ML20041D489
| ML20041D489 | |
| Person / Time | |
|---|---|
| Site: | Big Rock Point File:Consumers Energy icon.png |
| Issue date: | 03/01/1982 |
| From: | Vandewalle D CONSUMERS ENERGY CO. (FORMERLY CONSUMERS POWER CO.) |
| To: | Crutchfield D Office of Nuclear Reactor Regulation |
| References | |
| TASK-02-02.A, TASK-03-02, TASK-03-04.A, TASK-2-2.A, TASK-3-2, TASK-3-4.A, TASK-RR NUDOCS 8203050351 | |
| Download: ML20041D489 (5) | |
Text
. ____ ___ _________ _ -___________
Consumels power David J VandeWalle 0Nf Nuclear Licensing Alminis:rator General offices: 1945 West ParnaH Road, Jackson, MI 49201 * (517) 788-1636 March 1, 1982 m
co a
'\\
N n
- c. c _ ;-0
\\
L's
'~ M 4 1932% r H
~
.O u. m cm~f (9 Dennis M Crutchfield, Chief M
La a_
Operating Reactors Branch No 5
',\\
[
M Nuclear Reactor Regulation N /
<0 US Nuclear Regulatory Commission kU JT'. '
Washington, DC 20555 DOCKET 50-155 - LICENSE DPR BIG ROCK POINT PLANT - SEP TOPIC II-2.A - SEVERE WEATHER PHENOMENA, III WIND AND TORNADO LOADINGS AND III-4.A TORNADO MISSILES By letter dated August 3, 1981, the NRC issued a final evaluation for SEP Topic III-4. A for the Big Rock Point Plant.
This evaluation had been issued previously as a draft on December 17, 1980, and Consumers Power Company (CP Co) comments were provided in letters dated January 23, 1981 and March 9, 1981.
In the final topic evaluation, the NRC has taken the position that the tornado wind speed which should be used for plant analyses is 360 mph.
This value appears to have been selected because it.c the standard value from Regulatory Guide 1.76, rather than because of the site-specific characteristics.
The staff position uses for support a study by J R Mcdonald of Texas Tech University entitled " Tornado and Straight Wind Hazard Probability for Big Rock PogtNuclearPowerReactorSite, Michigan." This study determined that a 10 probability median wind speed value for the site should be 272 mph and an upper 95% confidence value should be 371 mph.
In his analysis, Mcdonald defined a very large local region which extends over 150 miles south of the Big Rock Point site in order to obtain a substantial number of data points (see Attachment 1). shows the distribution of tornadoes through-out the state. As can be seen from these plots, neither the global nor the local regions are homogeneous with respect to the tornado distribution, and are therefore not representative of the plant site. Few tornadoes occur in
/
the northern portion of Michigan's lower peninsula. Mcdonald's local region i
g does include a large number of tornadoes, but it is apparent that most occurred at the southernmost end of this regica through Kent, Gratiot,
/ [
Midland, Saginaw and Bay Counties. The inclusion of this area as part of the oc0382-0001a142 8203050351 820301 PDR ADOCK 05000155 P
n i,
D M Crutchfield, Chief 2
Big-Rock Point Plant SEP Topics II-2.A, III-2 & III-4.A March 1, 1982 local Big Rock Point site region, therefore, gives an upward bias to the probability of a tornado affecting the site.
In addition, as noted in CP Co letter dated January 23, 1981 tornadoes tend to have the most severe characteristics over land and are less intense near shoreline sites. Since nearly every tornado which has occurred in Mcdonald's regions has touched down well away from the shoreline, as is to be expected, Mcdonald's predictions of intensities are also biased toward more severe tornadoes than would be expected at the Big Rock Point site.
It must also be noted that the Fujita (F-scale) classifications of tornado intensities, on which Mcdonald's work is based, classify tornadoes on the basis of the most severe damage within each entire tornado track.
In fact, however, a tornado will be most intense over only a fraction of its patb.
This is discussed in " Tornado Data Characterization and Windspeed Risk" by Lawrence A Twisdale, appearing in the October,1978 edition of the ASCE Journal of the Structural Division. This contributes still another source of conservatism to bcDonald's analyses.
In light of the above discussfits, it is readiiy concluded that Mcdonald's occurrence intensity curve overpredicts both the probability and the intensity of a tornado postulated to hit the Big Rock Point sitc.
Imposing characteris-tics even more extreme than the already-conservative expected values is not warranted. Any additional margin of safety offered by using an upper 95%
confidence value rather than a median value for a 10 event would be extremely small, and completely overshadowed by other higher probability events having the same or wor 9e consequences.
In short, there is no safety basis for considering the 10 tornado wind speed for Big Rock Point to be anything in excess of Mcdonald's median 272 mph.
It has been noted previously in CP Co letter dated January 23, 1981, that an 7
independent determination of a 10 value plant design tornado characteristic has been performed using ANSI /ANS 2.
