ML19248C222
| ML19248C222 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse, Oconee, Crystal River, Rancho Seco, Crane  |
| Issue date: | 10/29/1980 |
| From: | Taylor M NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| To: | Fabic S NRC OFFICE OF NUCLEAR REGULATORY RESEARCH (RES) |
| Shared Package | |
| ML19247E428 | List: |
| References | |
| TASK-2.K.2.13, TASK-TM TAC-45202, NUDOCS 8105110190 | |
| Download: ML19248C222 (16) | |
Text
REFEREEE 1
/ *. " ' g f
J, UNITED STATES
- &@f h y yp v.(
NUCLEAR REGULATORY COMMISSION E
wemusTow. o. c. cosss
$,e!
a....
October 29, 1980 MEMORANDUM FOR:
S. Fabic, Chief Analysis Development Branch Division of Reactor Safety Research Office of Nuclear Regulatory Research THRU:
Gordon E. Edison, Section Leader Systems Development Section Division of Systems and Reliability Research Office of Nuclear Regulatory Research FROM:
Merrill A. Taylor Systems Development Section Division of Systems and Reliability Research Office of Nuclear Regulatory Research SJBJECT:
' INSIGHTS ON OVERC00 LING TPJ.NSIENTS IN PLAhT5 WITli THE B&W NSSS At our September 12, 1980 meeting SAB/SRR agreed to survey the B&W LER file for insights on actual overcooling events experienced by the B&W NSSS.
It was felt that information on actual plant transients bould be helpful in validating BUL calcLlations and in exploring potentially more severe overcooling transients.
Subsequent to the 9/12 meeting,4 torkscope with specific tasks for the RSR program on Analyses of Overcooling Transients was mutually agreed to. Tasks I, II and III (Phase I) were to be accomplished by SAS/SRR.
This memo is intended to fulfill these specific tasks.
It is r.ecognized towever, that further dialogue and DSRR inputs will likely be needed throughout the Phase I =rh Table 1 su=arizes the results of the LER survey.
Figure 1 illustrates the severity of the transients experienced.
As discussed in the following sections (I, II and III), we have recommended seVeral of the T:Dre severe events for ENL henchmark use.
Section IV mentions other events that may result in greater overcooling and some LWR failure experiences that might be viewed as accident precursors. We have also attempted to give you a rough perspective on the frequency of various overcooling transients.
We caution, however, that such estimates are made on a statistically limited base of experience with the B&W plants and they are rot of high precision. These should, therefore, be used with recognition that considerable uncertainty may exist around such estimates.
ffWI/M fD
S. Fabic I.
LER SURVEY RESULTS A survey of the B&W LERs was undertaken to identify the more significant overcooling events experienced. The LERs used covered a time 3
period through early 1979 when about 22 reactor years of operating experience had been accumulated by the B&W plant. designs. Table 1 sumarizes the results of the survey.
Other experiences have occurred after mid-1979 (e.g., Oconee 3,11/10/79) and these are also included in Table 1.
A total of roughly 28 reactor years is the present base of B&W operating experience.
Approximately 1/2 of this experience has been accumulated by the Oconee units.
ere identified that have exceeded the cooldown rate Fifteen eventg w/hr) sat forth in the technical specifications.
limits (v100 F This would suggest a frequency of roughly 0.5 events per reactor year exceeding the technical specification limits.
Review of these events indicate that about 60% (9 cvents) occurred prior to the plant having attained conraercial operations. This msy reflect the plant burn-in/ tuning-up experiences usually seen.
In regard to the LERs, these set forth little detail of the type that sheuld be of interest to the BNL analyses.
In a few cases, we have dug c;t additional details from the docket files. Examples are also being sent (separately) for your review.
(We could perhaps ask the licensees for recorded plant data on the transients if you think rare information detail will be needed by BNL. Let's discuss these additional needs at your convenience.)
BasedontheLERsurvey,five(5)overcoolingeventswereidentffied that should reflect a spectrum of cooldown rates from about 120 F/hr g
through 300 F/hr.
In terms of decreasing severity, the events are:
1.
