ML20115E015
| ML20115E015 | |
| Person / Time | |
|---|---|
| Site: | Peach Bottom  |
| Issue date: | 07/02/1996 |
| From: | Hunger G PECO ENERGY CO., (FORMERLY PHILADELPHIA ELECTRIC |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| NUDOCS 9607160060 | |
| Download: ML20115E015 (6) | |
Text
.
.m Station Cupport Deptrtment 10 CFR 50.90
=5lEir_
+
PECO NUCLEAR ncme. c-v Nuclear Group Headquarters A UN1T OF PECO ENERcy 965 Chesterbrook Boulevard F
Wayne, PA 19087-5691 I
i July 2,1996 Docket Nos. 50-277 50-278
]
Ucense Nos. DPR-44 DPR I U. S. Nuclear Regulatory Commission i
ATTN: Document Control Desk Washington, D.C. 20555
SUBJECT:
Peach Bottom Atomic Power Station, Units 2 and 3 Response to NRC Request for Additional Information Regarding i
Faculty Operating License Change Pequest 93-18, " Wide Range Neutron Monitoring l
System."
Dear Sir:
j t
j By letter dated January 17,1995, PECO Energy Company submitted Ucense Change Request (LCR) 93-l l
18, requesting changes to Appendix A of the Peach Bottom Atomic Power Station (PBAPS) Facility Operating Licenses The proposed changes were submitted to support PBAPS Modif'mation P00271, which would replace the Source Range and Intermediate Range Monitors with a new Wide Range l
Neutron Monitoring System (WRNMS),
By letter dated AprH 4,1996, the NRC requested additional information to complete their review of LCR 93-18. This letter provides the requested additional information. In the attachment to this letter, each
~ NRC question is restated followed by the PECO Energy response.
If you have any questions concerning this submittal, please contact us.
Sincerely, W.d. M f/
G. A. Hunger, Jr.
j Director - Licensing I
j
Enclosures:
Affidavit, Attachr.wnt i
cc:
T. T. Martin, Administrator, Region I, USNRC
//
W. L Schmidt, Senior Resident inspector, PBAPS, USNRC R. R. Janati, Commonwealth of Pennsylvania i
9607160060 960702 I
,7-PDR ADOCK 05000277 i
P PDR i
{
COMMONWEALTH OF PENNSYLVANIA i.
- SS.
COUNTY OF CHESTER D. B. Fetters, being first duly swom, deposes and says:
That he is Vice President of PECO Energy Company; the applicant herein; that he has read the attached response to the NRC request for additional information regarding License Change Request 93-18 for changes to the Peach Bottom Atomic Power Station FacBity Operating Licenses DPR-44 and DPR.
56, and knows the contents thereof; and that the statements and matters set forth therein are true and correct to the best of his knowledge, Information and belief.
lL A A__
~ Vice Prendent l
Subscribed and swom to
'M before me this c/
day of C
/
1996.
(
~1L u n $ 5 0A S
. f4 ry Public Notarial Seal May Lou Skrocki, Notary PubHc Tredyffrin Twp., Chester County l
Vy Commdsion Emres May 17,1999 b+2w, Porn @a aucuxutonc: fin 4
l L
l
)
ATTACHMENT l'
t NRC QUESTION 1 l
. SR 3.3.1.1.6 - APRM/WRNM overlap l
The licensee proposed to delete SR 3.3.1.1.6, which verifles overlap between the intermediate Range Monitors (IRMs) and Average Power Range Monitors (APRMs) during controlled shutdown The deletion of this SR is not consistent with Table 4.3.1.1-1 on page F 10 of NEDO-31439-A (Reference 5), in which this SR is retained.
How does the licensee propose to achieve assurance that the WRNM channels are functioninD properly when entering Mode 2 from Mode 17 l
The staff requests that the licensee provide justification for the deletion of this SR.
i PECO ENERGY RESPONSE TO QUESTION 1 l
This SR verifies overlap between the Intermediate Range Monitors (IRMs) and the Average j
Power Ran0e Monitors (APRMs). In the current system, this surveillance performs several l
functions. It confirms that the IRM gains, ranges, pre-amp, etc, are set up correctly. It also confirms that the IRM detectors are inserted into the core..Together with norrnal operator procedures, this assures that neutron flux monitoring and scram trip protection are in effect when entering Mode 2.
