ML17279A645
| ML17279A645 | |
| Person / Time | |
|---|---|
| Site: | Columbia  |
| Issue date: | 08/18/1986 |
| From: | WASHINGTON PUBLIC POWER SUPPLY SYSTEM |
| To: | |
| Shared Package | |
| ML17279A644 | List: |
| References | |
| TAC-62181, NUDOCS 8712070051 | |
| Download: ML17279A645 (24) | |
Text
INSieUHLHTAT ON 3/4.3.8 TURBIVE OVERSPE""0 PRO>cCTIOt't SYSTEM'i L'BITING CQHOITION FOR OPERATION 3.3.8 At laas one turbine overspeed prctaction system shall be OPLRABLK.
APPLICABILIY:
OPERATIONAL CONDITIONS 1 and 2.
ACTION:
a.
Mi ". cne turbine governor valve or one turbine throt.le valve per steam chest inoperable and not closed, restor the inoperable valve to OPERABLE status within 72 hcurs, isolate the affected s
earn
'hest frcm the st am supply, or isolate the turbine from the steam supply within the nex 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.
b.
c With one turbine interceptor valve or one turbine reheat stop valve f inoperable, restore the inoperable valve to OPERABLB status within 72 hours8.333333e-4 days <br />0.02 hours <br />1.190476e-4 weeks <br />2.7396e-5 months <br />, or close a
least one valve in t>>e affected steam line or isolate the turbine from the steam supply within the next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.
Mi h ai her of tha tha above required turbine overspaed pro.ac.ion systems otherwise inoperable, isolate the turbire from the steam supply within he next 6 hours6.944444e-5 days <br />0.00167 hours <br />9.920635e-6 weeks <br />2.283e-6 months <br />.
SUR'I--,LLAt/C REOUIRcycqTS 4.3.8.'he provisions of Speci ication 4.0.4 ara not applicable.
4.3.8.2 Tha above required turbine overspeed pro action system shall ba demonstrated OPLRAB~ 5:
a.
At least ence per days by:
Cycling each of the following valves through at least cne ccmalete cycle from the running position ',or the overspead protection control sys-em, the electr cal overspead trip svstam and the mechanical ovarspeed trip system; I&~
3.
A ~
Four high pressure turbine thrctt1a valves, S~o'~ pr~ssura turbine reheat s "o va'ves, cur 'lig 1 prassu u o ine governc.
- valves, and Six low pr assure urbine intarcactor valves.
87i207005i 8608i8 PDR ADDCK 05000397, P
'~A i!.VCTCN NUCL=iR -
UNT.T 3/ a ie4Q
WASH)NCTON PUBLlC POWER SUPFLY SYSTEM c'.0 BOX 966. RICHLAfiD,Vi'ASrffNGToiX 9"=362 WARRANT No. 05872 5871 AY TO
~
U ~ S NUCLe.'AR R-"frU'T CRY CGYt'. ~
o COI+l X" SIGN ORDcR o O
o 4Vashington Public Pov'er Supply System
'rrersore From Anv Moines in rne Ceuolee'rv,eC:
NV. 2 Oeienue ACCOur.:
EEAF ln$T QAicrc soot
~ aT~ A'u'E'uvE EEATT E, 'u>> 96sS4 f~i'
'lg ~ ir
~ Ir i~ ~ i ll'0058 7 2ll': 4 250000 24t:
9R 20 E
h 2ll'NVOrCE NO
';OuCN=-R l -" O CONTR.TOT GROSS 4'fOVN ~
l OfSCOJN Nc.T A>.fOVNT C ~1o 86 Sl R1 O ".
QL AoPLTCAT T CM F E=.
CHECK NO OAT=
l VENOOR NO YcNOO>
N4V=
TOT%V, AilCUNT
~l'.71I f:E1 "2 tl UU.".5o U ~ 8 ~
KUCLiAR RLroULAT~$?Y CQHHe
STATE OF MASHINGTON
)
)
County of Benton
)
lg )
Subject:
~x:> +.
~J I, R.
B. Glasscock, being duly sworn, subscribe to and say that I am the Director, Licensing
& Assurance for the MASHINGTON PUBLIC POMER SUPPLY
- SYSTEM, the applicant herein; that I have full authority to execute this oath; that I have reviewed the foregoing; and that to the best of my knowledge, information and belief the statements made in it are true.