Rancho Seco - 03/20/78 2.
Oconee il - 05/05/73 (occurred prior to comercial operation) 3.
Crystal Rivet !3 - 03/22/77 (occurred prior to comercial operation) 4.
Oconee il - 12/14/73 5.
Oconee !3 - 11/10/79 These events are illustrated in Figure 1 and briefly discussed below.
II. RECOPMENDED BENCH? ARKS 1.
The Rancho Seco event of March 20, 1978 is believed to represent the rest severe (and prolonged) overcooling tiansient experienced 1 ' sed on NRC Gray Book data on accumulated number of critical hours. Actual Etime from start of comercial operations is larger by less than a factor of two.
S. Fabic to date (m300 F/hr) and it is recommended as an important benchmark for the BNL analyses.
Not only did the Rancho Saco 0
event greatly er.ceed the cooldown rate limitations of ~100 F/hr specified in Technical Specifications, it also appears to have exceeded the pressure, temperature limits specified therein for the RCS.
These RCS limits are currently based on RPV irradiation of only 5 effective full power years.
2.
The 0conee 1 eve t of 15y 5,1973 involved a high initial rate 6
n (10 lb/hr @ 100 F) of delivery by the main feedwater system while the system was under manual control. This event occurred prior to commercial operation and is believed to be the reason for Duke installing a safety grade high level trip for the main feedwater system - this high level trip being independent of the Integrated Control System (ICS). H is event may also be of interest to the BNL analyses because of the initial high cocidown ran rate experienced and the fact that the RCS pressure diminished to N1330 psig and shrinkage caused a loss of pressurizer coolant.
The Crgstal River 3 event also involved a high initial cooldown 3.
(~164 F/15 min) rate but this stemmed largely from excessive steam relief occurring when the atmospheric dump (and possibly turbine bypass) Yalves remined partly open.
This event may also be of interest to the BNL analyses because of the system response to high steam loads. This event also occurred just prior to commercial operations.
4.
D e Oconee 1, December 14, 1978 event involved overcooling due to OTSG fill levels being specified higher than was found to be needed.
Emergency feedwater was used to fill the OTSG to about a 95% level - a specification that subsequently was revised downward. A cooldown to ~1430 psig occurred resulting in an actuation of ECCS.
5.
The Oconee 3, ?bvember 10, 1979 overcooling event resulted from some delays in stopping the main feedwater flow (slow valve operation or setpoint errors) combined with inadequate secondaiy pressure control due to apparent turbine bypass valve malfunctions and the presence of high auxiliary steam loads.
Tne minimum RCS temperature, pressure conditions 0
reached were 420 F and 1650 psig,~ respectively.
These latter two events should be of lesser interest to the BNL analyses.
O O
~
S. Fabic IIJ. DISCUSSION ON NATURE AND CAUSES OF OVERrnDLING EVENTS The Blot plant experience indicates that the rare severe overcooling events have happened after the plant has experienced an undercooling event causing near depletion or dry-out of the OTSGs. These have stemmad largely from operator actions taken to reestablish feedwater delivery into the depleted OTSGs. The cooldown has also been made somewhat rare severe by the presence of high auxiliary steam loads and/or by high steam relief via open turbine bypass or atmospheric relief valves (e.g., 50% open).
liith exception of the Oconee 1. May S.1973 event (prior to commercial operations), the above overcooling transients have involved various power faults that influenced the response of the plant non-nuclear instrumentation and controls (i.e., NNI/ICS). Such faults have contributed to the loss of (or somewhat erratic response from) the steam and main feedwater portions of the plant. Undoubtedly, these faults have also contributed to some degree of confusion by the plant operators in their actions taken subsequently to restore feedwater and achieve stable plant conditions.
The Rancho Seco event included a number of factors that contributed to the severity of the cocidown and to the fact that the extent of the RCS cooldown was not fully recognized for a prolonged period of time (i.e., nel hour).
Lome of these factors include:
Power fault (short) that affected the response of nearly 2/3 of the NNI/ICS equipment, Essentially " dried-out" the OTSGs via loss of feedwater, Gagged P2R-PORV existed such that relief through PZR safety valve occurred when feedrater was lost.
