ML20003C570
| ML20003C570 | |
| Person / Time | |
|---|---|
| Site: | Pilgrim |
| Issue date: | 02/04/1981 |
| From: | Morisi A BOSTON EDISON CO. |
| To: | Grier B NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION I) |
| References | |
| 81-18, IEB-80-17, NUDOCS 8103060476 | |
| Download: ML20003C570 (10) | |
Text
I TA BOSTON EDISON COMPANY asusn46 orreces eco novostou erns:T SceTON. M AemacMue Tre02199 A.V.MORISI NuctcA= apen4TS*i su,", ORT DEPARTMENT February 4, 1981
^'
BECo Ltr. '
+1
/
S Mr. Boyce H. Grier
(
E u.., kgfg2 hg,gI Office of Inspection and Enforcement 4
Region I
_A
/
U.S. Nuclear Regulatory Commission b\\
'% %r b g,. /
p 631 Park Avenue
/,
King of Prussia, PA 19406 gj
_~
License No. DPR-35 Docket No. 50-293 Response to I.E. Bulletin #80-17 Supplement #4
References:
(A) Telecon with BECo and NRC I&E on 12-24-80 (B) Telecon with BECo and NRC I&E on 1-05-81
Dear Mr. Grier:
Supplement No. 4 to IE Bulletin No. 80-17, " Failure of Control Rods to Insert During a Scram at a BWR" requested actions to provide assurance that the con-tinuous monitoring system (CMS) for the scram discharge volume (SDV) has been tested to demonstrate operability as installed, remains operable during plant operation, and is periodically surveillance tested to demonstrate continued operability. Boston Edison Company's response to the requested actions is as follows:
ITEM 1.
Bench Test of CMS Make available the following infonnation which describes the CPS design and the bench tests which have been performed to demonstrate system operability and sensitivity:
(a) System description including a schematic of the apparatus and associated electronics.
(b) Type of sensing device and characteristics (include response characteristics versus temperature).
(c)
Calibration criteria, including transmission losses.
(d) Training and testing of personnel performing the calibration test.
Items (a) through (c) above may be referenced by the licensee if the information has been submitted to the NRC by the equipment manufacturer, k 82os oso 476
.-_..2 r
CCCTON EXICON COMPANY Mr. Boyce Grier Februsry 4, 1981 Page Two i
r
Response
1.
General Electric Company submitted information to satisfy items
(
(a) through (c) above to Mr. W. R. Mills of the NRC's office of Inspection and Enforcement cn December 12, 1980.
However, for completeness we have included additional plant information in t
regard to items (a) and (c).
l (a) 1.
System Descriotion:
The system imolemented provides for the installation of four (4)
Ultrasonic Level Monitors (two in each Scram Discharge Header SDV)i The system is composed of:
Four (4) Ultrasonic Level Monitors, two per each Scram Discharge I,
i Four (4) U.T. Signal Conditioners, one for each transducer, in-stalled in the Cabinet Racks.
i System Annunciator and Status indicators, in the Control Room.
Cabinet Racks (2), in the Reactor Building, E1.23'-0" (by the CRD drives), one for each EAST and WEST header bank.
Power feeds with associated circuit breakers and fuses.
The U.T. Transducers mounted on the SP will sense a water level in the scram discharge header greater than 1.25 + 0.25 inches. The signal conditioners will generate trip signals whenever one and/or both of the following conditions occur:
u
- SDV high water level y
e
- Loss of hignal (loss of wall thickness) or power failure The above alar;i conditions will cause activat'on of alarn signals on annunciator' windows 9.1 and 9.2 (on Nucleat' Control Panel C905),
as well as status (lights) indication of the r)nitorino channels (on a lamp cabinet above C905). The system ir non-safety grade (non-class IE :nformation system), but is redtfadant to provide re-liable indication of SDV condition.
Y
^
Ec; TON E% ICON COMPANY Mr. Boyce H. Grier February 4,1981 Page Three As stated above, the installed system is non-safety grade, powered from two 120 VAC sources:
17L Emergency Lirsting Panel for sensors /
conditioners supply power, and Panel Y1 for scatus lanps (Instrument BUS).
Upon receipt of high water level alarm annunciation for both SDV's in the Control Room Panel, and confirmatory. status lights, plant operations will follow established procedures for orderly shutdown and problem investigation.
The layout and installation of the system equipment, local racks, conduits, etc. is such that no effect upon existing plant safety-related equipment is allowed.
Two annunciator windcws in the Control Room Panel C905 have engravings similar to "SDV.
EAST TROUBLE" and "SDY. WEST TROUBLE".
Eight (8) status lights (two per transducer channel) will indicate "HIGH WATER LEVEL" and." EQUIPMENT FAILURE" for each of the four (4) signal conditioners. See attachment #1 for the schemtaic diagram.
