ML19322C339
| ML19322C339 | |
| Person / Time | |
|---|---|
| Site: | Rancho Seco, Crane |
| Issue date: | 03/30/1978 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML19322C337 | List: |
| References | |
| NUDOCS 8001160877 | |
| Download: ML19322C339 (4) | |
Text
.
[ SAFE {EVM.uA,T,I_05]
~
PMCHO SECO C00LD0'AN INCIDENT OF 3/20/78 Ba t,ka round A joss of a;non-safetyzgrade i$strument power ~supplyicaused reactor and w--,~-,-~
.n ;..
tu rbine trips!ati4 i25_ A.ff[.oniMarch,20,l 1978. E As:.a Jesul t ofJoperatori Jaction=in restarting;t(he mainifeedwater: pumps ;plus. actuation,off thei bsafety injectionisyste the Er[ea_ctoFicoolant.. temperature idropped. tf
~m about 280'F in 75 minute's.
The cooldown during a one hour period was about 305'F, dropping from 590*F at 4:35 A.M. to 2d5 F at 5:35 A.M.
Because the rate of cooldown was more rapid early in the transient, the maxim;m cooldown rate was 470 F per hour for the first 30 minutes (590 to 355*F at 5:05 A.M.).
Pre'sure in the reactor coolant system remained s
abwe 1400 PSI throughout the event.
The rapid cooldown rate violated t'he Technical Specifications for primary system cooldown rite, and the pressure-temperature limits were also vio-lated.
In addition, it appears that actual Appendix G limits were exceeded i
very slightly, although this depends on details of the analysis.
The j
technical specification pressure-temperature limits for cooldown conditions i
prohibit a pressure over 1400 PSIG at a temperature of 280 F.
According i
to chart records, the lowest temperature the reactor coolant got to was l
280*F, but the pressure' was about 2000 PSIG at that time (about 65 minutes into the transient).
Therefore, they violated their technical specifica-tion by 600 PSIG.
~..
As a point of interest, 280*F is the temperature where a step change in f
pressure is permitted.
- Between 280 F and 185 F, the maximum permitted i
pressure is only 550 PSIG.
If the operator had delayed action to stop the
.j cooldown only slightly, the temperature would surely have gone below 280 F;.
i and he would have violated the pressu.re limit by 1450 PSIG instead of only j
f00 PSIG.
4
)
S3bcock & Wilcox Ana_lvses i
!......B&KperformeNffsctuFeinecha'nics analysii for the reactor vessel in i
general accoFdanc5 ~with~th'e' methods in Section XI of the code, but using some of the assumptions in Section III, Appendix G.
They concluded that s!
even if there were a flaw 1/4T deep in the worst weld in the beltline region, (Appendix G requirement) it would not have initiated rapid fracture, They also performed an analysis for the higher stressed nozzle region, with i+
i the same conclusion.
The c'- itions considered to be limiting by B&W were those vihen the pressure t
was 21. ; PSI, and the temperature was 468*F.
Using their calculations for
'l
. eat transfer, and code methods for calculating stress intensity factors, they arrived at the following for the beltline weld.
"~
s f
E j
8001160 [77
\\,
ce-M
2..
KI Pressure = 64.45 KsiM 1
K ? Thermal
= 45.47 Ksi/ili I
]
Their analysis considered "emeroency". xiditions where no safety factor
++
i is required on pressure stress. It s's that the KI applied, 110 Ksi M,
5L.
~
is well below their assumed toughnes (on the upper shelf) of 200 KsiM.
f.5.T.
l Even.considering the more ccnservative Arpendix G value of 170 Ksi/in 95 ;
for upper shelf toughness, the sa4ty factor is still 1.5..
=;.
1
~L The temperature used in this analysis, 468 F, is certainly well above l
the minimum upper shelf temperature of even the irradiated limiting weld a
metal in the beltline.
A i
- .1.
^
Independent Staff Analysis
+
4 We have performed an Appendix G analysis using the stress intensity factor t.
values calculated by B&W.
There are two important differences between the " Emergency" condition analysis and the Appendix G analysis.
These are a.
A safety factor of 2 must be applied to pressure stresses.
b.
The. maximum allowable upper shelf toughness u 170 KsiM to be considered is then (2)(64.45) + 45.47, or 174.37 KsiM.,"
~
The KI which is above the maximum permitted.
