ML20090B528
| ML20090B528 | |
| Person / Time | |
|---|---|
| Site: | Ginna  |
| Issue date: | 05/25/1982 |
| From: | Speis T Office of Nuclear Reactor Regulation |
| To: | Lainas G Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML082380335 | List:
|
| References | |
| FOIA-91-106, RTR-NUREG-0909, RTR-NUREG-909 NUDOCS 8206210075 | |
| Download: ML20090B528 (16) | |
Text
_. _. _ _ _.. _ _ _ _. _. -..... _ _ _.. _ _ _ _. _
/ -
~.
+
i o $.3gd1#'
p-Ju)geff" 1
U 26 O
'/
e)
ENDPMDUM FOR: Ce C. Laines. Asstatant Director Per seat
~
Division of Licensing FROM:
Thesis >
'911. Assistant Director for Reactor Safety Division c ystems Integrttion
$UBJECT:
GENERIC RECOPNENMTIONS TROM THE DSI (RSB/ICSB/CSB)
REYlEV 0F NURLG-0909 In the May 3,1982 referandum (H. Denton to the NRR Directors) we were requested to review the Task Force report (WREC-0909 'NRC Report on The January 25, 1982 Steas Generator Tute F:upture on R. E. Ginna Nuclear Powr Plant *) for the developnent of generic reconvendattors.
Enclosed are Reactor Systers Branch (R58), and thstrwentation and Control Systers Branch (ICSB) recoerendations for generic action based on their review. There are no generic recoewendations from containrent Systems Branc5(C$8).
or v us u n ns %
Jhemla P. Stels Themis P. Spets. Assistant Director
~
for Reactor Safety Division of Systees integration m..--
J
/
y/
l.:
A\\du
'4,
- g. - q3
\\9 pp i
1 n
+
u
=
~~ - - - - + -.., - - - - - -.. - -
-.-4,.
,,,_,_,..-~_.,,._m.-,-
J Ggt;RIC RECOWENDATIONS FROM tS!4RSB/lCSf[CS0)JEYlfW F h!!R_EG-0909 1.
_Recomendation - Letdown isolation __
A containnent isolation signal redundant to that being prvvided to the letdown control valve (LCV) should be provided to each of the three letdown orifice valves.
A safety grade air source, such as an accumulator, should be provided to assure the capability to close the letdown control valve on a containment isolation sicnal. This will provide redundant isolation of tre letdown steam from the relief valve to ensure the pressure boundary, t.icensees should perform an evaluation of their letdown control system to detennine if it can be isolated by redundant means to maintain pressure system integrity during containment isolation.
Discussion As described in Section 3.3.5 and 7.4.2 of NUREG-0909, the letdown control valve did not provide complete letdown isolation during the containnent isolation phase, since it was dependent upon the operability of instrunent air which is isolated by containnent isolation.
The outside containrent isolation valve closed allowing the pressure to build to the relief valve setpoint thereby discharging into a pressurizer relief tank.
I
2-Rochester Gas and Electric (RG&E) proposed a Itodification to prevent the
[
pressure build-up from recurring by providing a contair, ment isolation signal to the LCV inside containwnt. The LCV has an interlock which will auto-matica11y close the three letdown orifice valves.
If the LCV f ails to close (for whatever reason) then there is no guarantee that the orifice valves will close to provide isolation of the pressure boundary.
The implementation of thir, recorrendation should be such that the resetting of the isolation signal does not automatically reopen the letdown isolation k
- valves, p.copening of the letdown isolation valves should be a deliberate
^
operator action.
~
- - - - - - - ~ - - - - -
- 9 3
2.
Reconmendation Coo _1_ing Faulted 'SG Licensees should perform an evaluation of their SGTR procedures to confirm that a fornal nethod is outlined in the procedures t'or cooling the affected steam generator. Systein used should be of sufficit it capacity and capability for cooling the steam generator until the residual heat reroval system is placed in service and assuming offsite power unavailable.
t Discutsj,on As a result of the review of the Ginna steam generator tube rupture (SGTR) event.- the task force determined that the licensea's procedure did rot include formal instructions for cooling the affected steam generator following closure of its M51V. Oral instructions developed by the
~
t Technical Support Center, directed the operators to feed the steam generator with auxiliary feedwater, up to a evel of about 60% of narrow i
range and reduce primary side pressure by spraying the pressurizer to an I
indicated pressure of about 25 psi below the secondary side pressure.
The secondary water v.as allowed to flow through the tube break back into the primary side in order to accomplish the desired cooling.
m M
.._m_-..__.__._____
l e
4 3.
Reconmendations Pump Operation Puyo Trip Licensees should perform a forwel study to find a method for keeping the RCPs operational during SGTR events sucit as occurred at Ginn?. This study should consider, among other things, additional HP! capacity, more discrimin$ ting criteria for RCP trip, or, if RCP trip cannot be avoided, a qualified auxiliary spray capability.
