ML19220C494
| ML19220C494 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 04/12/1979 |
| From: | Ross D Office of Nuclear Reactor Regulation |
| To: | Case E Office of Nuclear Reactor Regulation |
| References | |
| NUDOCS 7905100384 | |
| Download: ML19220C494 (6) | |
Text
.,
}
/ - /
g A
f N
UNITED STATES
[.4, j, 'l NUCLEAR REGULATORY CCM.'.11Ss!CN g[
'NASHINGTCN, D. C. 20555
./ /
APRIL 1 2 1973 MEMCRANDUM FCR:
E. G. Case, Deputy Director Office of Nuclear Reactor Regulation FROM:
D. F. Ross, Deputy Director T Division of Project Managemerrt
SUBJECT:
SUMMARY
OF MEETING WITH WESTIt!GHOUSE - CORRECTIVE ACTIONS FOR WESTINGHOUSE NSSS PLANTS AS A RESULT OF THREE MILE ISLAND UNIT 2 INCIDENT On April 11, 1979, the NRC staff met with representatives of Westinghouse Electric Corporation (W) in Bethesda, Maryland, to discuss short term corrective actions to be implemented at Westinghouse pressurized water reactors (PWR) as a result of the incident at Three Mile Island Unit 2.
Several W PWR licensees were in attendance.
A list of attendees is attached (Enclosure 1).
The meeting opened with an overview of the events at Three Mile Island Unit 2 (TMI-2) which require imediate attention by all operating PWR's as these events are perceived by the staff in light of information available at this time. These events are identified as Items 1 thru 12 in the NRC Office of Inspection and Enforcement (01&E) Bulletin 79-05A of April 5, 1979 (Enclosure 2). The staff specifically noted that the responsibility for development of corrective actions for these items rests with W and the utilities.
The corrective actions that are needed are specific instructions to be issued immediately to licensees of W PWP's.
These corrective measures will be reviewed by the NRC staff and issued by means of an 0I&E Bulletin.
W representatives then presented a summary of the activities which they have initiated since the TMI-2 incident to prevent the occurrence of a similar incident at a W facility.
Since April 1,1979, W has been working with its customers on this issue, and on April 5-1979, a meeting was held between W and its customers to discuss the potential for the occurrence of a TMI-2 type incident at W facilities.
Since then, W has been conducting additional stuaies concerning scecific plant concerns regarding the T'1I-2 incident and has conducted some computer analysis of the incident.
W has also asked individual utilities to compile plant specific Information which may bear on the probability of occurrence of mitiga+. ion of a TMI-2 type inciden'.
W representatives stated that the efforts undernay with their customers covers all the items identified in IE Bulletin 79-05A and some additional areas of review.
7 9 0 510 038T q/g u
^ /
E. G. Case AFRIL 1 2 1979 h[ then discussed the response of a typical 4-loop (four reactor coolant system cooling loops) PWR to a loss of feedwater (to the steam generators) transient.
The transient response reported in individual plant Safety Analysis Reports (SAR) is more conservative than the actual response experienced at W facilities for loss of feedwater.
For actual transients, the large steam generator secondary-side inventory provides a buffer between secondary (steam side) transients and primary (reactor coolant system) response to the transients. (( is still investigating, but as of this date, they are not aware of any loss of feedwater leading to a primary system pressure increase that caused a pressurizer power operated relief valve (PORV) to open. Therefore, a stuck open PORV similar to that experienced at TMI-2 should not occur for a loss of feedwater transient under normal plant operating conditions.
- However, if no credit is taken in the analysis for non-safety grade plant control system, PORV lift will occur; and there exist other transients which can lead to a PORV lift (and to the potential for a stuck open PORV).
Because it is not impossible to preclude PORV lift and the potential for a stuck open FORV, h[ performed computer analyses using conservative assumptions to determine the response of a typical 4-loop PWR to a stuck open PORV. Using the W-FLASH code and 10 CFR 50, Appendix K assumptions, W analyzed a 2h" dia. Loss of Coolant Accident (LOCA) break in the vapor space of the pressurizer. This break size is similar to the size of LOCA caused by a stuck open PORV. W also assumed the steam generators were isolated (main steam isolation valves are shut) and no charging flow makeup to the reactor coolant systems is in progress.
