ML20246N093
| ML20246N093 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs  |
| Issue date: | 08/30/1989 |
| From: | Cintula T NRC OFFICE FOR ANALYSIS & EVALUATION OF OPERATIONAL DATA (AEOD) |
| To: | |
| Shared Package | |
| ML20246N050 | List: |
| References | |
| TASK-AE, TASK-E906 AEOD-E906, GL-89-04, GL-89-4, NUDOCS 8909080034 | |
| Download: ML20246N093 (5) | |
Text
- - _____ ___
f y '-
yq
.L p .
y,,, ,
6
{,'< AEOD TECHNICAL REVIEW REPORT *
' UNIT: Calvert Cliffs Units 1 and 2: TR REPCRT NO.: .AE00/E906 DOCKET NOS.: 50-317/318 DATE: August 30, 1989
. Baltimore Gas &. Electric Co.
t.ICENSEE: EVALUATOR / CONTACT: T. Cint':la R SS/AE: CE/Bechtel
SUBJECT:
FAILURE OF STEAM GENERATOR ISOLATION CHECL VALVE DISCUSSION Event Description
- On March'17, 1988, Unit 2 was in cold shutFiwn for maintenance and repair cf the manual stean: isolation valve (2-NS-109) to No. 21 auxiliary feedwater pump turbine. This valve had been previously-identified to be leaking through its seat- during' performance of _the ten-year IST Steam Generator (SG) hydrostatic .
pressure' tests. When the valve internals were removed, an additional piece of metal was found. Disassmbly of the S/G ' isolation check valve, 2-MS-103,
- revealea that the piece of metal was part of the internals of eneck valve MS-103.. Additional damage was observed and it was concluded that the check valve would not have been able to perform its isolation function. Disassembly of the identical' chect valve (2-MS-106) of the other S/G found the valvs internals-to be worn. :Both chegk valves (Chapman / Crane 6" - 600# tilting disk
- check valves) were replaced with nGw valves (Anchur/ Darling 6" - 900# tilting disk check valves) with an improve.d seating material (LER 316/88-003).
To' prevent recurrence _ of this type of failure, the licensee committed to reverte leak' tent the check valves every refueling outace, or one valve per refueling i outage et11'ne disassembled and inspected, If ti,e inspected valve shows sicts .i' of wear which indict.re degradation leading toward fa!1ure, the ot k r check valve will be disassembled and inspected. Partial stroking for pcsitive flow verification is accomplished by a monthly surveillance te t to verify the operability of the Auxiliary Feedwater Actuation Systein, !
00 October 29, 1988, Unit I was in cold sbddown for a similar maintenance putage. While performing a surveillance test cn the steam generator isolation theck valvcs,1-MS-103 teaktd by. 1-MS 103 was disassembbt and the disk was found misaligned appr0G%tely one-quarter inch dus to excessive wear of the hinge pins ar4 bushtng area. 1-MS-103 was tt:e Anchnr/9arling 6" - 903 tilting disk check valve that had 3een installed at Units 1 and 2 in Aoril 1083 as a result of the Ma=ch 17. 19G8 failure at Unit t (LER 317/66-014}.
3 ~
The cacse of (, heck valve tailure was excessive cycling. It was subsequently detentined that the cycling was due to the leak by of an upstread. isolation '
!alve lens-102, two feet away.
Check valve 1-MS-103 was rep' laced with an idertical valve. The licensee is evaluating improvements to the system, e.g., check valve design, location and mater 161s.
8909060034 890G30 PDR ADOCK 06000317 P POC
1 I Lic.ensee's Safety Analysis l l l l These check valves provide isolation of the S/G in the event of a main steam line break (SLB) upstreae of the MSIVs. Each unit has two S/Gs. The licensee i in the LER postulated that a ruptured steam line Would result in the blowdown of the entire inventory of that S/G. With the inoperable check volve, the '
unaffectedS/GcouldexperientealimitedbloydownthroaghtheaffectedS/G.
This would correspond R an additional 0.2 ft break.
The licensee calculated that noditional steam flow of an open six-inch AFW pump turbine steam line during a St.B would increcse the peak containment pressure to 52.1 psig. This exceed.s the design pressure of 50 psig, but is weli within the calculated ultimate (cntainment pressure of 124 psig. This additional steam flow wculd riot af Tect the SLB site boundary doses because the Counding assumptions envelope the six-inch c, team lire flow.
FSAR Accidunt Analysis Steam line breaks are charactedad as cooldown events due to the increased steam flow rate, which causes excessive energy removal from the S/Gs and the reactor coolart system (RCS). This results in a decrease in reactor coolant temperature in the RCS and in turn, an increase in core reactivity from the negative moderator and doppler reactivity coefficients.