_.?
We calculated the maximum pressure that would be permitted by Appendix G with the high thermal stresses present to be about 2000 PSIG,so they violated Appendix G by only about 100 PSIG, by this analysis.
.We have also performed independent analyses to determine if a more severe
^ "-
state would have occurred at longer times into the transient, and also to check their calculations for thermal stresses and stress intensity factors.
For this analysis, we also considered the most severe radiation damage i" 5 that we could assume, using the " worst" reported chemical analysis of the i ""
limiting beltline weld.
[F.".T e.
i p
1
- i=
l We performed these calcul.ations for 1/8T as well as the usually assumed 1/4T deep flaw.
In this case, the 1/4T flaw was always more severe.
With F"g - j a steeper temperature gradient, and more radiation damage, shallower flaws
=E could be more s,cVere.
Even with our most pessimistic predictions of radiation damage, the temper-j ature of the vessel at 1/4T remained above the minimum upper shelf temper-
~
ature, and the upper shelf toughness was over 50 ft. Ib.
Therefore..the i
upper shelf toughness value of 170 r,si/in given in Appendix G is appropriate f
for the analysis.
'4
}
- /
e77 Our calculated worst case occurred at 47 minutes into the transient when the pressure was 2030 PSI and the 1/4T temperature was about 410*F.
The pressure allowed by Appendix G was determined using our calculations for thermal stress and stress intensity factor.
This allowable pressure M
was 1560 PSIG, so at 2000 PSIG, our calculations indicate they violated Appendix G by 440 PSIG.
.... Y._
Our' calculations also confirmed that even using the conservative upper
.... f shelf toughness of 170 Ksi6, and' assuming an extremely improbable
@=
1/4T deep flaw the safety factor was still about 1.26.
[
Discussion
- q:
,,,,,_m--y.,
M though the! actual. safety-implications ;of-this;particular; transient _.
..... r
%were minimal,; this_is only.true because.it occurred -veryjearly infplant
?
p,li.fe,. :We-strongly: recomendcthatipositive steps.be taken to.. prevents I
a trcnsients of-this kind, and thatathe_genericiimplications be reviewed; nE promptly.
~~
=.
A prelipinary assessment'of the potential damage to the SMUD Steam Genera. tors during the 3/20/78 rapid cooldown transient at Rancho Seco has been completed by Babcock & Wilcox.
Based on that assessment, B&W feels that the steam generators have not been adversely affected.
This evaluation is based on an analysis of the stresses imposed on the
- =
tubes, tube-to-tubesheet welds, and head-to-tubesheet weld.
To assess i
whether or not the tube yield strengh was exceeded, an estimate of average E.[ l shell temperature vs. time was made.
From this calculation, the maximum
~'
tube-to-shell temperature differential (AT) is 170 F at 0515 hours0.00596 days <br />0.143 hours <br />8.515212e-4 weeks <br />1.959575e-4 months <br /> (50 minutes into the transient).
1
..jE A structural analysis was performed by the Component Des:]ner which demon-
.l 4 strated that a 200 F AT is acceptable without exceeding tube yield strength i
or without inposing unacceptable stresses in any of the welds and further p.if i meets the requirements of ASME Code Section III.
E:::.::
1.= =
In addition, B&W has evaluated the fatigue usage factor and they are con-
- N5l,
[ji==
vinced that the usage has not been appreciably increased.
This will be fE i
determined and reported in the final report of this incident.
..m
_ Conclusions sTE37'
- .=
1.
During the transient, the technical specifications on cooldown rate '
a=
and pressure-temperature limits were exceeded.
B+
I h
l J
a l
l u.
i
o'
_4 I
2.
Appendix G limits were exceeded.
3.
Because the vessel had limited service and radiation damage, appro-priate limits for emergency conditions were not exceeded.
~
4.
We conclude that the reactor vessel was not damaged by the transient to the extent that it reduced its expected service life.
5.
Positive steps should be taken to preclude similar transients, and generic implications should be reviewed.
6.
We conclude that B&W considered the appropriate regions of the steam generators and accept their findings.
e l
+=-
\\
- .i
!b
,; f. A p +:
- 3.
- 2: -
O m
W