Three cases were analyzed:
Case 1 [with auxiliary feed system (AFS) flow to steam generator and with safety injection (SI)]
Results:
a.
The reactor core remains flooded with cooling water throughout the duration of the analysis ( Approx. 40C0 sec.)
and the parameters indicate that no uncovering of the core would occur thereafter.
b.
The pressurizer steam-water mixture level increases and stabilizes at about a 2/3.
Case 2 [with no AFS and with SI]
Results:
a.
Same as Case 1, a. and b.
b.
2/3 of steam generator levei is still present at 4000 sec.
c.
Reactor coolant system pressure approaches 1100 psi which corresponds to the temperature in the steam generators with safety valves lifting.
io\\
o6 u
E. G. Case AFRIL 1 ; :379 Case 3 [with AFS and no SI (i.e., no water makeup to the reactor coolant system at all)]
Resul ts :
a.
The reactor core would start to uncover at about 2100 sec.
Additional analysis is being done by E for Case 3 without the steam generators isolated.
And a comparison of Cases 1 and 2 indicate that the results are not very sensitive to AFS initiation for the time periods of the analysis.
W discussed the sicnals which initiate SI. Analyses which they submitted previously (Zion Station and RESAR-3 dockets) show that a small LOCA
~
in the pressurizer steam space may not result in SI initiation because the pressurizer level may not decrease.
A coincident pressurizer icw level (Lp) and low pressure (Pp) is needed for SI actuation.
But their analysis of containment building pressure following this LOCA shows that SI would be initiated by containment pressure high (no.1) indication setooint which is set at about 10% of containment design pressure at about 1600 sec.
At 1600 seconds, reactor core fuel surface temperature would be at the same temperature as the reactor coolant system coolant which is saturation temoerature for 1100 psi. This is far below the temperature necessary for core damage.
To provide additional assurance that SI initiates and prevents the core from becoming uncovered, in addition to considering the high containment pressure setpoint 3I actuation signal, W has instructer' its customers that SI snould be manually initiated if Pp decreases to the low Pp setpoint regardless of Lp readinc.
E is still evaluating the question of when to manually shut off SI following its activation.
The concerns are (1) that the SI system would fill the reactor coolant system completely and thus increase the chances of an overpressure transient which could overpressurize the reactor ccolant system or (2) that the operator would shut off SI based cn an erroneous pressurizer level and thus increase the chances of a TMI-2 type incident (core uncovery).
g presented a logic " tree" that an operator could use to determine if SI should or should not be shut off following events which lead to SI and low or failing pressurizer pressure and/or level.
E agreed that a bulletin similar to Bulletin 79-05A should be sent to its custcmers, but. additional clarification of the need to shut off SI to prevent overpressure as discussea -bove should be included.
E noted that the bulletin provision regarding containment isolation reset is not applicable to its plants because containment isolation valves do not ocen following an SI reset (as occurred at IMI-2) unless the operator deliberate 1; opens the isolation valves.
1 ' 7, 4
,6 G
E. G. Case d-ArRIL 1 2 1373 Following the 'd presentation, the staff discussed the followto action to be taken in light of the h[ information.
A belletin will probably be issued to h[ facilities in the next few days.
The bulletin will be essentially the same as GI&E Bulletin 79-05A but additional information will be included to determine plant specific corrective measures dealing with:
1.
Manual shutoff of SI, 2.
Management checking of safety system operability status, 3.
Possible elimination of Lp as an SI initiation signal by placing it in a " tripped" state, 4.
Possible requirement for containment isolation on high radiation signal for all plant.
The bulletin will state that our best information shows that, under certain transient and/or accident conditions, a level may be present in the pressurizer simultaneously with a decreasing prima y system pressure.