(This factor has relevance to potential repressurization conditions.).
Confusion existed about the status of feedwater delivery into OTSGs largely because of NNI/ICS faulting, High auxiliary steam loads were present, Sergency feedwater design has an SIS actuation signal for starting unlike other E!JI designs (overcooling can be and was in fact made nore severe by this feature),
Initial overcooling occurred from human actions taken to reestablish nain feedwater flow.
Tnis challenged the SIS actuation signal and caused initiation of (1) 100% emergency O
I S. Fabic,
feedeter delivery of cold condensate to both OTSGs in the presence of main feedwater being deifvered into at least one of the OTSGs and (2) 100% high pressure infection (i.e., 2 HPI pumps) delivering cold EWST coolant to the RCS while an additional high pressure pump was also delivering to _the RCS in a rormal rode from the makeup tank, All r.ain RCS pumps were allowed to remain running to low RCS pressure-temperature conditions in violation of usual procedures and precautions.
Operators of the plant were apparently preoccupied with restoring NNI/ICS equipment and did rot promptly rectify the overcooling transient, although part of HPI and emergency feedwater delivery was secured during the overcooling transient. As a result, the operators did not realize until N1 hour inf.o the transient that the RCS temperature had decreased to 9285"F, Although not yet clear, it '
ssible that the NNI/ICS faulting also caused additional stea mds. This is so because the turbine bypass and/or atmospi.%1c dump valves are designed open to about 50% position as an expected null position on loss of power to the ICS.
(Not all NNI/ICS power was lost in this event however, and the particular response of the turbine bypass and atmospheric steam dump valves remains unknown to us at this time).
As mentioned above, most of the roderate to severe overcooling events have involved pawer faults of various kinds in the NNI/ICS equipment but were the result of human actions taken later into the transient to restore feedwater to the OTSGs.
I NUREG-0667 reveals that there have been 29 failures of the NNI/ICS in the B&W plants through about the spring of 1950.
Approximately 20 of these failures have resulted in a reactor trip while the plant was above about 30% power.
A feedwater transient was experienced in nearly all of these reactor trips.
About 6 of these events resulted in excess cooldown rates of the severity noted above and illustrated in Figure 1.
According to NUREG-0667 information, four (4) automatic actuations of the HPI were experienced during these NNI/ICS failures. These actuations could have resulted from RCS depressurization caused-either by a stuck open PIR-PORY or from the overcooling by the steam-feedwater conditions in the secondary side of the plant 1 See Table 4.2 of NUREG-0667
$. Fabic (these do not necessarily occur together). _This experience indicat,as a.fLtwexy n' Mrate ta vvcr.e-cw;r:.4A:LryxEr.:.5 LDL,33nts
.1pfligini.,fLactuate HPI) at_ rouchly_0,1 to 0 2, oer reactor vean If, on the other hand, dancho Seco were to be raten as the singular most severe overcooling event thus far egperienced in comercial operations, a frequency of roughly 3x10- per reactor year might be es tirated.
In light of the. post-TMI-2 and post-Crystal River improvements required of the B&W plants (particularly in relation-ship to the !MI/ICS failures and to the human training in response to such feedwater disturbances), the frequency of overcooling events as severe as Rancho Seco ray have been considerably reduced.
The Rancho Seco event could conceivably have been rade even more severe through prolonged inattention to the OTSG heat removal, by mainfeed delivering to more than one of the OTSGs, or.by failure of the humn to partly secure the emergency feedwater and HPI delivery after the such was actuated by the SIS signal. The probability of such additional errors M speculative on our part, but for purposes of establishing RSR analysis priorities for B&W reactors, y_a,L, s3tqq an overall frecuency of ev3x10-3/RY for the Rarcho Seco overcoofi q_1p_fave been mace more severe Inan actua iy experienced.
t IV. CVERC00 LING EVENTS I43RE SEVERE TMN REVEALED BY B&W EXPERIENCE As mentioned above, the Rancho Seco event might have resulted in somewhat greater overcooling largely through human inactions.