~
The replacement WRNMS uses fixed incore detectors, therefore, SR 3.3.1.1.6 is no longer required to assure that detectors are inserted. The upscale level based IRM scram trip is deleted for WRNM and replaced with the rate based "short-period" trip. The IRM range switches j
and the two range pre-empilfier are eliminated. Consequently, no set-up selections or adjustments are required to assure that the WRNMS monitoring functions are active. The trip
)
calculation and period extraction is performed digitally, therefore, no calibration is required. The period extraction depends only on proportional chan0e in the WRNM signal relative to neutron flux, a characteristic that is independent of WRNM gain adjustments With these system changes, no specific surveulance test is required to assure that neutron monitoring is available. Normal operator procedures exercised prior to entering start-up mode.
assure that the WRNM is operable. Operator panel displays are presented in such a manner that makes it easy to conclude that the WRNM channel is operating, and that si nal outputs are 0
reasonable. This is all that is required to assure that the WRNMS channels are functioning properfy when entering Mode 2.
NRC QUESTION 2 SRs 3.3.1.1.11 and 3.3.1.2.7 - Change Survellance Test interval (STI) from 184 days to 24 j
months.
i The licensee is proposing simultaneously to install new equipment and to increase the 4
j survetlance interval. The licensee states justification as follows:
1 The licensee stated in Reference 2, pa0e 2, item ill, paragraph 3, *... the survellance test 1
i j
1 i
k 1
i
frequency change of 184 days to 24 months is based on trip setpoint calculations using GE's j
l standard (NRC approved) setpoint methodology." The licensee further stated in Reference 4, j
(
page 410, Table 4-4, " Change channel calibration to reflect the 24 months frequency applicable 1
l to the WRNM."
t The licensee's statements appear to imply that the staff has granted generic approval for the 24-month CHANNEL CAllBRATION frequency for the WRNMS for these functions, which it has not.
l The staff requests to review the setpoint calculations and other considerations upon which the l
licensee bases the extension from 184 days to 24 rnonths.
)
PECO ENERGY RESPONSE TO QUESTION 2 l
The basic methode4ogy used to calculate WRNM setpoints is unchanged from the current IRM ard SRM system. The significant difference between the systems is the reduced effect of drift i
on the WRNMS.
The IRM and SRM both use analog components for initial signal processing, mean-square-i voltage (MSV) calculation (IRM), period extraction (SRM) and trip processing. The SRM and IRM l
use individual pre-amplifiers. The IRM pre-amplifier has two ranges and the IRM instrument has 10, each of which require calibration.
The WRNM uses a single, improved pre-amplifier. This pre-amplifier is a simpilfied digital design which has eliminated all ranges and adjustable components. Therefore, no hadware calibration is required. The MSV calculation is p dunried with fuHy integrated higNy stable components.
j All ranges are eliminated from the signal processing; period extraction is podunried digitally. As a result of these simplifications in processing hardware design, and the very significant improvement in component quality, term drift of component values that affect signal processing is virtually eliminated. All trip calculations are performed digitally, therefore, drift is not a concem. In addition, the change in the WRNM from level based trips (for IRM) and digital extraction of period makes the safety trips and indication of increasing reactivity very insensitive to any residual drift in the processing Since the calibration frequency is Inliuenced primarly by drift, the virtual elimination of drift in the WRNMS justifies the factor of 4 reduction in calibration frequency. This is cois";inied by using the WRNMS estimated drift effects for 24 months in the normal, NRC approved, setpoint calculation methodology when calculating WRNM setpoints. Since this yields fuHy adequate margins, the necessary Technical Specification Surveillance Interval is determined to be 24
(
months. If, after reviewing the above information, the Staff stRI desires to review the setpoint calculations upon which the extension from 184 days to 24 months is based, copies wRl be provided.
NRC QUESTION 3 SRs 3.3.1.1.3 and 3.3.1.1.4 - Change STI from 7 to 31 days The licensee proposed to change the STI for the CHANNEL FUNCTIONAL TEST (for WRNMS functions) from 7 days to 31 days. The licensee stated in Reference 2, page 2, item lii, paragraph 3 that, "The surveRiance test frequency change of 7 to 31 days is based on the WRNMS having (1) fixed in< ore detectors, (2) greater reliabRity than the SRMs and IRMS, and (3) self test j
features."
i 2
4 8
}
+.
1
On page 1.1 1 of Reference 2 the CHANNEL FUNCTIONAL TEST is defined as follows: "A CHANNEL FUNCTIONAL TEST shall be the in;ection of a simulated or actual signal into the channel as close to the sensor as practicable to verify OPERABILITY, including required alarm, interlock display, and trip functions, and channel falure trips. The CHANNEL FUNCTIONAL TEST may be performed by means of any series of sequential, overlapping, or total channel steps so that the entire channel is tested."
Does the WRNMS automatic self-test verify OPERABILITY as completely as does the CHANN'IL FUNCTIONAL TEST 7 List the portions of the channel for which the WRNMS automatic self-te at does not verify OPERABILITY.
PECO ENERGY RESPONSE TO QUESTION 3 The current IRM uses analog circuitry for signal processing and analog and relay circuitry for trip functions. Falures in either the analog or relay circuits could result in a loss of the trip function.