Stcam line breaks are analyzed iii the FSAR to assure the resctor ccre retains subcritical following the reactor trip and a post trip return to power does not occur. The analysis assumes the reactor ccre is at an optimum condition return to criticality and for conservatism, assumes a single failure 6ccurs that would add the most positiyo resctivity to the ce,re. Inis af3umptien is that the most reactive cor.tici element assecbly (CEA) is in its fully withdrawn position i after the reactor trip. The analysis assumes these conditions and p edicts no return to power foibwing a steam lir.e rupture. Leturn to power might resbit in high power peakirg far.tcrs and subsequent fuel damage (local boiling at mocerate power lwels). All of the above a::sumptions are analyzed for one steam generator blowing down. The Standard Review Plan requires safety rev1ews of the FWR design to cenfirm that the blowdown of more than one S/G assuming a concurrent single active component failure will not occur.
i For the worst case SLB, the Ca7 vert Cliffs FSAR assures a circumferential break
~
of one 34-inch steam line occurs at the S/G main steam ?ine nozzle. It is assumed that steam from tha other str :m generator flows through the intact steam lire and out of the break until the main steam isolation valves are closed. The automatic closure of both main steam isolation valves is assened to occur 6.9 seconds after a low steam generator pressure signal. Thereafter, only the ruptured S/A continues to blow down. The rapid cocidown of the RCS continues until auFiliary feedwater (ffW) flow is isolatt6 to the ruptured S/G and the ruptured S/G blows dry.
The AFW system automatically starts to deliver flow to the S/Gs upon detection of i low level in either steam generator. The AFW system consists of one motor driveri and two steem turbine driven pumps. Upon automatic initiation of auxiliary feedwater, one riictor driven and one turbine driven pump automatically start. The SLB analysis assumes that AFW flow is delivered tc the damaged 4
p J
c
- . steam generator at *.he time of the reactor trip. Subsequently, detection of a differential pressure'between the $/Gs in excets of set point blocks AFW flow to the ruptured steam generator. This feature ensures that AFW flow to a ruptured steam generator is terminated, tc prevent a return to power which could be caused by excessive cooldown of the primary system.
AEOD CONCERNS According to the LER, the license-a assumed that valves CV-4071 and CV.4070 will be ooen during the beginning of most SLB events, and if shut; these valves will not hold pressure in the reverse direction. In a follow-up discussion, the licensee explained that valve:; CV-4070 and CV-4071 are normally open air-operated salves; .1.e., spring to open and air to close. For these valves, the closing force of the air operator would be insufficielit to overcome the valve spring foice in combination with a reverse differa tial pressure. Tha nn pose .
of the downstream check valve (MS-103) is reverse flow protection.
If.the analyzed steam line rupture were to occur with the conditions reported by the Calvert Cliffs LERs, i.e., an inoperable S/G isolation check valve MS-103 on the ruptured steam line (Figure 1), an unanalyzed event inay occur because the positive core reactivity addition could be greater then the analyzed event in the FSAR., (It should be noted that CV-4070 is equipped with a handwheel and can be closed manually). With these conditions, the safety significance of an open S/G isolation check valve in the affected S/G in a SLB may include the following adverse system interactions:
. 1.. The unaffected S/G would also blowdown through the rupture in the steam line. The possibility of more than one S/G blowing down was not considered credible in the safety analysis. It is possible that the ensuing primary system cooldown with both S/Gs blowing down may add more positive reactivity to the reactor core than analyzed in the FS?.R.
- 2. Both turbine driven AFW pumps may be unavailable due to steam from the S/Gs hypacsing the AFW turbines and flowing out of the break in the stam line.
The single motor driven AFW pump would start and deliver feedwater to both S/Gs on low S/G level indication in the faulted S/G.
- 3. With both steam generators blowirig dn9n through the common steam line ,
rupture, the AFW isolation signal based on S/G differential pressure (365 psid used in the TSAR; 135 psid plant Technical Specifications setpoint) may not occur to isolate feedwater flow to the ruptpred S/G. The AFW isolation signal does not have a seal-in feature, so it is possible that the mot'or driven AFW pump woult intermittently or contuuously feed both S/Gs. Continu6 tion of feeawater to the ruptured S/G should result in a prolonged cooldown of the primcry system.
- 4. The blowing down of both S/Gs through a comnion steam line rupture will make it more diffic' ult for the operators to identify the ry tered S/G due to lack of training, lack of prior simulator demonstrations of this scenario, i ambigucus S/G signals, insufficient steam pressure for the turbine driven AFW pump, intermittent or lack of 5/G isolation and excessive primary system cooldown.
c____ m____.2 __ _ . _ _ _
s.
~ .
a.
l Flow WestriMor ,
- %S - %2 w,
>c -
m . . .-. ,,
CV-4070 MS102 MSIV 2
['] 4 6.35 fra 9reok Size Aomu ..m T- weer_%g RUPTUNED / MS*l03 SG NO.2 t y I I -
, t 11 V cJM7 MS-IO6 CV-4071 f ' f MS-105 w, fds!v i
- U ._ _-- l r p -
y., -
l f
INTACT SG 1 l N O.1 i
l i
l l
l s
1 l
l Figure 1. Anticipated Valve Line-Up During ;
an SLB at Calvert Cliffe l
i l
-s s
3 5-J CONCI.USIONS k
Licensees need to ensure that steam generrone isolation check valves are capable not only of passing steam to drive the auxiliary feedwater pcmp turbines, but also of preventing a simultaneous blowdown of the steam generators. Inservice testing of these check valves should be conducted on a regular basis for both the forward and reverse flow directions. NPC Generic Letter 89-04, " Guidance on Developing Acceptable Inservice Testing Programs," will address the need
'for back f?ow testing of these check valvss. In addition to testing, periodic disasseinbly of the check valves to inspect their internal condition could pro-vide valuable information for the detection of valve degradation before failura.
4
_ _ _ _ _ . _ _ . _ _ _ _ _ _ _ _ _ _ _ _ _ _