ML19225B741
| ML19225B741 | |
| Person / Time | |
|---|---|
| Issue date: | 06/12/1979 |
| From: | Oreilly P Office of Nuclear Reactor Regulation |
| To: | Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 7907250604 | |
| Download: ML19225B741 (6) | |
Text
.
DISTRIBUTION LIST Westingnou;e & Combustion Engineering Coro. - TMI Related Cer:. ;:endence T. Ncvak, DSS P-ll32 '.
C. Heltemes, DPM 242 R. Reid, ORB 330 A. Schwencer, 00R 316 Z. Rosztoczy, DSS P-1030 D. Ziemann,00R 314 R. Audette, DSS P-1030 T. H. Cox, DPM 242 R. Frahm. DSS P-1132 W. F. Xane, DPM 242 P. Norian, DSS P-1030 M. Cunningham, RES MNBB 3103 A. Thadoni, DSS P-il32 H. Gearin, DPM 256 C. Graves, DSS P-ll32 R. Bursey, DPM 357 S. Newberry, DSS P-ll32 J. Shapaker, DSS P-904 G. Xelly, DSS P-ll32 NRC PDR I 3
R. Scholl, DSS P-722 NRR Reading File y q
p @3o P. Matthews, DSS P-822 L. Riani, DSS P-222 SB Reading File J. Wermiel, DSS P-SC2 Meeting Attendees J. Calvo, DSS P-822 Central File M. Taylor, RES MNSB 3103 N. Villaiva, DPM 242 P. O'Reilly, DPM 242 R. Tedesco, 055 P-il32
- 5. Varga, DPM 116 F. Williams, DPM 110 R. Woodruff, ISE E/W 359 N. Wa gne r, CSS P-il32 T. Wamcach, CCR 314 R..Mattsen, CSS P-llc 0 S. Hanauer, CSS P-822 D. Ross, DPM 265 V.
Stello, CCR 528 E. Case, NRR P-404 K. Mahan, CPM 357
.rt 9 DTBWs4N
]90] 25()T C N tD SW/
C
/
UNITED STATES
- ' $ w h
NUCLEAR REGULATORY COMMISSION
-Y5. 74[ E WASHINGTON, D. C. 20555 l
\\ ((.
- JUN 121979 MEMORANCUM FOR:
Distribution FRCM:
Patrick D. O'Reilly, Project Manager Standardization Branch, DPM
SUBJECT:
SUMMARY
OF MAY 9, 1979 MEETING WITH WESTINGHCUSE ELECTRIC CORPCRATION REGARDING RECCMMENDED RESPONSE TO ITEM 7(c) 0F I&E BULLETINS 79-C6A & 79-06A, REVISION 1 On May 9,1979, representatives of the Westinghouse Electric Corpcration (W) and 1 censees with W-designed operating PWR's met with the.'niC staff at NRC Headqt'arters in Bethesda, Maryland. TF meeting was brld at the staff's reque.t to discuss W's recc::rendations so its custcmer -licensees regarding the response to IIem 7(c) of I&E Bulletins 78-C6A and 79-06A, Revision 1.
A list of attendees is contained in the enclosure.
In its intrrductory remarks, the staff stated that Item 7(c) of I&E Bulletins73-06A and 79-06A, Revisicn 1, requires that the reactor coolant pumps be kept operating in the evert of high pressure injection initiation when the reactor coolant pumps are running. The staff further stated that it is aware that the containment isolation signal (in a number of operating plants, this occurs upon initiation of high pressure injection) results in the isolation of certain auxiliary systems for the reactor coolant pumps, such as the ccmponent cooling water system. The staff then summarized the role of the task group reviewing 'd and Ccmbustion Engineering-designed Ocerating plants in the light of the Three Mile (TMI-2) accident.
The staf' expressed its belief that W had been in general agreement with the I&E Sulletin 79-06A and 79-CEA, Reviiicn 1, requirements concerning reactor coolant pump cperation during an accident. Several discussions wi:n i representatives during tne preparation of I&E Bulletin 79-C6A were cited as basis for this belief.
W replied that it had never indicated agreement wi:h the requi ements on reactor coolan pump 0;:eration of Item 7(c) c' I&E Sulletin 79-C6A.