There also exists the possibility of other severe cooldown events beyond those revealed by B&W experience.
Examples would include a rupture of large rain steam (or feedwater) piping - perhaps with additional coincidental failures taking place. At this time, we do not have very good estimates on the frequency to be associated with these more severe overcooling events. More work would b2 needed to derive such frequency estimates and these could vary somewhat from one plant design to another. We know, however, that the world experience wiyh various comercial LWR designs is roughly 1000 reactor years with ai; cut 1/2 of this being U.S. experience. To our knowledge, there has been no large rupture of main steam piping during this comercial experience which suggests a frequency of large pipe ruptures of the order of 10-3 per reactor year or less.
Failures coincidental with rupture of large piping would either be caused by comon interactions or result incependently from the rupture.
Some examples of coincidental failures that might result in greater evercooling would be: failure of the ICS, failure of rain steam stop valves, failure of the steam line rupture matrix, uncontrolled emergency feedwater flow, rupture of steam line interacting and causing failure of another, human error, etc.
In the overall, 1This cstimate reflects actual time from start of cor:rrrercial operations. The accuoulated number of critical hours should be about 50-60% smaller.
S. Fabic events involving coincidental failures vould be expected to be at )
lower frequency perkps by as much as an order of tagnitude.
Bush has made a survey of incidents pertinent to the reliebility of piping in LWRs and has considered these incidents as possible precursors to piping failure.' failure statistics confinn 10~j basic {inding of his was to 10' pipes,with hi Earlier, WASH-1400 had estimated 10~gher rates as size decreases."per reactor year as a redian value large piping (,t 6" dia.) gging an,grror spread of 10 up and down about the median (i.e.,10 to 10
).
For purposes of the RSR analysis tasks, we believe it reasonable to assume a frequency of 10~' per reactor year as being applicable to rupture of large steam piping. Given a rupture of large size steam piping, the plant operators muld be faced with events not yet experienced.
As the Rancho Seco and TF.I-2 incidents muld suggest, the rossfbility of human errors being conaitted is large when the operators are faced with a new experience or one for which training and procedures are lacking.
Hindsights from the TMI-2 accident muld suggest roughly a 30" chance of error existed given the unanticipged sequence of events. Ee,,39uld therefore succest us,e_ of ~~MID
/PY n.s_.in e tated freoue govercoohno caused bra 1arce steam pioino ruoture
.olnc,en;y c
idental with other failures.
V.'
SUMKARY The EEW experience reveals a spectrum of overcooling transients including several that have been rather severe.
This experience suggests that as the plant equipment and operators become " tuned-up" prior to commercial operations the frequency of overcooling transients diminishes. All of the more severe overcooling events that occurred during commercial operations appetr to have some coronality in that they involved various faults in the instrumentation and controls (NNI/ICS) causing disturbance and/or loss of feedwater to the OTSGs.
The severe overcooling events resulted largely from human actions to restore feedwater to depleted or " dry" OTSGs.
In these cases, it is likely that the human faced some degree of confusion about equipment status because of the faulted NNI/ICS.
Therefore, he cannot, in our view, he held totally accountable as the root cause for the overcooling events.
We have attempted to give a perspective on the frequency of overcooling events experienced and our jsdgments on the frequency of events that r.ay result in greater overcooling. These frequency estimates 1 See raper IAE.A-SM-218/ll, Reliability of Piping in Light Water Reactors, S. H. Bush, USNRC, ACRS, Washington, DC, USA (International Symposium on Applications of Reliability Technology to Nuclear Power Plants, Yfenna 10-13, October 1977).
- 5. Fabic are intended to assist RSR calculational priorities and sh>uld not be taken to be of high certainty and precision.
For RSP, convenience, these estimates are surcarized in Figure 2.
A Y[/aT' [M M -
Merrill A. Te"ylor, Systes Develop ent Section Division of Systems & Reliability Research Office of h'uclear Regulatory Research
Enclosures:
1.
Table 1 2.
Figure 1 3.