Since the purpose of the functional test is to demonstrate that the entire circuit is operable, the existing 7 day test frequency was based on the perceived weakest link in the chain; the analog circuitry. This interval was based on engineering judgement rather than actual predicted or observed faBure rates. Relay logic systems typically have test intervals greater than 31 days (e.g., SR 3.3.1.1.9).
in the WRNMS, all functions except the initial signal processing and the final NUMAC trip output relay are performed digitally. Self-test covers signal reading from individual modules, all CPU processing, and signal transmission to the output modules. Self-test also monitors the relay outputs to confirm that the output state matches the logic command. If a trip were to occur, and the output of the NUMAC chassis did nct change state correctly, the system would t
automatically set the Inop trip.
With this coverage, the only portions of the trip circuit that are not monitored continuously by the self-tests are the initial signal processing and the relay logic. If all parts of the circuit were continuously monitored, it could be argued that there would not be a need for a channel functional test. However, since that is not the case, a functional test is appropriate to demonstrate operability of those parts of the circuit not covered by the self-test or another more frequent SR.
The initial signal processing is covered by a channel check every 12 hours1.388889e-4 days <br />0.00333 hours <br />1.984127e-5 weeks <br />4.566e-6 months <br /> (SR 3.3.1.1.1). This wil detect a loss of, or significant difference between, channels. Therefore, the functional test interval should be one that is appropriate for relay logic circuitry. Considering that certain logic systems presently have test intervals greater than 31 days, the proposed 31 day test interval is conservative.
NRC QUESTION 4 SR 3.3.1.2.5 - Change STI from 7 to 31 days The licensee proposed to extend the STI for the WRNMS (operating in the Source-Range mode) from 7 to 31 days when the reactor is in mode 5.
The current PBAPS TS aNows a frequency of 31 days for SR 3.3.1.2.6, but requires a briefer 7-day frequency for SR 3.3.1.2.5. The current TS Bases require the briefer 7-day frequency for SR 3
3.3.1.2.5 to ensure "...that the channels are operable whle core reactivity changes could be in 3
~,
progress." In the proposed change to the TS the licensee would allow the frequency for SR i
3.3.1.2.5 also to be 31 days and bases the 31 day frequency on the WRNMS " fixed incore detectors, overall reliab8lty, self-monitoring features.* (See Reference 3, page B 3.3-42).
Describe how the WRNMS provides indication of increasing reactivity to the operator when in mode 5. To what extent does the WRNMS self-test verify OPERABluTY of the components
)
which provide this indication?
l According to Reference 5, Section 3.9.5(1), the period trip (ramp) test is not part of the automatic self-test, but must be selected by the operator with the instrument in the Inop mode.
Does the operator rely on the WRNMS Period-Short function for notification of an increase in reactivity when in Mode 57 PECO ENERGY RESPONSE TO QUESTION 4 When the units are in Mode 5, the WRNMS would be expected to be in range 0. In range 0, i
both the count rate and period are displayed to the operator on the NUMAC and the operator i
display assembly. If the power level were to change such that system protective action was required, the automatic action would be initiated in response to a detected short period l
The current SRM uses analog processing, and in particular, uses analog processing for the i
period extraction function. In the WRNM, all processing except initial signal processing is
}
performed digitally. The period calculation is performed entirely digitally from the same signal that is displayed for count rata. Self-test monitors the signal access and operation of the CPU.
The path for indicating period is the same as that for indicating normal count signals.
i l
A period trip test can be performed off line, but it checks no additional hardware beyond that i
already covered by other tests. It does check the program stored in EPROM, but that is also monitored continuously by self-test, therefore, the period trip test is not actually required to be performed at all to confirm hardware OPERABluTY.
i The normal Channel Check will confirm that count rate indication is still reasonable, it confirms i
the total algnal processing path and, in corr,bination with self-test, confirms that the CPU j
processing is still being performed Together, these confirm OPERABluTY of all hardware required to indicate a short or decreasing period, which is the indication of unexpected reactivity s
changes in the core. With this coverage, an increase in the survetlance test interval (STI) to much greater than 31 days, or elimination of the surveRiance requirement entirely, is justifiable.
Therefore, the proposed increase to 31 days is conservative.
As an aid in setting the proper discriminator levels, and for accurately determining signal to noise ratio, the NUMAC WRNM utuizes a graphic on screen discriminator analyzer showing total pulses counted versus the discriminator threshold The discriminator levels are set digitally thus eliminating any drift in setting. Any significant change in signal to noise ratio due to introduction of EMI would cause the stable count rate and period indications to chan0s. The normal channel check will detect this short period which is also recorded on a strip chart recordw. Therefore, as in the case above, the proposed increase of STI for signal to noise ratio to 31 days is conservative.
4