- lasscock, Director Licensing
& Assurance On this day personally appeared before me R.
B. Glasscock to me known to be the individual who executed the foregoing instrument and,acknowledge that he signed the same as his free act and deed for the uses and purposes therein mentioned.
GIVEh under my hand and seal this~ day of ~y iv.i < h C
, 1986.
I
-o.7-'~; i",~-..
h i) otary Pub ic in and for the State of Mashington Residing a
t.i'Y.
iri s 'i'~ i llgq~l
Safety Evaluation For the Impact of Reduced Turbine Valve Test'ng on the Operation of Washington Nuclear Project - 2 INTRODUCTION HistoricaU.y, Nestinghouse has recmxnended that turbine valves be tested periodically.
A weekly test recoamendation originated in the mid-1950's primarily as a result of service experience assoc'ated with fossil plant application at that time and in recogniticn of the importance of reliable turbine generator operation as it relates to operating personnel and equipnent protection.
'Ihe mp0rtance of frequent valve testing to maintenance of the integrity of systems necessary for the safe operation of nuclear plants has never been clearly established.
Nevertheless, the per'odic valve testing recommendation has evolved into a license r equiranent for certain recent nuclear power plants by virtue of its inclusion as part of plant technical spec'fications.
The technical specification requirenent appears to be arbitrarily applied-in some cases - weekly test intervals, in others - monthly, and in still other s - ro requirenent at, aU..
Based on the lack of a demonstr ated safety need for frequent turbine valve testing and the inconsistency with which the requirement to test turbire valves has been applied, an evaluation to min'vtze turbine valve testing frequency is warranted.
In 1982 and 'n support of Alabama Pcwer Company's Farley units, Nestin@ouse performed a probabilistic study that demonstrated that less frequent testing of turbine valves does not influence valve reliability or failure rates and frcm a safety point of view yearly testing is adequate.
Th's study which is docunented
<<n NCAF-10161, "Evaluation of 2npact of Reduced Testng of Turbire
s Y
Valves", is specific to Farley 1
and 2 but is genera11y applicable to Westin@ouse nuclear turbines of the type installed at Farley (i.e.,
BB296 units with steam chests).
Another report WSTG-3-P,
~Analysis of the Probability of a Nuclear Turbine Reaching Destructive Overspeed" which was prepared by Westin@ouse Steam Turbine Generator Division further docunents the impact of turbine valve test frequency on turbine valve reliability.
Based upon the results docunented in these reports Westin@ouse revised the reccamended turbine valve test interval from weekly to monthly.
In addition to the Farley units, the results of these studies have been successfully applied to the McGuire, Crystal River, and St. Lucie units (all BB296 turbines) to extend their turbine valve testing intervals frcm weekly to monthly.
SYSTEM DESCRIPTION The main turbine is a tandem-compound unit, consistent of one double-Qcw high pressure turbine and three double-fLow lcw pressure
- turbines, rmning at 1800 rpn with 44 inch last-stage blades.
Exhaust steam from the high pressure turbine passes through two moisture separator/reheaters (two stage reheat) before entering the lcm'ressure turbine inlets.
The exhaust steam from the three 1cw pressure turbines is condensed in the main condenser.
The generator has a hydrogen cooled rotor and a water cooled stator.
It is a three
- phase, 60 cycle, 25,000 volt, 1800 rpn unit rated at 1,230,000 kUA at
.975 power factor.
Steam is tt ansported frcm the reactor by four main steam lines and flcws through the turbine stop valves and contr ol valves to the high pressure turbine.
The steam lines are combined upstream of the stop and control valves.
The turbine bypass valves are located upstr earn of the turbine stop valves to permit steam bypass to the main condenser during transient condi ticns.
~
~
The turbine generator 's equipped with a dig'tal elect ohydraulic (DEH) control system.
The DEH system consists of an e3.ectronic governor using solid state control in combination with a high pressure hydraulic system completely independent of the turbire lubricating system.
The high pressure fluid supply is from a dual punp system in which one punp is a backup for the other.
The system includes electrical control circuits for speed control, load control, and valve positioning.
