ML19262A246
| ML19262A246 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 09/10/1975 |
| From: | METROPOLITAN EDISON CO. |
| To: | |
| Shared Package | |
| ML19262A242 | List: |
| References | |
| 75-05, 75-5, NUDOCS 7910260582 | |
| Download: ML19262A246 (2) | |
Text
"etrcpclitan Ediscn C: pany Olet-Ed)
Three ' tile Island Nuclear Station, Unit 1 (TM:-1) c-
.CC4et
..c, v-cc9 s
Cperating License Uc. CFR-50 Ncnroutine 3C-lay Report 75-05 nLw c..r.o..
r.l ce n & care.-cv n-aL-. ~
- n. e - L s.
.w.
s.
Pecently it was disecvered that the ultraviolet (U7) detectors used in the Air Intake Tunnel Fire Frctection System fcr TMI-l are of a different type than criginally specified by the vender (Fenval) drawings. Further cre, the basis fcr Technical Specificaticn 5.5 and pcrtiens cf the Final Safety Analysis Repcrt (FSAE) describe the incorrect UV detectcra.
Cn Auguct 22, 1975 it was determined that this discrepancy was repcrtable under Secticn 6.7 2(b)2 cf t'a TMI-1 Technical Specificatiens as a sub-stantial variation frc= performance specifications.
The tvc U7 detectcrc in question are:
a.
Fenval Part #1C-190017-222 which is installed at TMI-1, and b.
Fenval Part #10-190016-00L which is shcvn en Fenval drawings 952h5-01, 2, 3 Rev. A/3 and implied in the basis fcr Technical Specifications for TMI-1.
The differences between the tvc types cf detectcrs are given in Attachment #1 to this report. The scst significant of these differences is the total reaction time from introducticn of the triggering ccnditice to full flev cf the Ealen which is 68 =illiseconds with the "criginal" detectcr and 17h tillisecond with the " installed" detectcr, fcr a 5" diameter 100 cctane gascline panfire at 3 feet, (this reacticn time is 33 milliseccnds and 150 milliseconds for the "criginal" and " installed" detectors respectively, for a 1.75" prcpane flame at 8 inches.) This difference in reacticn time is caused by an inherent time delay cf the " installed" detecter - 156 millisecond for a 5" dia. 100 cetane gascline panfire - as against the 50 millisecond time delay for the "criginal" detector for the same condition.
Frcm Fenval test data, an unsuppressed explosien requires 1.36 seconds (1360 til11 seconds) to go to ecmpletion in the vertical secticn cf the tunnel and the intent of the specification '.
to detect this explcsion and discharge halen within a small fractic this time period.
This time pericd has been specified as 250 mill.
- cnds in Sec. 4
- 15.1 in Ded gn Specification SP-5701. The installed detector satisfies this criterien adequately.
'he next aspect for censideration is *he change in the cone cf visien.
6 For tt.e criginal detector, it was 120 and for the installed detectors it is 90~.
Fenval Inc. has made a study cf the existing lay 9ut at IM:-1 and has concluded that the installed detectcr with the 90' cene of visicr provides adequate ecverage.
The other difference is in the Spectral Eange fcr :he Tvc detectcrs:
for the criginal detecter it is 1,000-29^0 Angstrc= units, whereas for the installed detecter it is 17:C-26CC Angstrc= units. The detecticn range 1:
still veil within the ' ltraviclet band of the electrcragnetic spea- "-
? 939g60 m
. Frca the abcVe discussicn, it was ecncluded that the " installed" UV detectcrs,
Fenval Fart No. IC-l?C017-222, do not in any way change the design criteria cr any cther censideration and therefere are saticfactcry in all respects.
1.
CE2:GUATICS CF CAUSE CF LISCREFASCY lt has been determined that the cause of this discrepany was a breakdevn in the quality Assurance (QA) pregram that existed at the time that the UV detecters were installed, in that the QA program either allowed installatien of the inccrrect detectors cr did not prcperly decurent this change so that it could be incorpcrated into the FSAP..
It shculd be noted, that the Cperation C,A pregram had not been fully implemented at the time that the inccrrect detectors were installed, and that no pctential for a si=ilar descrepancy to occur exists under the QA progrs= as it ncy exists.
an
+--e u
1
. ~.c rc.m.. s - n. a.w Ocrrective acticns will be taken as fcllows:
a.
An investigation vill be conducted in crder to provide adequate assurance that similar circumstances do not exist in other systems.
b.
A Technical Specification Change Request will be submitted by October 15, 1975 c.
An stendment will be generated to correct applicable portions e
C.e +h
- c.n.n. ky ew e +. v w e.
-k
,.-ls.
,y a
j o
w 1484 223
!TACHMENT
- 1 n
7APIAELE 1)-190016-COh 10-19C017-Z;2 L'/V OETF"TCE U/7 DETFCTCI-2pe' tal Bange ( A Units )
19C0 - 2900 17C0 - 2600 Max. Eesponse (A Unitss 2200 CISC Cene of Vision 100 92 hesponse at Static Cettir.g (M3)
NiA N/A Response to 1 75" Propane Flame at 3 Inches (MC) 15 ikC Respcnse to 5" Dia. 100 Octane Gasoline Pan Fire st 3 Feet
'3 )
50 156 Centrol Relay Full In (MC) 5 Burst Disc Crens (MS) 5 5
3 tart of Full Flcw cf Agent frcm Spreader (MC) 8 3
Total Reaction Time (MC)
To 1 75" Propane at 8 Inches 33 158 To 5" Dia. 100 Octane Pan Fire st 3 Feet 63 174 Y
sd i484 224