MSivr. The tests will be conducted during the first few k months of plant operattTon. The prototype valve configuration will duplicate the valve and actuator design used in the plant. The test objectives include:
V O Verification of the mechanical integrity of the value ,
and actuator for the expected operating and test cycles.
8709090670 070028 PDR F O.1 A WETTERH87-438 PDR
' ?.,
- \
b!' c t ,
l
, l 7
>- 0 DeraonstFdian of' valve leak tigStr;ess; for tha. expected valve osty e M es.
b .
9 Demnstraticn of the ability to close the vahe within le Technical Specification limits under normal operating j pressure and temperatu:-e steam conditions.
' , 0 Verification of the conservatism of the between-the-seat leak test twthoc' as an alt.ernative to across-thuvalve siat lokage tests.
Other objudvet af the prototype test. program are to provide l baseline (.sta for evaluation of (1) the long terr. suitability of the valve Tnd-(2) potentiti design and matairial improvements.
2.0 EvitVAT10N 5 In March, 1986 all eight MSIVs were ". Type C" Local Leak Hata Tested (LLFT) in accordance with tne requirements of Appendix J of 10 CFR 50.
The results of cests were acceptable. In August,1986 these : valves were each cycled in excess of 100 timesp (75 valve cycles are esticated (uring &
the first fuel cycle) in an attempt to correct actuator problems. .,
Subsequent w these operations, leak testing in September, 1986 indicated c.
that all ek.n MSIVs had leakage rates ateove the limiting valyk of 6 SCFH A in the Tecnnical Specification. vat tig to '
r Tne applicant in their submittal of October M,1956 and during the A meeting with the staff on October 15, 1986, provided their evaluation *
' ~
describing tha conditions that led to degradation of the tungsten carbike costing on the ball arid to the seat uhtch forms tie seal.' ' the applicant ,
p has pro 7tded the results of mechanical aralyses showing that when tne i
origi[nally cduse the test configured valve is in to " rock", resulting opened from high local f ts closed stresses on the position, ball. the t loa From the results of site frictional tests combined with analytical reSuits !
' l stated above, 'the applicant has hypothesized the degradation occurt as follows:
4
- 1) High local contact t, tresses khen " rocking" occurs caese the softer stellite 6 seat material to be # '
s smeared onJo the harde/ tungsten carbide ball surface whenahe valve was cycled open.
- 2) When the steihte 6 material is present both on the ball surface and the seat, the high local contact stresses muse localitad c'oldielding i.
of portions of the seat to the ball surface.
- 3) subsequent cycling of the valve csuses local delamination of portions of the tungsten cerbide coating when the local cold welded creas are pulled apart. This premise is supported by the otsservations of tne ball and simulated seat surfaces in the site friction test performed at about 55 Ksi ball pressure.
Pock marks were observed on both the seat and bdl after 30 strokes of tSc valve daring the test.
1
l s
- 4) The louse hard tungsten carbide flakes removed from the ball surf ace cut the softer seat surface wher. the valve is cycled, causing the seal between the seat and ball surfaces to be degraded.
The staff concurs that the above stated failure mechanism appears plausible from the information obtained from the analytical and t test data. Most assuredly the data supports the conclusion that the tungsten carbide coating degradation is the result of a faulty valve design, rather than related to a material problem with the coating itself.
To correct this design deficiency, the applicant decided to rearrenge %
the seat springs.7n this regard, four 115 springs were removed from the left side of the seal ring (facing the ball) and eight 57.5 pound springs were added on the right s16 of the seal ring to determine if " rocking" would be eliminated by the revised spring pack. The results of the analysis indicated that the modification should be effective in eliminating j the " rocking" ph,enomenon. If the seat does not rock, the contact area of the seat on thetwall is much larger. The applicant states that the results k of their calculations are that an order of magnitude reduction of stress is realized from the modification, and that the calculated bearing stress is reduced to approximately 2700 psi.
The applicant has performed two tests to date for assessing the performance of the modified Sall valve. One test consisted of stroking N e valve with a new ball and modified spring pack 75 times. This was followed by the leak checking and subsequently examining the ball and seat surfaces for degradation. The applicant reported that the ball surface showed some evidence of wear similar to that observed in other tests where the contact pressure was low. No other degradation of the ball surface was reported. Inspection results reported for the seats are
, that the surface finish meets requirements and only some slight scratche/NG * -
and polishing were observed. The applicant reported the modified valve passed the leakage test following cycling 75 times. The results of l simulated Mock Friction Tests f r=hs in which a simulated stellite 6 y seat was stroked across a tungsten carbide ball at a contact pressure of 3500 psi were reported to be similar t6 test results for the modified 4 valve. essentially the same wear observations were presented for 75 I
cycles for the ball and simulated seat in the Mock friction Test.
