ML20010F844
| ML20010F844 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 09/04/1981 |
| From: | Hukill H METROPOLITAN EDISON CO. |
| To: | Stolz J Office of Nuclear Reactor Regulation |
| References | |
| TASK-2.K.3.07, TASK-TM L1L-254, NUDOCS 8109110439 | |
| Download: ML20010F844 (5) | |
Text
- _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.
5 Metropolitan Edison Company l
Post Office Box 480 I$
L Middletown, Pennsylvania 17057 Writer's Direct Dial Number September 4, 1981 L1L 254
<3 fo Office of Nuclear Reactor Regulation Attn:
J. F. Stolz, Chief b--
Division of Licensing
(
8[p y 1 [O8N.g Operating Reactors Branch # 4
-pghfgg U. S. Nuclear Regulatory Commission g
Qj Washington, Is. C.
20555
,s s8
Dear Sir:
/
\\
Three Mile Island Nuclear Station, Unit 1 (TMI-1)
Operating License No. DPR-50 Docket No. 50-289 PORV Opening Probability (II.K.3.7)
Enclosed please find responses to questions 1 and 2 concerning the subj ect item.
Responses to the questions concerning Report of PORV Failures (II.K.3.2) will be provided in the near future.
Sincerely,
/
u i
H
).
Ilukillt Director, TMI-l HDH : LWH:vj f Enclosure p
( [f 0109110439 810904 PDR ADOCK 05000289 l
P pop Metropolitan Edison Comoany is a Member of the General Put ic Utilitie; System
. CU S.,. 1
. : u..
The licenste is to show how he deriv 2d pressure rollover data, which is the prirary system pressure rise above the high pressure reactor trip set;aint, as listed in Table 2.1-1 of the licensee's report.
.R_E_SPO:;S E The RCS pressure rollover was deternined as follows.
l.oss of feedwater transient data from the cperating plants were reviewed.
For those sets
~
of data in which RCS pressure plots were available, the peak RCS pressure was read.
Only the loss of feedwater events prior to March 1979 were considered.
For all of these events the Reactor Protection System high pressure reactor trip setpoints were 2355 psig.
The difference between the peak system pressure and 22's was calculated.
This simple difference represents the pressure overshcot for each of these particular events.
"o normalization or adjustment of the data was made to account for the difference in power levels or any other plant condition.
Twenty-six different loss of feedaater events were included in determining the rollover distribution.
Because there is a tendency for site personr.el to adjust reactor trip setpoints in a conservative fashion no data point was throcin out, even in those cases where the rollover was a negative numf>er, j
1.e., the reactor trip occurred before the design trip setpoint was actually j
reacneo.
This was done to provide a realistic estimate of the true expected i
roll:ver distribution.
i Q'iESTIC'i2 The licensee is to sheet heu he obtained the maans and the standard deviations for the three normal distributions used in the l'onte-Carlo analysis to I
I, t
6 simulate the opering of PORV during an overpressure transient as discussed in 2.1.1 POR" Opening Probability based upon analyses in the license's report.
!!SFONSE '
The actu 1 setpoints for PORV actuation and high RC trip are assured to be a normally distributed random variable with means of 2450 and 2300 r.spectfully.
The standard deviaticn of the setpoints were calculated from instrur. ant accuracy values provided by the manufacturer.
The small standard deviation (1.4 psi) is due to common redules in the PORV actuation and RPS high pressure trip channels.
Both signals are processed through the same modules with the exception of one amplifier and one bistable in each path which are not common to both signal paths.
The errors in the common signal paths cancel out leaving only the error in the two non common modules for each setpoint.
It is assumed that the errors remain constant over the entire pressure range.
Both modules have an accuracy of.1%
and the pressure rcnge (narrow range) of interest is approximately 1000 psi.
The. standard deviation is derived as bl.001x1000+.001x1000=
Ndipsi?1 psi =1.4 psi.
It should be noted that the standard deviation of the overall calculations (1.4)2 (j,4)2 + (27.52)2 is dominated by the 3rd tenn which is associated with the rollover data.
The mean of the rollover data has been calcuated to be 9.2 psi with a standard deviation of 27.5 psi.
The supporting calculations and 'd test follow.
The data given in Table 2.1-1 is also plotted in the attached figure.
The parameters of the data were calculated with the W value of.07756 supporting the assumption of normality.
- ANSI N15. 15-1974 Assessment of the Assumption of Normality e
j w
o 2 = 26 (Xg - x)2 = 18934.62 Coefficients used in the W test for l
S
~,
i=1 nonnality with sampic size 26.
o k=13 A.g,j n
k
.4407 b = t.
An-i+1('n=i+1~ i) = 136.05
.3043 i=1
.2533 i
2
.2151 W
=b
=.97756 lF-
.1836 S
.1563
.1316
.1089
.0876
.0672
.0476
.0284 2
.0094 e
O e
I
+
I i
1 4
1 i
i I
v t
4 ROLLOVER DATA I
l Gr K
50 i
x x
x 40 x
1 X
1 30 x
x i
f 20
^
j E
10 - -- K X
- ~ ~ ~ ~-~~ - -------
AVER AGE ROLLOVER = 9. 2 v
q.
g
- .y '
- s.,,
l g
v X
X
-10
- 5 10 15 20 25
- 35 40 45 50 55 00 65 70 75 80 85 00 05 100 l
g l
1 c:
20 X
t
-30 xx l
i i
-40 Power, %
-50
-00 i
4
(
I i
1
!t 1
l f
i, 4
e.