Figure 2 cc:
G. Edison F. Pawsome R. Derr. orc <
c L. Shao J. Strosnider M. Vagins C. Serpan P. K. Niyagi J. Jenkins W. Yesely M. Cullingford D. Basdekas O
e O
0 ENCLOSURE 1 Summary of Overcooling Events in Plants with the B&W NSSS I
e O
Qn'
)
yA d
f c(l s wnn p
co n
oa 4
owa rd.so
.i eerw Wl o y eon)d o rre tFfdi Klt at al reet Mgntronl t uM l s i c no ewha 1 e oooo hop 0h adil1 t opr
~ md crnb ego 3 weeo rps E n t
Ahi c0 2
rt pmn e
on aend th 3
r rt ai v%mw.
t roi d3 eoraee n0t od
,ti1 ht et st r i4ade a.yucd l c dsapN t ct s sh N
dl s es a
af l
/
f d on eel set paevnnesr f easayF o
noe l
i p c r li glG f
or ang taci h ii p a o i S eo swia0 sa r
wrt vpsr h
nd e gl 4 s
dAe t ar t eehTe ont pil1 ose p
a u m r i >n l ap x
d mesedtt r
i t
nuoz wp.e eyopss ao
.u a W i p( u o
i%
swb sl pl tumapead
) se l F m af.
t rr0s pl iaapyr e0Fev oM rl o n ot5a eceNw ru m b u p e f,-
re s
7eeh e
dt 3
t nti sif 0 pl if t v g gh ar esr d0 otaeint l an s
smaonetl e1 mr e
u wlhi o eioe p
tti r ah ue spod l
srbtt mss b p ri s 6 m z uib e% doe u
r u
ua er oti ii a r A e5 noh b eudn pwt uuetil rnr ct f9act lt ei r
li nm odThcnbhiu
.y~l l s en ahd e Wrn t aia/ms s
Fo t
e tb s
aurc evi t w n p5 n
y Mtdu.drrsl ae nl enovo i
ao1 g
r ciopn eg r
gpwet c n vS
/
s og in5 CrsF i
a dad oant e0np i
s m
ne do ifr1 i sdprs a
CI og f
r eGl 5
A t v4 e
m aRrell a s~d o
sa% mSld3 ocl 6 D
u ei ae S
r t rtussa( e u y0 c i e1 ft aa1 S
e. ad e n a e w t s oh7 dfc" o
ev e
u rr t
S t
t rw haet sl s i c N p,l rt sd se pu S
n sedGt mruweuo imd swemo l
e owcS cnot sl ui il i
n oohub v
coeT nnnil ae pwt rih W
E 3pfOiiimf rra SsatfTpa L pT d a D
s e
t t
n a
4 7
e D
7 7
v
/
/
E r
7 0
e 1
1 g
t
/
/
n t
1 3
i e
0 0
l L
oocrev O
e t
3 4
7 a
7 7
7 D
/
/
/
1 5
4 2
t 0
0 0
E n
/
/
/
L e
5 4
3 B
v 0
0 0
A E
T o
N
)
0
)
t 9
7 3
e 6
2 1
k 2
c 0
r o
0 5
e D
5
(
v i
(
d 2
R n
1
)
a l2 e
e a0 t
c c
t3 n
n n
s -
a o
o y0 l
c c
r5 P
O O
C(
2 3
1
r i
h t
t d
~
,e n
1 e
~
d s g
ne l
eh dWu c
rI -
n at n
l t rT s nF uh eP r nei u
aWi nu G
aE) s dt wll e areyen iFf Sas SW rriei o
v ubll i cme Cl s
.TFI dh ot rw p
o i
r I petOMP n/tt
,s lb m t s0ui aa
/
rs I
aF annn
%t o.