In fact, W clai: ed that its disagreement with :ne staff's ;csi-icn had been ccnveyed Txplicitly to tne staff. W then explained.iny it had recommended to its
~
customers that the reactor ccolant ;: umps be trip;ed in cases sucn as that specified by Item 7(c) before the D1I-E accident and wny its posi:f cn en this matter has not changed since TMI-2 50(o4 452 286
JUN 121979 Memo for Distribution W gave the.following reasons for its pre-TMI-2 position that, given the conditions stated in Item 7(c) of I&E Bulletin 79-CEA, all reactor 3colant pumps should be tripped:
(1) The bounding LOCA analysis presented in SAR's assumes a loss of power at the start of the calculation.
(2) Other analyses in Chapter 15 of the SAR assume either that a loss of power occurs (and, therefore the pumps trip immediately) or that the pumps centinue running, whichever is the more conservative assumption.
( 3') Scme transients result in loss of power to the pumps.
(4) Damage to the pumps must be avoided.
(5) Certain conditions of containment isolation isolate tne ccmponent cooling water system.
W then presented a description of its 93A type reactor coolant pump, which is typical of the pumps installed at operating W-designed PWR's.
With the use of a large diagram of the 93A type pump, the following components were located: driver head, three pump seals, pump bearing, thermal barrier, and the impeller, which is located inside the pump casing.
The locations of the component cooling water inlet and cutlet were pointed out. The main bearings of the pump are cooled by the ccmponent cooling water which dissipates heat tnrough the thermal barrier heat exchanger. The pump seals are cooled by the scal injection water flow of approximately 8 gallcns per minute.
W described the function of the seal injection system in detail. Seal injection
'Jater feeds seal #1. Seal #1 is a so-called film seal. The seal injection water flows througn seal #1 (rne pressure difference across seal #1 is approx-imately 2200 psi) and feeds seal #2. Seal #2 is a rubbing type seal.
It operates at abcut 20 psi (volume contra' :ank pressure). Seal #3, which is the same design as seal #2, operates at aporcx:mately 3 psi. Seal injection water is introduced below the pump bearing at a rate of about 50 gallons per dnute.
This water is clean, filtered water provided by the charging oumps.
If the seal injection system should be lost, cooling water for the pumc seals is taken frca the primary system and directed past cooling coils.
This water is not filtered.
Before a reactor coolant pump is scarted, a laak rate is established nrcuen seal
- 1 witn a J.P across tne seal. Otherwise, tne seals wculd ccme into contaci.
Seal 12 serves as a backup to seal il and is capa:le of withstanding the primary system pressure.
The function and operation Of the ccmcon<.nt ccoling water systn were then discussed by W.
In addition to its primary unction, ccmcenent cooling nater also orovides cackup for seal injecticn wc.ter.
Therefore, if the ccm::anen:
NOR DiEnM 452 289
Memo for Distribution JUM 121979 coeling water system were lost, there would still be water available (although unfiltered) for seal injection. However, since the component cooling water also cools the main bearings, the bearings would heat up to alarm-temperatures in about 10-12 minutes. There is no pump trip provided upon loss of component cooling water.
The discussions which followed brought out a number of noteworthy points:
(1)
M seal injection is lost, the leak rate past the seal does not increase ovec three gallons per minute since primary system cooling water is still available for this function.
(2) A reactor coolant pump can operate with the seal return line shut off.
(3) Reactor coolant pump operation could continue under phase 3 containment isolation with the leakoff line closed. However, there might be a problem with the pump seals.
(4} Reactor coolant pump operation could continue under phase 3 containment isolation with the component cooling water system isolated until the main pump bearing burned out.
(5) Normally, failure of seal #1 would result in a leak rate of about 50-60 gallons per minute. This compares with an estimated leak flow of 400 gallons per minute (based on the amount of water in the containment) for the pump seal failures which occurred at the H. 3. Robinson plant.
W requested that the staff explain the basis for its Item 7(c) requirement That the reactor coolant pumps be kept running.