The turbine control system includes an overspeed trip mechanian, steam adnission valves, emergency stop valves, crossover intercept
- valves, and an initial pressure regulator.
There are four methods of turbine overspeed control protection.
They are:
a.
Governor (DEH) b.
Overspeed protection controller (OPC) c.
Hechanical overspeed trip meehan'n d.
E3.ectrical overspeed trip.
The governor, digital electrohydraulic contro3.
system maintain the turbine speed within 2-3 rpn.
The speed is maintained as long as the load demand does not exceed the capability of the turbine generator tnit.
The overspeed protection controller's primary fLnction is to avoid excessive turbine overspeed such that the turbine trip is avoided.
At 103$ of rated
- speed, the OPC solenoids open, closing the governor and intercept valves to arrest the overspeed before it reaches the trip setting of 111$ of rated speed.
After the turbine coastdown to synchronous
- speed, the digital system takes control and maintains the turbine generator at synchronous speed.
The turbine generator is then ready to be resynchronized.
Zf the turbine accelerates to 111$ of rated speed, the mechanical overspeed trip mechanian trips the turbine.
Tripping is accomplished by outward movement of the mechancial overspeed trip mechanian weight due to h'gh centrifugal forces caused by excess've turbine speed.
The mechanical trip mechanian cames the h'gh pressure hydraul'c trip fluid to be released to the dra'n.
All of the
steam valves will trip closed the! eby excluding all steam from entering the turbine.
The turbine speed is thereby maintained belcw 120$ of rated speed and the unit will coastdown to turning gear operation.
In addition an electrical overspeed trip, set at approximately 0
RPH higher than the mechanical overspeed will energize the solenoid trip which also dumps the high pressure hydraulic trip fluid to drain.
The results are the saae as the mechanical overspeed trip.
This setpoint differential permits each trip device to be tested separately.
HfALUATION OF TURBINE CVERSFEED There ar e three turbine overspeed cases of increasing severity which may occur as a result of equipnent malfunction or failure.
They are, design overspeed, intermediate overspeed and destructive overspeed.
The events leading to each of the overspeed cases are described belcw.
'Ihe turbine speed will reach design overspeed if:
A.
During normal operation load is lost, the output breakers open and a
turbine trip does not occur at even onset.
B.
Both the speed control and overspeed protect'on systems fail to close at least one or more governor valves or one or more interceptor valves.
C.
The emergency trip system functions properly and 'nterrupts the steam Q.cw into the turbine.
'Ihe conditions that lead to 13(5 of rated speed, given a full-load system separat<<on are:
A.
All throttle or governor valves are closed before design overspeed is reached.
B.
One or more steam l<<nes from the MSB's to the LP turbines remain open after the unit trips.
The turbine speed may reach the destructive overspeed
<<f the fallcwing events occur simul taneously:
A.
System separation with suffic<<ent steam supply into the turbines, e.g.,
this can happen if the load is lost and the breaker opens during normal operation, and B.
A combination of failures in the overspeed protection and emergency trip
- systems, causing a high pressure turbine <<nlet to be kept open.
It should be noted that for any overspeed event to occur, a system separation is necessary, that is, loss of load accompanied by or due to opening of the generator output breaker.
Otherwise, the events which result in the occurrence of the overspeed conditions are clique.
This necessitates evaluating each of the three overspeed occurrences separately.
WCAP-10161 docunents a probabilistic evaluation of turbine over speed for a BB296 unit with DEH.
The methodology and the fault tree analysis and results and the data used to quantify the fault tree are d<<rectly applicable to the
'urbine found at Washington Nuclear Project - 2.
The results docunented
<<n t¹s report wi<<l1 be used to discuss design and <<ntermediate overspeed.
0
~\\
WSIG-3-P docuaents a probabilistic evaluation of destructive overspeed for a BB296 unit with DEB.
The contents of this report are directly applicable to Washington Nuclear Project - 2.
The results docunented in this report weal be used to discuss destructive overspeed.
Finally, Westin@ouse Steam Turbire Generator Division has prepared a report specific to Washington Nuclear Project - 2 describing a turbine inspection program ba ed upon turbine missile generation probabilities.
