ML19271E256
| ML19271E256 | |
| Person / Time | |
|---|---|
| Site: | Limerick  |
| Issue date: | 05/17/1984 |
| From: | Thompkins B, Tompkins B NRC ATOMIC SAFETY & LICENSING APPEAL PANEL (ASLAP) |
| To: | DEL-AWARE UNLIMITED, INC. |
| References | |
| LBP-83-11, OL, NUDOCS 8405180443 | |
| Download: ML19271E256 (2) | |
Text
z
's
, t* p*o' 'E cq o,
UNITED STATES g *, [e-
.' )
WASHINGTON, D. C. 2055R
e,,*~
j NUCLEAR REGULATORY COMMISSION
- 3..
g%
/
MAY - a sgr MEMORANDUM FOR:
C. E. Rossi, Chief Events Analysis Branch Division of Emergency Preparedness and Engineering Response Office of Inspection and Enforcement FROM:
Gary M. Holahan, Chief Operating Reactors Assessment Branch Division of Licensing Office of Nuclear Reactor Regulation
SUBJECT:
PROPOSED IE BULLETIN ON UV TRIP ATTACHMENTS ON REACTOR TRIP BREAKERS The following analysis was performed in order to be responsive to V. Stello's concern that additional monitoring of parameters (such as trip time) could be detrimental to safety ir that it cokld increase the frequency of reactor trips. The analysis indicates that:
- 1) few reactor trips are caused by RTB testing; 2) the monitoriig of parameters is not likely to noticeably increase the frequency of reactor trips; and 3) that the contribution to core melt from these activities is extremely small.
~
The event sequence of interest is a reactor trip followed by loss of feedwater (main and emergency). The estimated probability for this
.dence (for a PWR) for all reactor trips is:
P (core melt) = P(trip)
P(Low Aux Fw)
P (inadeo Feed & Bleed) 10-4 0.5 P (core melt) = 7/RY 0.1 P (core melt) = 3.5 x 10-5/RY The probability of such a sequence being initiated due to RTB testing is:
P(cm from RTB testing) = P(trip from RTB testing)
P(Lo in FW)
P(Low Aux FW)
P(inadeq. Feed &B}eed)
P =.02/RY x
.1 x 10- x 0.5 P=1.0x10g/RY h
T)h,0 Q
Where the probability of a reactor trip due to RTB testing is based on a point g
estimate from 1983 PWR operating experience, (i.e. O trips due to RTB testing if f in50PWRRY's).
Based on INP0 statistics for the last six years, only 9% of h f reactor trips are associated with ectivities on any part of the RPS and only f
4% are associated with errors c'uring testing.
It is therefore not surprisina that there have been no recent, examples of reactor. rips caused by RTB testing.
8405180433 840504 CF SUBJ TD&R-11A IEB
(,7
MAY - 41984 C. E. Rossi -
Under the most conservative set of assumptions (i.e. assuming that nonitoring would be the cause of all cases of RTB testing resulting in reactor trips) and using a conservative point estimate of the frequency of RTB testing resulting associatedwi'hRTB'swouldbe1.0x10ftyfromthecurrentnonitoringactivities in reactor trips, the core melt probabil
/RY.
Since the proposed bulletin would increase the amount of monitoring of parameter during RTB testing by a factor of 2 to 4.
/RY to 4.0 x 10-ym increase in core melt probability would be from 1.0 x The maxim 1.0-7
/RY. The actual contribution is likely to be significantly lower, perhaps by several orders of magnitude.
In addition to being the direct cause of reactor trips, RTB parameter monitoring could contribute indirectly to the frequency of reactor trips by keeping a given RTB out-of-service longer than it would be otherwise.
During such a period, the plant could experience a reactor trip due to a single failure in another part of the RPS. Currently RTB are out-of-service for about 8 days / year (2 days of servicing twice per year plus 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> of testing each month).
For most plants the current monitoring of results in an additional time out-of-service of
.5 hours5.787037e-5 days <br />0.00139 hours <br />8.267196e-6 weeks <br />1.9025e-6 months <br /> / year ( 25 hours2.893519e-4 days <br />0.00694 hours <br />4.133598e-5 weeks <br />9.5125e-6 months <br /> twice per year) or.3% of unavailability (i.e., plant susceptibility to trips on single failure). Under the proposed Bulletin, this component of unavailability would be increased by approximately a factor of 4 resulting in a 1.2% contribution.
Since all RPS testing only contributes 4% to reactor trips, the contribution to reactor trips of parameter monitoring would be less that 1 would be 1.7 x 10-8 2% of 4% or.05%.
The corresponding core melt contribution
/RY.
An analysis of the improvement in RTB reliability due to early identification of degraded performance is not available, however, it would appear to also have only a very small contribution to changing the estinated probability of core melt since RTB reliability is not the controlling factor in the RPS reliability for most systems and since the probability of a core melt from an ATWS event is already relatively small.
In surmary, the proposed Bulletin will clearly not cause a significant reduction in safety; will probably cause a small improvement in estimated core melt probability and will address an identified deficiency in an important safety related component of the RPS. The mr eitoring parameters during RTB tests wiit identify, and result in repairs, of dV devices which are undergoing progressiu degradation and which are likely ',o full if not given attention.
On this basis I continue to support the propoNd bulleti.,,
Ug'"W 5, ed By:
Gary M. Holahan, Chief Operating Reactors Assessment Branch Division of Licensug Office of Nuclear Reactor Regulation cc:
F. Miraglia E. Jordan E. Buche J. T. Beard J. Conran OR[B d DISlRIBUTION
- GHolahat, Jertral files GHolahan 5/y/84 ORAB r/f