ML20116E630
| ML20116E630 | |
| Person / Time | |
|---|---|
| Site: | Calvert Cliffs  |
| Issue date: | 05/14/1990 |
| From: | Office of Nuclear Reactor Regulation |
| To: | |
| Shared Package | |
| ML20116D885 | List:
|
| References | |
| FOIA-96-237 NUDOCS 9608060066 | |
| Download: ML20116E630 (6) | |
Text
.
3 SAFETY EVALUATION CALVERT CLIFFS SURVEILLANCE INTERVAL EXTENSIONS DOCKET N05. 50-317/318
1.0 INTRODUCTION
The following safety evaluation considers proposed Technical Specification changes 3, 4, 5, and 6 described in the Baltimore Gas and Electric (BG&E) Request for Amendment for the Calvert Cliffs Nuclear Power Plant (Units 1 and 2) dated December 14, 1988.
2.0 EVALUATION
2.1 Change 3 proposes to change Surveillance Requirement 4.8.1.1.1.b from extant wording: " Demonstrated OPERABLE at least once per 18 months during thutdown by manually transferring unit power supply from the normal circuit to the alternate circuit." The underlined words are those which licensees proposes to change to:
" REFUELING INTERVAL". The alternate circuit is the 69kV SMECO offsite power circuit. This is the only proposed change for this TS section.
i The licensee states that he has evaluated the proposed change against the standards in 10CFR 50.92 and concluded that: "The proposal... to demonstrate operability of the alternate power circuit every 24 months, versus every 18 months, will not signif-icantly increase the probability or consequences of a previously evaluated acci-dent.... no new failure modes are induced. The margin of safety will not be significantly reduced...."
The staff recommends allowing the licensee to change the TS Section 4.8.1.1.1.b per the December 14th Request For Amendment. A discussion of the staff's basis follows.
4 l
The probability of any time-related, randomly occurring, 69kV preferred power supply (PS) failures will not be affected by surveillance test intervals unless the random process has an increasing average frequency, and time dependent effects are significant. Such an incrt6 sing frequency has not been established.
9608060066 960731 PDR FOIA DINICOL96-237, PDR
. l The change from demonstrating operability of 69kV offsite power supply every 18 l
months to every refueling outage, instead, is a change, for Calvert Cliffs, from l
18 months to a possible (513/.70) + 90 days = (823 days /30.5) = 27 months. The l
assumptions upon which this calculation are based are: that Calvert Cliffs has the same 513 day fuel cycle as San Onofre 2,3; and that the Unit Capacity Factors l
cumulative Maximum Dependable Capacity Net and Design Electrical Rating Net of
.7, as of August 1989, will not fall much lower; (See NUREG-0020).
Unit Capacity Factors (UCF) lower than.7, e.g., in 1989, Calvert Cliffs 1 and 2 had UCF approximately equal to.3, are generally aused by outages long enough i
to allow surveillance before return to power, so that there should never be a need to extena any surveillance interval beyond thirty months. Such a limit j
would keep associated risks from doubling the risks owed to the present eighteen month refueling and surveillance intervals. Accordingly, the staff recomends limiting the subject four proposed T/S changes to thirty months.
The basis for 18 months surveillance intervals used by many licensees and the NRC was not a PRA, but the convenience due to performing surveillance on the 69kV power supply when the operating units did not have to be tripped to make the associated tests. Since that was the only basis fcr the 18 months maintenance /
surveillance interval, and the staff knows of no discovery since which shows that the basis has proven unacceptable, or that an increase of six to twelve months would reduce, excessively: any margin of safety inherent in the 18 months interval, the viability of the basis remains the same. This is the basis for the test interval for many safety systems and is not peculiar to electrical systems.
Further, surveillance testing has no effect on the probability of the next demand for a transfer being a success; it can only demonstrate the success of the last transfer. The reliability of the breakers is primarily based on the number of switching cycles, decreasing as cycles increase, and only distantly on elapsed time alone.
The staff has provided in this Safety Evaluation an intensive discussion of its basis for accepting the licensee's proposed Technical Specification change 3, because the basis offered in the licensee's Request for Amendment of December 14, 1988 was judged, by the staff, to be incomplete, and insufficient by itself.
r~
\\
. l 2.2 Change 4 proposes to increase the Emergency Diesel Generator (EDG) surveil-l lance interval in T/S Section 4.8.1.1.2.d from 18 months to REFUELING INTERVAL.