The results docunented in this repor t, Uurbine bkssile Re~rt, Results of Probability Analyses of Disc Ruptur e and N.ssQ.e Generation",
- Parch, 1981 will be used to discuss design overspeed.
MCAP<<10161 docunents two methods for calculating design overspeed probability.
'The first method utilizes a fault tr ee evaluation and gave consideration to the impact of varying turbine valve test intervals.
This method results in a design overspeed probability, using a 95$ confidence bound, of 4.7 x 10 and 5.3 x 10 per system separation for weekly and monthly turbine valve testing, respectiv ely.
'Ihe second method of calculating the probability of design overspeed that was
- used, relied on operating experience.
Based on operating experience Westinghcuse has estimated the probability of design overspeed to be 3.2 x w3 10 per system separation using a 95 percent upper confidence bound.
This calculated probability value is believed to be conservative since Westin@ouse is not aware of any occurrence of a design over speed event in a Westin@ouse nuclear turbine which was camed by failure of turbine inlet valves or the control system.
MCAP-10161 utilized a fault tr ee evaluation with consideration for turbine valve test interval to determ'ne the probability of 'ntermediate overspeed.
As
~ ~
docuaented, the probability of ~~ermediate overspeed using a 95$ confidence bound was 5 x 10 and 1.1 x 10 per system separaticn fcr weekly and monthly turbine valve testing, respect'vely.
WSIG-3-P utQized a fault tree evaluation with consideraticn for turbine valve test interval to determine the probability of intermediate overspeed.
As docunented, the probability of destructive overspeed using a 95$ confidence
-8
-8 bound was 2.8 x 10 and 7.8 x 10 per system separation for weekly and monthly turbine valve testing, respectively.
MISSILE GENERATION PROBABILITY This section evaluates the potential for missile generation assuning a turbine overspeed has occurred.
Such an assessment requires consideration not only of the likelihood of turbine overspeed, but also of the cond'tional probability af missile generation.
The entire analytical operation can be conveniently expressed in the form of the follcwing equation:
P=P1xP2 P1 repr esents the estimated probability of turbine overspeed, P2 represents the conditional probabD~ity of missile generation and P represents the absolute probability that a turbine missile wQ.1 be generated.
Missile generation probability for each overspeed event described is calculated in the follcwing paragraphs.
The sun of the three probabilities indicates the total missile generation probability per system separation.
l t
l
The report, Wurbine M.ssile Report:,
Results of Probability Analyses of Disc Rupture and Missile Generation" establishes the probabilities (P) of generating missiles at design over speed for various lm pressure tur bine rotor inspection intervals.
UtQ.izing the design ov erspeed missile generation probability for a five year inspection interval obtained from Table IV Output Sumnary, a
m'ssile generaticn probability (P2) of 5.2 x 10 for the design overspeed case can be obtained.
This value was arrived at by converting the five year probability to a yearly probability and then calculating P2 as described in MCAP-10161-.
'Ihe probability of design overspeed has been shown to be 5.3 x 10 ut&ming the fault tree approa&
and assuning monthly testing.
Using the P2 value fcr design overspeed obtained above results in a probability of generating a
missile at design overspeed of:
P=P1xP2
=5.3x10 x5.2x10
=2.8x 10
'Ihe probability of intermediate overspeed has been shown to be 1.1 x 10 per system separation assuning monthly turb'ne valve testing.
As discussed
<<n MCAP-10161 no detailed analysis has been performed to determine the conditional probability of missile generation for the intermediate overspeed case.
It is
- believed, however, that this condition probability 's at least one order of magnitude lever than the condition probability of generating a missile at
destructive overspeed.
That is, P> -.1.
Therefore, the probability of generating a missile at intermediate overspeed per system seperaticn is:
P =
P1 x P>
=1.1 x 10 x.l
=1.1 x 107 The probability of destructive overspeed has been sheen to be 7.8 x 10 per system separation assuning monthly turbine valve testing.
The conditional probability of generating a missile at destructive overspeed is assuned to be 1.0.
Therefore, the probability of generating a missile at destructive overspeed per system separation 's:
P =
P1 x P>
= 7 8 x 10 x 1.0 "7.8x 10 Total missile generation probability is the sun of the missile generaton probabilities for each turbine overspeed case.