I epod e rua o 5ftd er i2 coo Nl l nhfI"0 e ne4 f g0 h
n I
ei 1 i o
m pd tl et Nie at oP0pebt uldi p
n o gt t
g hhs I 3 oz a
r er I
s idp ot e yaneu d n rwt atyn i
ub no aei o
wsbof nt nau.
tt up f rf o
i o( n gh c) a b
q 9d b
,l l uonl gt ao ct sir o
ed3isa e
da rso bc o
nansl ceon eeru l vt o nsV wn eic l c3isiuoeperee il nh wuWs i
whr uo/t altit rurm oaFct oTe b
2 awl ct ay t pp hao dc ynF p
v t o c
i aarb, a i
wvcl httirf u
rrf u a
t sl g
.o useai1 g
ud eedt gd eeoq tfd h a v v p0 0St itont rncnevp e
sen swl 1
0Cn l rbia aAireme oiar adf 1I e oA wt l uwesS ll e
etvno" N
u ah ld ng ocotuC ed v oeh aI sreo t n a
oq s
eeen.
oeW 1f t e g
.veuircs mc/
a t
miF nco t
s.
n a r e f ql er
,eeI wmaF
)
esMoy atb r i
el eot eyttbN ag
, soavl os(N met 0 em.
ruh cd w t
rn i
n artl wd sos /
aeoddb cwot ny ut c1 C
y_
f roa oe ah F
l spneurd r
sFf usc1 o
any~
c otf rrdc wsg pu a pS e e e a d g o
(
r bel n n0 ea%
p v e r pi y5 e
dr0n vil
.of a
t griwoe tCa2 rc0ei ve a2 o al t s
m snel ooi nD 1
7 rr rmaa sd utt nyersm i
n m
eit aPcr a
t an"t uacseritt e
u g
u tl tc e
l ant i
S li l p p
eu wa aro n
worua sgng s
gi m f
.l x
dio tro d ed t por nnan e
t nf soeae eesb nl aetd egesa bo eimi D
n i rn sm l sna ae pcesrsprt at d s l
i ua mnsmaiueumea s
nuso e
reat oss S
v uvrul a
harf u a a e e r a m a u ph oe oaao S
E Dotocaw Wct o Hpwt r d c e c pO t t r Ccwc S
N W
D e
t a
7 9
0 9
s D
7 7
7 7
t
/
/
/
/8 9
1 n
r 1
e e
2 1
3 0
v t
/
/
/
/
E t
4 1
3 0
e 0
0 0
1 g
l i
i i
looc e
r t
7 9
8 e
a 7
7 7
v D
/
/
/
O 5
5 0
t 1
0 2
n
/
/
/
3 e
4 1
1 v
0 0
0 E
E.
l BAT
)
)
)
o 2
2 2
1 N
1 1
3 3
3
+
3 e
f 0
0 0
k 5
5 5
(
c r
(
(
o e
D v
o o
o i
c c
c d
R e
e e
n
)
S S
S a
l2 a0 o
o o
t t3 h
h h
n s -
c c
c a
y0 n
n na l
r5 a
a P
C(
R R
R 4
5 6
7 i
~
Y g
R f
d
- e n
%O rno t
e sr i
0P oo p
a n c nou.
e dt 0 - ti g di s
hsm wn anss dl nea 1R ct n errhp wea goe mt se g
eieru Z aai st et m eveendi uAev n
shh rt mP ezl u
wiut ml t ril r rio arowpnoaset oe h
rl si uwt uc me ro cop esv waop paul a
ca n.
vl o ct ;c fh euc t
gl u y
l cx i pd po nyot thsr rcS nl qesrso e
inf c
eer s
d,Tse eaCoeed s a pi A ev weIl wsaaimv s
drae) r vidgp e
et it e as n
cn ro or a t b m pl u a
t dl av hnti0 uw
.p p
ou yipF.w reeeo anl3 doeeo.