The staff replied that the purpose is to maintain forced flow in the primary system to enhance core cooling. The staff noted that a modification to this requirement was under consideration. This would require keeping the reactor coolant pumps running unless it leads to an unsafe plant condition.
W then explained that, after the TMI-2 accident, the position on post-accident operztion of the reactor ccolant pumps was reevaluated. 3ased on this reevalu-ation, W reccmends to its custcmers with c::erating plants that the reactor coolant pumos be snut off during an accident if either of the following conditions are satisfied:
(1) After it has been verified tr.
safety injection sa:er is cooling the core, or (2)
Five minutes after the loss of ccmpcnen: ccoling nater.
P90D cnprum
- 9 ~
i mr 452 290
Memo for Distribution JUN 121c79 W cited the following reasons for this position:
(1) The FSAR safety analyses are still valid.
(2)
In the opinion of W, the loss of core cooling at TMI-2 was not attributable to shutting off the reactor coolant pumps. There fo re,
the focus on the operation of these pumps as a cause of core damage is not appropriate.
(3) W is ccaff dent that, with the configuration of its pWR design and the U-tube steam generator design, natural circulation will be established when needed to prove a means of core cooling.
(4) W questioned how the issue of which two pumps in a four-loop plant would be kept operating should be addressed.
(5) The philosophy of keeping reactor coolant pumps operating is an as-yet unanalyzed situation.
(6) Pr blems have been encountered in the attempt to explain the shift in thermocouple readings at TMI-2 which were observed when one reactor coolant pump was shut off and another st:eted up.
Frcm the discussion that followed this final W presentation, the folicwing points were significant:
(1) Neither phase A (isolates seal return) nor phase B (isolates ccmpenent cooling water) containment isolation trips the reactor coolant pumps.
(2) The staff noted that 'd needs to consider the question why ccaponent cooling water has to be isolated.
(3)
W explained tnat the 1550 psi criterien for high pressure sa fety injection shutoff contained in reccmmendations to its customers with operating plants assures an adecuate margin of subccoling.
(4) W mentioned that leaving tne reactor coolant pumos running could aggrevate Ccnditicn II events. Altncugn it would help in same transients, it iignt hurt in others.
(5) f remarked that the staff shculd either require shutting off all pumps or else keep all tne pumps running.
The cicsing discussion centered an the role of positive dispiacemen c: a rgi n g pumcs in tafety analyses.
In res::ense to the staff's inquiry :c this end, y replied that, si;ce positive displacemen charging pum::s are n00 sa fety-grace and do not have a safety function, no credi is ta.<en for these pumcs in safety analyses.
QMu(c. D D '~?)
M. t r A4 M 4 patrick O. 0 9eilij, Project Mana ger Standardization 5rancn Division of ?roject Managemen h
3 U U d) 1)l} Ml$[
Enclosure :
As stated 452 Ni
NRC MEETING WITH WESTINGHCUSE - MAY 09, 1979 LIST OF ATTENDEES Name Affiliation P. D. O'Reilly NRC/DPM T. V. Wambach NRC/ DOR R. W. Woodruff NRC/IE T. G. Satryan Westinghouse K. R. Jordan Westinghouse N. H. Wagner NRC/ DSS C. L. Gottshall Westinghouse A. J. Weisbard Shaw Pittman O. C. Richardson Westinghouse J. M. Kovacs NRC/ DSS H. L. Srammer NRC/ DSS F. C. Cherny NRC/ DSS X. R. Mahan NRC/DFM F. Williams NRC/DFM S. Varga NRC/ CPM A. Thadani NRC/ DSS E. Reeves NRC/DCR George Liebler Florida Pwr & Lgt Co G. D. Whittier Florida ?wr i L;t Co R. J. Acosta Florida ;wr i Lg: Cc Ste:: hen T. Ha r; Florida Pwr 7 Lg. C:
Canald M. Rcsh Ocn Ed af N.Y.
Caniel Call Westingncuse John Gilmcre Commenweal:n Ediscn James Deress Commonwealth Edison J. Sha:aker NRC/ CSS W. Butler NRC/ DSS J. Scin:o
'IRC/CELD M. E. Stem Wisconsin Puolic Service Clark Steinnardt Wisconsin Publi S e rv i '. e 452 392