Based on the analytical and test results presented by the applicant, the staff concludes there is reasonable assurance the modification will be effective in reducing degradation of the ball / seat. However, the tests were performed on a new ball. The actual balls which will be used by NMP-2 will be recoated. To ass g the difference between a recoated vs.
a new ball, one must consider tMs fabrication process. The balls for r the MSIV valves were initially fabricated as follows:
- 1) The balls were cast using a 316 type austenitic stainless material. '
The castings were extensively weld repaired to remove casting g d[ continuities (this is not unusual for this type casting). The ,
casting then received a solution anneal heat treatment following weld repair.
( .
?
L I 2) A Haynes 25 weld overlay was applied. The final coating L
thickness was between 0.080 and 0.100 inches. The Haynes 25 coating was ground and then liquid penetrant tested (PT).
, 3) The Haynes 25 coating was grit blasted as a surface preparation ,
prior to coating with a material rich in tungsten carbide. A j minimum 0.010 inch tungsten carbide coating was applied by Union !
Carbide Corporation using a detonation gun process. This process ,
was used to obtain a low porosity coating with a minimum of distortion of the ball. The ball coating was ground and subsequently lapped to the seat.
g.d Since the ball surface was damaged bykstroking in the originally configured valves, it was necessary toVref&hh their surfaces. The refurbishment consisted of mechanically removing the tungsten carbide conting, re-applying it and re-finishing the ball surface. Upon removal of the tungsten carbide coating, the Haynes 25 surface was PT tested and any unacceptable discontinuities were removed. In some cases the excavation went into the costing material. The excavations were PT examined to assure no unacceptable discontinuities were present. The casting and Haynes 25 coating were locally weld repaired as necessary.
Final Haynes 25 coating thicknesses ranged between 0.076 to 0.097 inches. The Haynes 25 coating was ground and PT tested. Application of the tungsten carbide rich coating and surface finishing was as previously described for the original ball.
The only difference between a new and refurbished ball is that a solution anneal heat treatment was applied to the ball following weld repair of the casting. The applicant stated the repairs to the casting as part of the refurbishment were minor. Considering the weld repairs to the ball were minor, repaired the areas ballclad were itselfwith is aHaynes casting25(with a high which ferrite is highly number, and corrosion resistant), we consider the likelihood of susceptibility to stress corrosion has been minimized to an acceptable level without a second solution anneal heat treatment of the ball. Since the refurbished ball has been processed in a manner essentially the same as the originalf with +
the exception described above, we conclude the refurbished ball is equivalent to a new ball and, therefore, the 75 cycle test of that performed with a new ball is appropriate for a refurbished ball.
With respect to the use of one MSIV in each main steamline to provide secondary containment integrity prior to criticality, all four MSIVs will have the modified seat spring configuration and recoated tungsten carbide balls. However, the actuator modification will not be completed. Each valve will be cycled a few times to assure proper seating prior to a
" Type C" LLRT. The between the seats MSIV leak rate must not exceed the limiting te ical specification value of 6 SCFH to be considered acceptable.
acIMp^ The actuat will be deactivated in the closed position by opening the '
circuit breakers. The MSIVs will, therefore, be deactivated in the closed position during fuel load and startup testing. Since the MSIVs will meet " Type C" leak rate TecMical Specification limits and the valves will not be operated after the leak testing, we find it acceptable to use the modified MSIVs for secondary containment integrit Operational Conditi~ o ns 4 and 5 (Cold Shutdown and Refueling)y durin
4 i
I
.' The second issue pertains-to the use of the refurbished and modified inboard and outboard MSIVs for their intended function of providing primary containment isolation capability through the initial fuel load.
While four MSIVs provide secondary containment integrity during fuel ;
load and startup testing, the four remaining MSIVs (with modified i actuators, recoated tungsten carbide balls and modified seat spring ,
configuration) will be installed in the main steam lines. These four i valves will then be placed in an operational status after successful l completion of " Type C" le6k tests with leakage not to exceed 6 SCFH and j valve closure times of less than 5 seconds required by the Technical Spi.cifications. The four valves used to provide secondary containment integrity will then be upgraded in a similar fashion to operational j status. .