(a) 2.
Comoliance with Vendor Recommendations:
PNPS-1 installation followed the Vendor-supplied installation speci-fications, including:
Use of brackets specifically designed for holding the UT sensors with required alignment and pressure.
Renoval of scale and paint from the pipe surface at the contact area.
Use of calibration fixture (6" nom. diameter c.s. pipe, sch. 80),
to match actual conditions at the field installdtion on the headers.
(b) 1.
Desian Differences:
a) Diversity of Power Supplies - Instead of one, two 120 VAC sources were used:
120 VAC (Instrument Bus) for status lamps circuitry, and 120 VAC (Panel 17L) for UT Water Sleuths Modules.
The original GE design called for a single power source. The modification intro-duced at PNPS #1, enhances the capability of status indication at the Cor.crol Room, in the event of a power failure at the modules.
b) Test of Status Lamps - Two " test" buttons were included in the PNPS-1 CMS design, to allow on-line testing of the status lamps' circuitry.
This feature was not present in the vendor design.
COOTQN EDCON COMPANY Mr. Boyce H. Grier February 4,1981 Page Four (b) 2.
U.T Sensors Coupling Compound:
The UT Sensors at PNPS-1 were tested, calibrated, and field installed using Alemite Type-H Lithium high temperature grease.
This compound provided adequate UT coupling, as verified by accurate header wall thickness channel measurements.
The monitoring modules at PNPS-1 provide a wall thickness aralog output (voltage), and high and low alarns that monitor the proper sensor coupling to the pipe. Any deviation from normal expected cut-put will trigger a trouble alarm (annunciator) and status indication in the Control Room.
(b) 3.
UT Sensors Location on the Scram Discharge Headers:
GE recommended that the UT Sensors be located "within two feet of the lower end of the header, without any other constraints.
At PNPS-1, each header bank (East and West) is comprised of two sloping dcwnwcrd C.S. pipes (6" pipe disc. Sch. 80) joined together by 6 transverse; 4
6" pipe. During field tests, a gross amount of " cross-talk" was detected when both sensors were placed close to each other (at tF; joining 6" "T" piece). To reduce the " cross-talk" phenomenum,'one of the UT Ser. ors was repositioned back along the 6" transverse piece, until a significant reduction in the noise was obtained, while still complying with the requirement of a low-point location on the sloping header pipe.
(b) 4 UT Beam Penetration Caoability:
The UT Sensors were tested on the calibration fixture, with water ranging from zero up to filled pipe. conditions, and no functional problems were encountered.
During the last outage (Spring '.360), the whole Scram Discharge Header System was hydrilazed, to remove any extraneous build-up of residues. A reductior in attenuation or dispersion of the UT beams 4
at PNPS-1 should not be expected, due to existino deposits.
(b) 5.
Sional-To-Noise Ratios:
Due to noise problems during field tests, it was decided to route each coaxial signal cable in separate conduits. The modification resulted in no spurious tripping and "raisonable" signal to noise ratios.
9-g,
-m
CO2 TON EDCON COMPANY Mr. Boyce H. Grier February 4,1981 Page Five (c) 1.
Calibration Criteria:
Implemented by a PNPS-1 Calibration and Acceptance Procedure, the bench calibration, test, and field functional acceptance criteria called for:
- 1. - Calibration of the four UT Water Sleuths Model 280, with the respective UT sensors. Every UT Model 280, tagged together with
~
a respective UT sensor, underwent a series of calibration steps
[
(as recommended in the UT Water Sleuth Model 280 Operating Manual).
. Adjustment of alarm set-point voltages and corresponding verifi-cation of contact output condition.
t
- Aojustment of both water depth and wall thickness channels cir-cuitry.
For the wall thickness channel, standard steel wedges were used.
For the water depth channel calibration, the specially-built calibration fixture (carbon steel 06" pipe, Sch. 80) was employed, filled to the appropriate water levels.
A UT coupling compound, proven to be suitable for UT coupling and withstanding the field temperatures was selected. Video cuputs were recorded for every module /sensnr pair, for later field in-stallation acceptability and surveillance procedures. The output voltage from the wall thickness channel was useful in " measuring" the thickness of the pipe wall, to verify the quality of the coupling method.
To eliminate the signal attenuation effect betwten calibration and i
actual field installt. tion, the same type and field length of coaxial cable was used during the calibration procedure.
- 2. - Field Installation Test (before and after UT sensor installation).
Every monitoring channel was functionally tested, using the cali-bration fixture to simu,' ate the SDV headers, verifying appropriate Control Room indications v.d alarms, f
C ZTON EZON COMPAN'T Mr. Boyce H. Grier February 4,1981 Page Six After the UT sent 's were placed on the headers, every wall thickness output (video and analog) was compared to bench-obtained data, as an acceptable proof of a good UT coupling.