]
This change is described in the December 14th letter, misleadingly, on page 7, l
as a change from 18 months to 24 months. [ Note: The actual period between sur-veillances could depend on the net Maximum Dependable Capacity as was discussed inChange3.]
l The basis for recomending that the Emergency Diesel Generator surveillance inter-val be extended to REFUELING INTERVAL but in no case to over thirty months is the same as above; namely, the basis for the interval will remain the same, the con-j venience of testing when plant is not at power. The staff has no evidence to l
irecicate that the relationship is inappropriate for the EDG surveillance specified.
Further, the staff also s'spports the notion that it is prudent to conduct certain l
tests with the reactor plant in refueling versus at power operation.
2.3 Change 5 propose 5 to delete T/S section 4.8.1.1.2.d 3(c), " Verifying that the high jacket coolant temperature (JCT) and low Jacket coolant pressure trips (JPT) are automatically bypassed on a Safety Injection Actuation Signal," and reinserting it as section 4.8.1.1.2.d.6.
This removes the need to simulate a concurrent loss of offsite power when verifying that the jacket coolant trips are automatically bypassed on a Safety Injection Actuation Signal as was required by section 4.8.1.1.2.d.3(c).
Licensee's basis for change 5 is stated as follows: "The logic circuit that bypasses the jacket coolant trips only requires an input from the safety injec-tion circuitry; there is no input from the undervoltage circuitry." Thus, the i
proposed change would bring the T/S into correspondence with the hardware in j
place. Absent a clear need for installir.g a LOP input for JCT /PT bypass, the staff finds the proposed change acceptable.
2.4 Change 6 would delete from T/S section 4.8.1.2, Surveillance Requirements for Shutdown, the requirement in T/S section 4.8.1.1.2a.7, " Verifying that the l
automatic load sequencer timer is OPERABLE with the interval between each load
r block within 10% of its design interval." The staff finds the proposal to not test the automatic load sequencer timer when the unit is shutdown acceptable.
because the automatic load sequencer is not necessary to plant refueling or any other operations during shutdown. However, the entire on-site electric power system including the sequencers will be demonstrated Operable before either unit returns to criticality.
l j
l l
1 l
-. ~.. -
i ENCLOSURE 2 I
l l
SALP INPUT FACILITY NAME:
Calvert Cliffs 1 and 2
SUMMARY
OF REVIEW:
l Baltimore Gas and Electric, by letter dated December 14, 1988, proposed six Tech-i nical Specification changes. SELB reviewed proposed changes 3, 4, and 5.
The staff accepted these three reviewed changes.
l Change 3 increased the 69kV offsite power circuit demonstration of operability interval from 18 months to 30 months. Licensee had proposed, in the December 14th letter, to increase the offsite power demonstration interval from 18 to 24 months-and from 18 months to refueling interval in the Technical Specification marked-up pages, attached to the letter.
Change 4 increased the Emergency Diesel Generator Surveillance Interval from 18 to 30 months. Licensee had proposed changing surveillance interval from 18 months i
to refueling interval.
l Change 5 deletes from the technical specification the need to simulate the concur-l rent loss of offsite power when verifying that emergency diesel generator jacket coolant trips are bypassed on a safety injection actuation signal. Because there is no input to the jacket coolant trip logic from undervoltage relays sensing offsite' power, the staff accepted the change.
l Changes 3 and 4 were accepted because the 18 month surveillance interval for many systems is based solely on the length of the refueling interval and should l
therefor < change if the refueling interval length has changed.
i hARRATIVE DISCUSSION OF LICENSEE PERFORMANCE FUNCTIONAL AREA: ENGINFERING/ TECHNICAL SUPPORT:
Changes proposed by licensee were administrative and of little safety signifi-l cance, and only for licensee's convenience - i.e., performing surveillance when the plant is shut down. As such, they were accepted by the staff. But licensee's bases for the surveillance interval changes were: first, that he had performed i
l extensive research to determine if any consistency problems or detrimental trends i-had developed between inspections at 18 month intervals;-and second that the NRC had permitted another licensee to increase his surveillance intervals and, there-fore should do the same for Calvert Cliffs. As to the first basis, staff found after a discussion with the licensee's staff that no formal procedures were used in the extensive research, that no criteria were established, no data gathered, j
no. statistical techniques employed, and no documentation resulted. The research consisted of a search of some unspecified log books and records. This is not
r 2
whst the term extensive research currently means.
It is, in fact, such a radical definition as to amount to, if used without immediate clarification, an attempt to mislead the staff.
The second basis, that of consistency, should apply, but only when licensee demonstrates in his proposal that his plant is sufficiently like another.
Proposed changes were simple, yet, in one place, the licensee wrote that the new surveillance interval was to be 24 months, and in another place, in the same sub.
mittal, that it was to be the refueling interval, a variable period limited to thirty months by the staff. Requisite care and review in the preparation of this submittal does not seem to have been exercised by licensee's staff.