Put@__ Restart Licensees should perfonn a study, in conjunction with the RCP trip study, l
to determina the criteria for RCP restart following RCP trip. The criteria should not.icopardize plant safety or result in operator confusion.
In partl-cular, the
- dy should address:
o The need for a bubble in the pressurizer prior to restart.
o The minimum volune of water needed in the pressurizer prior to restart, considering the maximum volume of steam that could potentially exist in the vessel head, i
o The need for a system subcooling margin prior to restart, and the effects of bulk system flashing if the pressure drop due to head bubble collapse following RCP restart drops the system pre'sure below saturation.
s t
va v w,-,,e n
p.--
+-n' ere-
+
-.m-a
~v
-n e-e.,
n
-+-,
---r,
,,ve-e
,,w, a=--w
Licensees should also examine the procedures to require the earliest possible restart of a RCP, if the RCPs are to continue to be tripped under existing criteria.
Discussion The staff has been studying the issue of PCP trip and restart since tne TMI-2 accident. The issue has also been extensively discussed with both the utilities and the PWR vendors. The hbove recommendations, while reflecting specific conclusions drawn from the Ginna SGTR event, will be superceded by mare specific, generi.: guidance from the staff that is now in final preparation.
Thus, if these specific recommendations are forwarded to the industry,
~ they should be properly clarified as specific concerns, resulting from Ginna, and that they probably will be superceded with a generic resolution guidance letter which will incorporate all of the elements of the RCP trio and restart Conc 0rn.
g.,
- 4. Recorrendation - Cooling Intact SG Licensees should evaluate the need for and consequences of securing the con-
~
denser as a means of removing energy from the intact steam generator, and utilizing only the atmospheric relief valve on the intact steam generator.
))
Discussion _
The decision to secure the condenser and use only the ADV relieving directly to atnospheric during the Ginna ever? may have reduced some in-plant equip-
]
nent contamination, but it added /
r'
.ite releases and reroved the normal means of plant coolcawn. Also, the decision to break vacuum on the condenser removed this means 6.1 a cack-up means of energy removal should a problem have occurred with the inuct SG ADV, or its cv.: trol equipment.
f 5.
Recomnendation - Overfilling Faulted SG Licenlees should evaluate the potential for and consequences of overfilling the damaged steam generator, including the effects on the safety valve.
Discussion Overfilling of the damaged SG can result from excess feedwater addition due to untimely feedwater isolation, or from sustained primary-to-secondary leakage. During a Cinna-type event, sustained leakage i be caused by excessive prinary pressure, or a det.rease in the secondary pressure due to maloperation of the safety valve, atmospheric relief valve, main steam isolation valve (or its bypass) ur the blowdown system.
The consequences of overfilling the SG can be severe.
The excess weight may damage the steam line and result in
()
7 continued releases and a loss of primary coolant and RWST water outside containment. Also, filling the damaged SG can result in challenges to its safety or relief valves, and releases offsite. The iodine partitioning is also affected by the snount of water in the SG. Generally, more iodine is in the water than in the steam or vapor above it.
If water releases out the safety or relief valves occur, the radiological consequences would be more severe, than if steam releases occurred.
6, Recommendation - Operator Guidance Licensees should provide additional guidance for operator action in the event of real or suspected void formation in the reactor vessel upper head.
Discussion Several events includino the Ginna SGTR, St. Lucie Cooldown and the TMI accident have produced voids in the upper head which proved difficult to eLJminate and were, in varying degrees, unexpected.
RG&E in their SGTR submittal requested further guidance on RCS management when hubbles are present. During an SGTR, there is a need to depressurize the RCS to minimize leakage to the steam generator and to the atmosphere. Since void formation nay occur, and operators may perceive the void to be a threat to core cooling, nore positive guidance should be given on bubble nanagement and condensation. This is clearly generic in nature and is an offshoot of owner's group work on operator guidelines. This should include detailed TH analyses verifying the behavior of voids fonned in the RCS.
,~.
(')
~8-7.
Retomendation - Achieving Cold Shutdown _
Based on the high frequency of SGTR events and our evaluation of the requirerents for achieving cold shutdown, licensees should evaluate the plant systems' capacity and capability for achieving cold shutdown follow-ing the SGTR event.
Conservative assumptions should be made reoarding the amount of primary and secondary inventory lost through a failed open SG safety valve, conservative times for achieving shutdown and loss of offsite power.