Thisanalysishasshownthatthe10} Draft 2 Revision 4datedApril,1980.
wind speed applicable to the site would be 250 mph. This agrees well with Mcdonald's median value of 272 mph.
.In the Probabilistic Risk Assessment (PRA) for Big Rock Point, submitted to the NRC on March 31, 1981, the total probability of core damage 3
contributors at the existing plant was determined to be 1 x 10 Although this value may be perceived to be somewhat high, the effect on the public of a core damaging accident coupled with containment failure is very small because of the remote site and the small core size. Following planned modifications, the total core damage,grobability from all contributors will be reduced to approximmately 1 x 10 With this in mind, it is illogical to use as a plant design basis a single event which has a probability of occurrence several orders of magnitude less than other events which could cause core damage. For Big 7 Rock Point, therefore, it is not appropriate to evaluate the plant to be 10 tornado. The goal instead should be to ensure that.the risk from individual events such as tornadoes, when viewed independently, represents only a'small fraction of total risk; or as a complementary alternative, that oc0382-0001a142
1 D M Crutchfield, Chief 3
Big Rock Point Plant SEP Topics Il 1.A, III-2 & III-4.A March 1, 1982 the sum of the residual risks from all contributors be less than some overall goal.
For the specific case of Big Rock Point, a tognado wind speed value with a probability on the order of approximately 10 would be appropriate to ensure that the risk from the single event is small compared to other risk contribu-tors.
If it were assumed that a tornado wind in excess of this value would result in core damage, a very conservative assumption in itself, then tornadoes would still represent only a small percentage of the total residual core damage probability. _This is certainly true now, and will remain true af ter the other planned plant modifications are complete. The cost associated with analyses of lower probability conditions more extreme than these are simply not warranted for a plant of small core size like Big Rock Point.
Analyses of Big Rock Point structures under Topics III-2 and III-4.A are currently in progress. These analyses are being performed assuming a tornado wind speed of 250 mph. Although this value is very conservative in viey of the above discussion, it was selected some time ago because it was the 10 wind speed value determined by our tornado anaiysis._7This wind Speed also corre-sponds to a probability of approximately 2 x 10 in Mcdonald's work.
It is our intent to continue these analyses using 250 mph for wind speed.
In the event that specific structures are identifed which can not withstand this wind load, then lower values may be selected for further structural evaluations.
If the wind-induced failure (below 250 mph) of important structures from gross loading or missiles becomes significant with respect to other risk contribu-tors, then these structures will be evaluated using the PRA during the Integrated Assessment. The methodology for the Integrated Assessment has been described previously in CP Co letter of February 2, 1982.
O ll &. YO W] d h U David J VandeWalle Nuclear Licensing Administrator CC Director, Region III, USNRC NRC Resident Inspector-Big Rock Point Attachments 1 & 2 oc0382-0001a142
l i
l kn ca t og hRi g
h i gc aii rBM tS r oe dft n
i a yS t
oir dlo ait nb c raa ob e tor
" rP r c
q m
a od e
]
rrl l
f ac
-(i y
e zu Ng::
d.
t aN l
pl r
t ed n cni xio EWP T
N c
I O
P E
LN K
wp<
N AO C
I CIG O
OE R
LR ED G
I G,
D R
R EN-O f
F 3
~
l u
D SiO c:[-
NIDTI w
S 1
p-4 T
APA A N
s
- ..[ 7 s
Q O
T RA E T e
GRWS h
I N
G E
l a*'
sB E
M
=
R I
I s
s 0
C i
L A
e A
T B
T c ::.
H g
0 A
- W S 0::
l H
1;.
k G
i 0;:
i s
D
!:l
' i.
N d :.
A s
0, :.
h L
A
- g;f
- bC-,.
I C
E
.2:
3* N h
O
\\
': i L
t
- k
$ j!
l I
h
..h t
t 1
~
E R
U
~
I F
G I
S O
N
~-
l E
l l
lll 1
l
- '\\
ATTACM CIT 2 h
~i A
Ir; 4
_b
'1._.i p
4 l
1, i.
t_,
m- -
g',
d
'_._.1 2.
N %
.r
'. I i
-~
u
_js -
Qa i
,O
. j _*
I f
c s_j 7-Ihl" j""~" 5 Q
a l
I)
! es
.j.._,
- g i
ly I
- l... j* 3.9 a
,p__
l
[
l j*
o q _. - _. ; q -
i-I
'S
~
'I s*_._je.
g",
g
..-=!
k.~T._.
Tornadoes I
'l e l o
.], Q _, 4,.
f,h:* j' y,
1874 -1899 G
~ d5* { [& r O** TrM.
1900-1949 O
L[o
.,l***
L 1950 -1973 APPROXIMATE _
j[k j[
M io *k o
j.*
- rY, 4tA*-
6-i 9
y
- t g dg ~7 <0 4
g /c oo f : *.n *
.uj r. -(T y 1
0.
.. __ 9-gu--
u.
Extracted from
" Michigan Hazard Analysis" May, 197L
- -