ML19319C199
| ML19319C199 | |
| Person / Time | |
|---|---|
| Site: | Davis Besse  |
| Issue date: | 10/08/1976 |
| From: | Martin R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| To: | Knop R NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION III) |
| Shared Package | |
| ML19319C193 | List: |
| References | |
| NUDOCS 8002050788 | |
| Download: ML19319C199 (1) | |
Text
.
UNITtD ST/.Tec
/
NUCLEAR REGULATORY COMMISSION
~,
e REGlorJ til 793 RoOSEV' LTROAD Ct.EN ELLYN. ILt.f t4015 Got37 October S, 1976 R. C. Knop, Chief, Reactor Projects Section 1 RESPONSE TIME TESTING ASSUMPTIONS - DAVIS-BESSE I (DOCKET NO. 050-346)
During a recent inspection, I reviewed the methodology to be used by the licensee (Toledo Edison) in determining response times of safety related components and systems. In general, the program appears satisfactory and has good technical bases for the techniques being utilized.
However, the conduct of the program will utilize several basic assumptions, two of which I would urge be concurred in by Headquarters.
g'77 The first 13 that the licensee is assuming that analog devices, in this
' ;/
case pressure transmitters and possibly RIDS, exhibit the same response time characteristic when subjected to increasing or decreasing input.
. j/.,y,y'This is an essential assumption to the test technique being utilized.
t
,f,.
, f. J' 4
i Q,The second is that for "large" populations of identical devices (in this
{J # j pf(,i' l case approximately 240 relays and 70 analog amplifiers), the response g
g,7 time of these two devices can be determine ( by measuring a sample of the 13 '-f devices. Their acceptance criteria includes a randomly selected group of
[d/
- )/g measured response times must then fall into a spread of j-20% of the av l
at least 10% of the devices (24 relaya and 7 analog amplifiers). The i
2[
/ y value obtained.
If these conditions are det, the response time will bc
((
the largest measured value for the act.
11*
p 3'
I request that Heade,uarters be requested to provide guidance with respect Y
to Davis-Besse I in the following areas:
1.
Is the assunption that analog devices have equivalent response time 1
characteristics for increasing or decreasing signals sufficiently H]
s.% V Y j#g* n.sg w/.
valid to assure Technical Specifications compliance?
. }&.
Is the sampling =cthod to be used for identical devices accept.bleM. fu--
~
2.
or must a larger sampic or closer reprcducibility be utilized?
(An A /L P W alternative approach would be to assign the population a response time 5'E, --
3 -f
- 7. vs of approximately 1.5 times the average value which would provide a 95%
confidence level that the true response time is less than the assignel
~~
7. '7 74I value.)
k/!5Yk, s7, W6b
/
R. D. Martin WTIo3",
Reactor Inspector q.p y jgf Reactor Projects Section 1 g '
,in c.
6 (.
)
cc:
G. Fiorelli
. s,u -...,
~.
D7c.tn8 8 0 02 05 gyg 1
3 g
y-
- --