a. shssb x s
luot rh oreo4 edsdt r re a
uC0es st a n
ttwo1 rt o u
9detddweaR5re ecl yya ca ulSd o 9 nwunn n
4ut gh cCch
~ a o n a a gi) r d u raota tWGr pi nb2uoen esecs hFS e nseRi mrif ed gMTvh i
r t
p meGir( ot ci,f om u
adn u
Ooc ee tns t m%
tSl u Fr udes a
od a
bct ner au5l aTot ne n o e pi p7 a s O o p0 cnr em h n "B a r sanuof c
rn c)%3 2
e~ee" aaeml e t
a r10(0s an q
v watd se" i
s et
,o wed ege it vdBe(5 5 nt oeNdnNio v
o u rest 1d isedt ti snh h a o e n" eves 1 a nh e rndbit wst so sl v
( o t oTlS ei un n
cn s( a tb l
l p l C w a d si g t ekr ieu mua
)
nal y evab r us iR omn an cdcof han
, oa t
d v
em ae
.f p
aemi anafoeTceh rcr n
nP d wa frreem a
her ob s
epgfee
.boi rv o
y oBfeoe erh e r
dt ru ecntt a hgpo C
r cTel pt dSw tt ore d
c usre n
roargey n
(
a il s
rCos s
t ondY a "A m
gfl ow oI pady cfonRr lt cnrl eidF s
m noeroy.
c wes a
i aOh n"
S sniet rnn e
vrena2 t f. C n e i loeawu n
u i
rtl rd s%
s e%t P/
I eR l
g S
se n
ae ru8t uI r5 cd F
neg no h
dosp rsi arcooir eo9iaP 1
i tN u ph 1 i
rt ec s
e ~ f pt 0 cH wiNn s
t euictl r ape
.sreedti o r e
n rsr iu l
mit sI uerraf m D
e coesexC pu ed e i
l i 1
f h o a r y a A t a w N c o n wlll rf aNavoeoei v
elhauuC pth nh Wt mtb o
S E
'DcpwDao Asawat S
S H
W B
e 3
t a
4 9
55 s
D 7
7 77 t
/
/
//
n r
2 5
78 e
e 2
1 20 v
t
/
/
//
'E t
0 1
68 e
1 0
00 g
l n
i loo e
c t
8 9
r a
7 7
e D
/
/
v 4
0 O
t 1
1 n
/
/
e 2
1 1
v 1
1 E
E L
BAT
)
o 2
N 1
3
)
)
)
t 9
7 7
e 0
6 8
8 k
5 2
2 2
c
(
o 0
0 0
D o
5 5
5 c
(
(
(
d e
n S
1 3
3 a
o e
e e
t h
c c
c n
c n
n n
a n
o o
o l
a c
c c
P R
O O
O 0
1 8
9 1
1
t c
e.
e-rt e
rid p
S en s
errel eu C
t ve eas h out a vt s
R nei h ea T csai rnG S
ass t rw G
sut i
t eti arcn' w
l n
c yS tai 5 I
ct T
So pea gntl n
rr nilld r
pai esdO
,l oyot i
uue et e f
e emt e no bm soaf n wad se dolh wt s e
itf e
oc ac rut o
el eh a
op psSwn
,f s d evodt rd et o eC i
g ef ul ra eh
% rRSs ngro gd a t mo duTedl 0
I in n
vn t
nn vnc 3t gS w s ay e on os disg i) oo ai
.l c ni
,l e
nhin tiWcse t
a c
echi tsF l s t nl vsI rnsr ctTd u pM l an noe n
tduw g
si e
h e
o acvWof hiDoso s0gvsp l
eFio d
d ar aw e.
2neas pl l t
e wwp s's. f d re d6il we e
fcl l s. o dnneer n1 n r
evdoao i e pl en ed g hee r
hvib si aivvr a 6, p n n r l ou tli ekt wl r n
rda c
r(oaio fron at eol eevnc e
lt neiuoo dv h pl
)
ws ao wd yw o a wh cl t c e
et si oug oeb oor ot ei pfh et t
p ndS pr l ce d
parr petyrs n
einI rdf rp
,sf oo ii b
o
% rtaS
%ue eo g2 tf rl r ms C
y 2 ea 0csmv n-moa t eed ea
(
r 2wl pf 2cuoOt il ut rs rt et w uio e oar c
tGm ee r
fss a
9sgr c
cf
.e tSid pv omauyg s
m r
u xeOl t a asn n
m eet n n
g u
t rr o
t oerrr hf at a
ceec i
at r Wn i
S au ri aireoo somu
.v s
dWot tuft c b
esunF u c
t eeFt a cassan r1 %ien r
soMt O
n l cMcu i usnri ec6 rsi nsi e
io at f t eae t y8t oa eiet eS S
v hroec hcrrpy f c tl r harahC S
E W pt rA W a pt ob A i '/ a c d TmpzTI S
ti W
t.