I The modified seat spring configuration was tested in conjunction wf th a l new tungsten carbide ball for 75 valve cycles, which is an upper bound estimate of the number of cycles to the first refueling outage. After 75 valve cycles, " Type C" leak testin,g indicated acceptable sealing WJ 'I integrity of less than 6 SCFH. Pfsttestint.pectionalsoindicatedno y evidence of ball coating degradation. Based on these results, we conclude there is reasonable assurance tb3- the MSIVs with the modified seat spring configuration and recoated tungsten carbide balls, will provide acceptable leak tight performance until the first refueling.
The staff also concludes that the tests performed to date indicate the root causes of the high leakage has largely been corrected. However, it must be noted that these same tests were not totally protypical of in-plant conditip&y a more n d _ ause base,ofthe thesestaffdifferences concludes and that recognition confirmatoryofinformation the need for Y' is needed to verify continued acceptable MSIV performance during the first fuel cycle. This needed confirmatory information will be obtained from the MSIV prototype test program and during Nine Mile Point Unit 2 operation during the first fuel cycle. Matters to be addressed are:
1 - Will the modified actuator affect the perfomance of the refurbished staling surface? The successfully tested MSIV leakage corrective action has not presently been tested in combination with the modified actuator. This is an objective of the prototype test program.
2 - Will modified MSIV rformance be affected by steam flow and k operating temperature, pressure and steam conditions during valve closure? Tests to date have been at ambient temperatures with no steam flow. The prototype test will address this concern.
Also the in-plant MSIVs will be leak tested after the MSIV full isolation test, in addition to the mid cycle shutdown to provide in-plant performance data after MSIV exposure to operating temperatures, pressures and steam flow.
3 - Will the alternate between-the-seat leak test method be an acceptable test alternative to the conventional across-the-valve seat leak test after many cycles of valve operation? The actuator loads on the sealing surfaces may change with seat wear and affect the leak test results as valve closure cycles accumulate during )
plant operation. Verification of the between-the-seat leak test
I'
- l. .
1o.
l l
o method as an alternate to across-the-valve seat test will be an objective of the prototype test.
Due to the importance of the above test data, the continued operability of the Nine Mile Point Unit 2 MSIVs will be conditional on the successful performance of a concurrent prototype test program and in-plant leak tests 50cn after the MSIV full isolation test.
CONCLUSION On the basis of testing to date and contingent on prototype test performance we conclude that the refurbished MSIVs are acceptable for.
plant use up to the first refueling. We will evaluate the prototype test results in tems of demonstrating long term operability prior to .
making a determination on continued use of the refurbished MSIVs beyond 1 the first fuel reload.
Specifically, we find that:
~
1 - The four MSIVs (one in each main steam line), with codified seat spring configuration and recoated tungsten carbide balls, are acceptable for use to support secondary containment integrity for initial fuel load and pre-criticality testing.
2 - The inboard and outboard MSIVs', with modified actuators and refurbished seat surfaces, are acceptable for providing primary containment isolation capability for the first fuel load conrnencing with initial criticality. This conclusion is contingent upon verification of acceptable MSIV perfomance by a special in-plant test in Nine Mile Point Unit 2 and successful completion of a prototype testing program during the first fuel cycle to assure the achievement of the testing goals.
The licensee should be conditioned as follows:
Nine Mile Point Unit 2 MSIVs: g 7 ggf p3, Aprp p 4<
A. Confirmatory leak tests sin'g acceptance criteria that are the same as outlined within JAS T/S Ai?. J. LLRT program will be perfomed no later than the first outage (but within 30 days of) the 100 hour0.00116 days <br />0.0278 hours <br />1.653439e-4 weeks <br />3.805e-5 months <br /> warranty run, whichever is earlier. This test is '
not intended to replace any APP, J. test requirement.
PtrEHP w B. The prototype test program that will test the modified valve and actuator configuration similar to the ones installed in Nine Mile Point Unit 2 will include:
o Verification of the mechanical integrity of the valve and actuator for the expected operating and test cycles, o Demonstration of valve leak tightness for the expected valve duty cycles, o Demonstration of the ability to close the valve within i Technical Specification limits under normal operating pressure and temperature steam conditions.
I 1
)
- 1. '
v...
l l; , ..
- E ' o Verification of the conservatism of the between-the-seat leak' test method as an alternative.to dcross-the-valve -
seat leakage tests.
l- o Other test objectives are to provide baseline data' for evaluation of (1) the long term suitability of the valve and (2) potential. design and material improvements.
i~ >
C. The prototype test report which will 6ddress the confirmation of the valves' acceptability for the first operating cycle shall be provided to NRC by May 15,1987. This testing is scheduled to be completed by April 1, 1987.
l l
l 1
- - _ _ _ _ _ _ - _ _ _ _ _ _ _ _ _ _