e
/* r Atw_Qd
/
i Denwood F. Rcss, Daputy'Cirector' Division of Project Management
Enclosures:
As stated cc w/ encl:
See next page q(
\\i Gg
E. G. Case 5-APRIL i e :373 Di st ri bution Docket (50-320)
NRC PDR Local PDR DOR Reading NRR Reading H. R. Denton V. Stello R. Vollmer W. Russell B. Grimes T. J. Ca rter D. G. Eisenhut A. Schwencer D. L. Zierrann P. Check G. C. Lainas D. K. Davis T. A. Ippolito R. W. Reid V. Noonan G. Knighton M. Fletcher D. Brinkman Attorney, OELD R. Fraley, ACRS(16)
J. R. Buchanan TERA NRC Participants oI
\\
c0
LIST OF ATTENDEES WESTINGHOUSE MEETING 04/11/79 NRC Carolina Power & Licht R. S. Boyd, DPM D. B. Wa te rs T. A. Ippolito, 00R J. J. Sheppard M. H. Fletcher, 00R R. Lobel, DOR Shaw, Pittman, Potts & Trowbridge E. G. Case, NPR F. Orr, DSS J. H. O'Neill A. Ignatonis, DSS N. C. Moseley, I&E 4:erican Electric Pcwer Serv. Coro.
J. L. Crews, I&E Region V E. A. Reeves, DOR J. G. DelPeriro N. Anderson, D0R E. Wenzincer, DOR Public Service Electric & Gas M. Mendonca, 00R L. B. Marsh, DOR P. A. Moeller D. Neighbors, DOR J. Wetmore, 00R Westinghouse T. V. Wambach, DOR A. Burger, D0R R. W. Stutter B. C. Buckley, DPM K. R. Jordan A. J. Szukiewicz, DSS V. J. Espusito J. Guibert, CCM W. J. Johnson A. Schwencer. DOR T
M. Anderson G. Zwetzig, DOR S. H. Hanauer, DSS Southern California Edison F. Schroeder, DSS L. P. Croker. DPM J. Rainsberry D. Vassallo, DPM A. Thadani, DSS G. Lainas, DOR D. F. Ross, DPM D. G. Eisenhut, DOR 26 195
U.NI.D S.
5 i:
ini NUCLEAR REGULATORY CCPMISSICN OFFICE OF INSPECT!CN A.10 ENP00. CEMENT WASHINGTCN, DC 20555 APRR 5,1979 IE Bulletin 79-05A C EAR INCIDENT AT THREE MILE ISLAND - SUPPLEMENT Cescription of Circumstances:
?relimicary information received by the NRC since issuance cf IE Eulieti. 7C-05 en April 1,1979 has identified six potential human, cesign and cechanical failures which resulted in :n c re damage and r:diaticn releases at the Three Mile Island Unit E n;cin plant.
The information and actions in this supplement clarifv m ca. tend the cricinal Eulletin and transmit a preliminary :h-enclegy cf tnt N I accident thrcugn tM first 15 hours1.736111e-4 days <br />0.00417 hours <br />2.480159e-5 weeks <br />5.7075e-6 months <br /> (Enclosure 1).
1.
At the tima of the initiating event, less of fee: water, both of the auxiliary feedaater trains were valved cut of servicae 2.
The pressuri:er electrematic relief valve, which c;pe t d '. b ;
the initial pressure surge, failed to close
..".en
- n3 press ee decreased belew the actuation level.
3.
Folicwing rapid depressurizatien of the pressuri:er, the cressurizer level indication may have lead ;o erronecus inferences of high level in the reactor coolant system.
The pressurizer level indication apparently led the operators to prematurely terminate high pessure injection ficw, even though substantial voids existed in the reatter c:clant system.
4.
Se:ause tne c:ntainment does not isolate en high prescura infectich (HPI) initiati:n, the highly radicactive water frcm the mlief valve discharge was pu ped out of the containment by the lutamitic initiation of a transfer pump.
This water entered the radi: active waste treatment system in the auxiliary building where seme of it everficwed t: the ficer.
Cutgassing frem tnis water and discharge through the auxiliary building ventilatien system and filters was the principal source of the offsite release of radicactive noble gases.
5.
Subsequently, the high pressure injection system was intermittently cperated attempting to control primary coolant inventcry icsses through the electrematic relief valve, apparently based on pressuri:er level indication.
Due to the cresence of steam and/or ncncondensible voids elsewhere in the react:r coolant system, s
e mm m
. 5. i s l e.. o a.,J r. 5.e r n. s,.,.. h. m_
- s 4 - :
4
.-m gJ i, -
DUPLICATE DOCUMENT g<
jq' cU i,t dh Entire document previously entered
_O' linto system under:
ANO 746Ml7 9 O
a7 0 Lf No. of pages:
$