D e
t a
9 8
8 0
4 s
D 7
7 7
7 t
/
/
/
/
n r
8 8
0 8
e e
2 2
2 0
v t
/
/
/
/
E t
2 7
4 5
e 0
0 0
0 g
l n
i looc e
r t
8 8
8 8
e a
7 7
7 7
v D
/
/
/
/
O 2
9 5
3 t
0 2
2 2
.e n
/
/
/
/
e 2
4 3
4 1
v 1
0 0
0 E
E L
BAT lo i
te k
)
c 0
o 2
D 3
e e
s s
d 0
s s
n 5
e e
a
(
B B
t 2
s s
2 n
i i
a I
v v
I a~
a M
M l
D D
T P
2 3
4 5
1 1
1 1
Figure 1.
ILLUSTRATION OF APPROXDiATE SEVERITY OF OVERC00 LING EVENTS Events Exceeding Severe Overcogiing,
Rates to ~300 F/hr.
Tech Spec. Cogidown Small Overgooling ModerateOvegcooling liigh Initial Pmlonged Rates of ~100 F/hr.
Rate. 4110 F/hr.
Rate,2;120 F/hr.
Rate Overcooling I.
PRIOR TO C0"MERCIAL OPERATION 1.
Rancho Seco
-July 19'74) 4-Dov<xx>o, 7) 4110"F/hr*)
2.
Rancho Seco ~ October 1974)*
u,.m4 3
Davis Desse 03/25/78) p'xwg, 7) 4.
Oconce#2(01/04/74)
M E2221 ( ~140 F/hr.)
5.
Tiil-f 2 (12/02/78) e m m x6M4+ (?)
6 THI-f2(04/23/78) 7.
DavisBesse(04/29/78) t-arx e M*(?
LI 8.
Crystal River f3 (03/02/77)
(tim M A+
epwouga (~164"F/15 m 9.
Oconce il (05/05/73) g u,ca e (7, To
~1330 p II. AFTER COMMERCIAL OPERATION 1.
Oconce #3 (~ June 1975) a g:s:5k ( ~10l F t
2.
Crystal River #3 (04/16/77) gggggE3(~101g /hr.)
F/20 min) 3.
Rancho Seco (01/05/79 m mitM ( m120 F/gr.)
g 4.
Oconcef3(11/10/79) 9 ysrarA (~112 F/20 min)
Oconce il (12/14/78) )
g gj_fwQ-+ (?, to 1430 J 5.
Rancho Seco (03/20/78 6.
- i
~ ' < '
(es300Ffh to ~ 205 F
- Possibly includes 2 occurrences when the plant was at low power (~15%)
(?) Actual rate of cooldown not set forth in LER writeup.
f These cycnts involved various power faults affecting NNI/ICS response and main feedwater delivery.
Overcooling resulted largaly during human restoration of feedwater to the OTSGs.
~ " Recommended Benchmark Events.
I Figure 2. ' Estimated Frequency for Overcooling Transients All Events Experienced
<v0.5/RY Small Severity N7x10-2fgy Overcooling Events Experienced Transients in in Comercial x Moderate to Severe
~10~j/RY B&W Plants Opera tions
{
Severe N 3x10~2/RY (Rancho Seco)
Rancho Seco and rv3x10-3/RY Additional Failures Events Not Yet Experieved that 2
Large Rupture of ul0-4/RY Might hrsuit in
- Steam /Feedwater Piping Grer.ter Overcooling 2
,large Piping Rupture tv3x10-5fny and Additional Failurcs I
These estimates are. intended as en aid in RSR analysis priorities and should not be taken to be of high -
precision and certainty.
2 This freque:cy estimate assumes actual time of experience as the time from start of comercial operations and it may be larger by less than a factor of two if the relevant time of experience is measured by the accumulated number of critical hours.
~-
j i
/
t
~
s.