The total missile generation probability per system separation assuaging monthly testing then is:
T" DSO lO DO
= 2.8 x 10
+ 1.1 x 10
+ 7.8 x 10
=3x10
According to published NRC guidelines in Standard Review Plan Section 2.2.3 and Regulatory Guide 1.115 the probability of unacceptable damage frcm turbine missiles should be less than 1 x 10 per year.
Historically, as discussed in Regulatory Guide 1.115 and in the paper,
~Probability of Damage to Nuclear Ccmponents Due to Turbine Failure" by Spencer H. Bush, this value has been segregated irto two probabilities with values of 1 x 10 and 1 x 10 The value of 1 x 10 represents the probability of generating a missile.
The
<<3 value of 1 x 10 represents the probability of a missile striking and damaging cr itical components.
Historically, then, acceptance criteria for the probability of generating a turbine missile is 1 x 10 per year.
More recently the NRC has been considering limiting the probability of turbine missile generation probability to 1 x 10 for turbines with the rotor axis located parallel to plant structures as in the case with Washington Nuclear Project - 2 (NRC letter to R. L. Ferguson from A. Schwencer dated March 16, 1983, subject: Turbine Maintenance Coamitment for MNP-2 Turbine Missile Issue),
The total turbine missile generation probability per system separation
-6 calculated above is 3 x 10 To obtain the yearly turbine missile generation probability the value of 3 x 10 must be multiplied by the number of system separations experienced per year.
MCAP<<10161 assuned 3 system separations per I
year.
Using this value, the total yearly turbine missile generation
-6 probability is 9 x 10
~
This total yearly turbine missile generation probability is less than the historically accepted value of 1 x 10 per year and less than the more recent r equirement of 1 x 10 per year.
The total yearly turbine missile generation probability assuning weekly
-6 testing is 7.6 x 10 The increase in turbine missile generation probability from weekly to mcnthly testing then is anall and considred acceptable.
STATISTICAL HfALUATION OF VALVE TESTING INTERVAL AND VALVE FAILURE MCAP<<10161 and HSING-3-P contain a statistical evaluation of valve test interval and valve failure.
The statistical comparison considered valve failure rates
for different turbine valve test intervals.
The conclusion reached was that statistically there can be shown no dependence of valve failure rate to the valve test inter val, or, valve failure rate
<<s independent of valve test interval.
This conclusicn is supportive of the same conclusion reached by evaluating valve failure mechanians.
WCAP-10161 concluded that considering known valve failure modes the required periodic testing cannot influence valve failure rates since it does not readily identify failure precurmrs.
In sugary, it is expected that less frequent turbine valve testing wQ.l not adversely impact turbine valve reliability.
CONG.USION As a result of the evaluations performed and docunented in WCAP-10161 and WSIG-3-P Westinghouse has revised the reccnmended turbine valve test interval for BB296 units from weekly to monthly.
Westinghouse further reconmends that at a minimun plants with 88296 units revise plant technical specifications to allow monthly testing of turbine valves.
Th<<s recomnendation is based on the high reliability of the turbine overspeed and trip system which has been demonstrated by plant experience and which is supported by the foregoing evaluaticn.
Additionally, it has been shcwn that the probability of generating a missile from turbine overspeed is acceptably lcw and satisfies all publ<<shed acceptance critet ia.
REFERENCES 1)
WCAP-10161, "Evaluation of Impact of Reduced Testing of Turbine Valves",
September 1982, Westinghouse NES.
(Proprietary) 2)
MSTG-3-P,. ~Analysis of the Probability of a Nuclear Turbin Reaching Destructive Overspeed~,
July 1984.
(Proprietary) 3)
"Probability of Damage to Nuclear Ccmponents Due to Turbine Failure",
by Spencer H. Bush.
Taken from Nuclear Safety, Vol. 14, No. 3, May-June 1973.
4)
~Protection Against Low-Trajectory Turbine M.ssiles~,
Revision 1, July 19?7.
5)
Mashington Nuclear Project - 2 Final Safety Analysis Report 6)
"Turbine Missile Report, Results of Probability Analyses of Disc Rupture and Missile Generation~,
Revision 1, March, 1981, Nestin@ouse Steam Turbine Generator Division.
(Proprietary)
l.