Descrintion Cold shutdown of the Ginna Power Plant was achieved approximately 33 hours3.819444e-4 days <br />0.00917 hours <br />5.456349e-5 weeks <br />1.25565e-5 months <br /> following the SGTR event. Auxiliary feedwater was used for decay heat removal for about 20 hours2.314815e-4 days <br />0.00556 hours <br />3.306878e-5 weeks <br />7.61e-6 months <br /> prior to initiation of RHR cooling. There fore,
in the event of a continuous leakage of primary or secondary coolant through a failed open steam generator safety valve or other malfunction, coolant inventory could have been depleted to unacceptable levels before achieving cold shutdown.
8.
Recomendation - PORV Operation Licensees whose pressurizer PORV is air operated should ensure that they have not crimped or otherwise altered the air system solenoid valve vent path so maloperation of tne PORV is not experienced. The potential for blockages in the air supply and vent path should also be examined.
.~
.g.
Discussion Although' not conclusively demonstrated, there eppears to be a strong possibility that the Ginna event demonstrated that-the pressurizer PORV air system vent flow p6th should not be tampered with to avoid restricting the PORV air operator vent path and causing the PORV to stick open. With the PORV stuck open in any scenario, the operator must rapidly detect, diagnose and correct the problem (isolate the PORV).
It is not acceptable that any over-pressure event,_like a turbine trip or loss of feedwater, be turned into a small break LOCA due to maloperation of the pressurizer PORV.
9.
Recorrendation - RCS Pressure Control The optimal. means of controlling and lowering RCS pressure in the SGTR accident should be determined to avoid rapid depressurization, void formations, and continued primary-to-secondary tube leakage.
Discussion The control of RCS pressure during the SGTR determines, in large part, the tube leakage and therefore, releases out the damaged SG relief or safety valve.
Too rapid a depressurization can cause void formation in the upper head and hot legs, especially if the plant is being cooled with natural circulation.
In such an eventuality, it may be difficult to reduce system pressure without restarting the RCPs and initiating normal pressurizer pressure control. The use of the PORV as a neans to rapidly depressurize the system should be evaluated, as well as the use of auxiliary and normal spray systens. The I
review shnuld be performed with the goal of minimizing complicating conditions, expeditious plant cooldown and reducing,to the maximum extent possible, offsite consequences.
l l
.~
1 10 -
10.
Reconnendation - The Single Failure -Criterion NRC should determine if single equipment failures should be assuned in the course of analyzing a SGTR accident. Also, determine if the loss of offsite power should be assumed in a rechanistic manner. That is, should licensees or applicants assume offsite power is retained until the turbine generator has been tripped.
Discussion The SRP for the analysis of a SGTR does not require applicantt to assume a coincident or consequential single failure. The analysis generally assumes the damaged steam generator safety or relief valve is assumed to correctly operate. Based on the steam generator overfill experience (Point Beach and Ginna SGTRs). and the resulting challenges to the safety and relief valves, and the significant impact the failure of the safety or relief valve would have on system cooldown, inventory control, offsite consequences and in general, the operator's ability to cope with the event, it is prudent to corsider if single failures should be postulated along with the event.
With respect to offsite power, the loss of offsite power at sometime beyond the initial event may place a more severe restriction on systems and the operators than a concurrent loss of offsite power.
11
- 11., Recommendation - Credit For Control Grade Equiprent NRC shvuld detennine if credit should still be given for the use of control grade pieces of equiDinent in the mitigation and recovery from a SGTR.
Discussion _
j Analyses and procedures for operating plants are perfomed assuming a number of control grade pieces of equipment. The prcssurizer PORV is relied on to lower RCS pressure and the intact SG atmospheric relief valve is used to remove primary energy to cool down the RCS. The PORY quality has been improved after TMI, but the val've control system is basically control grade.
Likewise, the ste 9 generator atmospheric relief valve ( ARV) on operating plants is not a safety grade piece of equipment, although on newer plants, the ARVs are required to meet RSB BTP 5-1 of SRP Section 5.4.7.
NRC needs to determine if credit for control grade pieces of equipment should continue to be given in light of the f ailures that have been experienced and the f requency of the SGTk event itself.
This determination should be I"
ude concurrently with items 6, 9, and 11.
- 12. Recomendation - SGTR Analysis Requirements NRC cetermine if the SGTR should be reclassified and the requirements for analysi.s made more stringent.
Discussion 5
No other PWR " accident" has occurred as frequently as the SGTR. The Ginna event is the fourth domestic SGTR and, considering the general philosophy that accidents should rarely if ever occur, it is prudent to reconsider the event classification and the SRP assumptions and criteria.
It may be appropri-
- - - ^
['
O ate to consider mre than a single broken tube as the initiating everd, or, an additional single failum, or the offsite consequences must rieet the 10 CFR 20 limits rather than the 10 CFR 100 limits.
13.
Recorrendation - RCS Dilution Evaluate the ootential for and consequences of RCS dilution as the RCS pressure is reduced below the damaged SG pressure and relatively dilute
~
fluid flows back into the ECS.
Discussion Since the past SGTR events have demonstrated that filling c' the steam generator is possible in the cooldown phase following the event, dilution of the RCS can occur.
It is not readily apparent that without proper manitoring ana chemistry control, recriticality is not a concern. The maximum dilution of the RCS should be deterTained, and procedural steps taken to ensure operators are cognizant of this phenomena and take the appropriate actions.
C 14 Recommendation - Upper Head Thermocouples All PWRs should have thermocouples in the upper head as a valuable aid in diagnosing and managing various cooldown events.
Discussion for most PWRs the upper head is a rather stagnant location especially after the pumps are tripped. Also for most PWRs, the upper head has relatively high fluid temperatures.
Therefore, for those events where depressurization occurs, the upper head would tend to flash prior to other locations in the
- - =. -.
( )
i primary system. Thermocouples located stratigically in the upper head would be extremely valuable in monitoring the approach to saturation and anticipating bubble fonnation.
Ginna happens to-be one of the very few PWRs with thermo-r couples in the upper head. Although not well placed for saturation monitoring, they still proved valuable in following the course of the accident. Obviously thermocouples arranged vertically in the upper head would give the best prediction of incipient saturation.
The recommendation for thermocouples is not meant as a substitute for reactor vessel level instrumentation (RVLis) as required by the THI Action Plan. Thermocouples cannot provide positive level
. indication but are meant as a prediction for the or. sat of bubble formation.
- 15. Recommendation - Pressurized Thermal Shock 1.
Cold leg Temperature-Monitoring in the long-term, as part of the overall PTS resolution, evaluate the need for _
supplemental temperature monitoring-instrumentation on the RCS cold legs between the reactor vessel and the safety injection inlet nozzle.
~
- / f-T U51 i c9 C.n.n ;13, crid 17
- crators (+' ect
- rs a re lo:t'ed :n iho. ng 'c; t-i en t" s I Ut y i r l :::' L.-
" -t r.: : :l e u.c t is s ' e r
- cre ct.t
- r sc that daring E. E r.t S
'2:e the f if5121 is Ec'.L30bd, carate dett)cinPtion of the fit'd t:
?rEtL:E I%tir$D; the I c 3 Cic r '. it! &l ( ? n . be 2de.
The current
'. S C i t i : r. c f '.E'*0 E ri t b r S d6t'".tcr5 CC ~-p } ' G J! Itf E":'yf i! 0f.-;*f!vri!ed thF r-dl I'.0Ci C0r:5i t'6 ration a f ter the E.h r.0, tred G.is if.S ;, : f t '.Or,0 t E n ti a lly 5
~
c -cm cative i'.f0 oticn 2 4 carding the cold I?g t-
. S r s '. '; c e s.
5c r e t o r 2ct is '
' Ztr01 fff'3*'iM !L'a ds h?3 PIS M nt SM ld M IHed On the r
- t l' itir;
- , ;e<
- re c:a. tion, aith : uld te Me
't id w ur
- re cf ;Le 51 i-itt ner:le.
Actices t : 9f cn 't;
.l estru-(
_1 c.
'c
,+.
.y
.t _e'mi m
t
~
13 -
2 Enhanced Cold loop Flow Licensees should review and identify all potential transient and accident scenarios that could produce relatively stagnsnt flow conditions in a coolant loop, and examine the effect of the operator actions which could draw the cold water into the vessel (e.g., oper PORY high point vents, depressurize secondary).
In the near term, for the scentrios identified, review and modify procedures and train operators as necessary to prevent or minimize the flow of cold leg water into the vessel.
In the long term, as part of the overall PTS resolution, perform detailed PTS evaluations of the scenarios identified.
Discussions The staff examined the Ginna scenario to determine if any actions or variations of the scenario coulti have directed the flow of "rold" HPI water into the vessel.
If it is assumed the enhanced fiev of cold water away from the vessel was due to thE open PORY in the hot leg of the broken generator loop, then the same of this scenario, but with pressuruzor located in the loop with the intact steam generator, needs to be examined, This is because opening the PORV in the hot leg of the loop with the intact steam generator could put the low pressure point in the system in a location such that HPI water wouir]
preferentially flow towards the vessel to rear.h the PORV rather than away for the vessel,
9 A similar situation could exist if high point venting at the top of the vessel occurred. Ibwever, it is not known if high point vents exist on the top of the Ginna vessel.
M]reovar, the NRR has not yet allowed any licensee to operator high point vent;.
The above two scenario describe a general class of events in which one Icop is stagnant and cold with respect to the other loop (s).
The general concern is that a ventieg,iction could draw relatively cold HPI water into the vessel L
prior to being mixed by a loop circulation flow, l
i D
~
Y a
w O
O
- - _ - -