The system was installed, tested and calibrated by technicians qualified to ANSI N18.7-1976 and a Level III NDE examiner with certification in accordance with SNT-Tc-1A-1975.
ITEM 2.
Ope-ability Test of CMS Prior to conducting the operability test, verify that the CMS on the SDV is installed and calibrated in accordance with the vendor recom-mendations.
In order to provide assurance of operability of the CMS, if not already performed conduct an operability test within 14 days of the date of this bulletin.
In this test, inject a sufficient amount of water into each SDV header to determine that the ultrasonic transducers are adequately coupled to the SDV piping and that the trip alam function of the CMS will perform satisfactorily. The test may be perfonned by a single (multiple) rod scram tests while operating. No water may be introduced into the SDV header shilt. the reactor is operating except using the scram function.
Independent level measuranent must be used to verify CMS operation and proper calibration.
Response
- 2. As per telecon with the NRC on January 5,1981, (ref.8) the CMS was I
tested by single rod method and all four sensors were witnessed to respond to water.
The water level indication was verified by utilizing an oscilloscope response which monitored individual CMS sensor output.
In addition, PNPS experienced a full reactor scram on January 28, 1981, during which the CMS instrumentation was verified to function as re-quired.
c ITEM 3.
Interim Manual Surveillance In the interim 14-day period before the operability test is completed, nerform a manual surveillance for the presence of water in the SDV at least once per shift and after each reactor scram.
In order to provice assurance that manual surveillance can detect water accumulation in the SDV, verify that the method and the operator have been qualified by testing which uses or simulates the SDV piping and has the ability to detect differenct levels of water in the SDV.
Surveillance of SDV manual measurement techniques should be done before completion of the operability test described in Item 2 above.
db!'dhy"ce If'$YikE" February 4,1981 Page Seven
Response
3.
As per the telecon (ref. A ) with members of your staff involved in the review of this bulletin, pennission was given to PNPS to continue to utilize the continuous water level monitoring system for the Scram Discharge Volume (installed and made operational pursuant to Commis-sion Order dated October 2, 1980) in lieu of the manual surveillance requested by this item.
ITEf1 4 Full Test of CMS to be Conducted During a Planned Outage P
During a planned outage within six months, perfarm a full CMS test using the SDV headers:
(a) Admit water into the SDV to establish fill rates for several (not less than three) in-leakage flow rates.
The in-leakage
~
rates should range from approximately the minimum which results in water accumulation in SDV to a full scram.
(b)
Establish and record the response of the CMS indication and alarm functions from the trip level to a full SDV.
Provide criteria for replacement or adjustment when exceeding design specifications of the system.
(c)
Verify by independent measurement that the alarm initiates at the proper level setpof.t.
Response
4 As per the January 5,1981 telecon (ref. B ) we are waiting for the response from the NRC regarding test methodology and necessity.
Our proposal called for a single fill to the alarm point and a relaxation i
of tne six month test to one which coincides with the planned refueling outage commencing on or about September 30, 1981. We also cansidte this requirement to be a one time only test.
We will consider our alternate g oposal to be acceptable to you unless informed otherwise.
l ITEM 5.
Operability of CMS During Reactor Operation The CMS shall be operable prior to reactor startup and during reactor opera tion.
If the CMS becomes less than fully operable, within 8 hours9.259259e-5 days <br />0.00222 hours <br />1.322751e-5 weeks <br />3.044e-6 months <br /> perform a manual check for water in the SDV and institute procedures for a manual check of the SDV each shift and following scram until the CMS is fully operable. When not fully operable, the CMS should be used to the extent practical in addition to the manual checks.
COOTON Ect:ON COMPANY Mr. Boyce H. Grier February 4,1981 Page Eight If the CMS is not operable within 7 days, the frequency of the manual check should be increased to once every 4 hours4.62963e-5 days <br />0.00111 hours <br />6.613757e-6 weeks <br />1.522e-6 months <br />.
If the CMS is not operable within 30 days the plant shall shutdown.
To demonstrate continued operability of the CMS during reactor oper-ation, perform periodic surveillance tests for operability of the CMS.
For these oeriodic surveillance tests, test as much of the CMS as practical during reactor operation without injecting water in the SDV.
Establish criteria for repair or replacement when the system design criteria or estimated service life limitations are exceeded.
The frequency of these periodic surveillance checks should be determined by the licensee.
These periodic surveillance tests should include the folicwing:
t (a) determination that the response and power output of the transducer l
has not degraded; (b) visual inspection for adequate condition of the transducer to SDV coupling material; and (c}
a calibration check of the electronics to assure alarm initiation in the control room.
Water should be periodically injected into the SDV to perform a CMS operability and calibration check similar to that specified in Item 2 above. This check should be performed semiannually and during startup after plant outages where maintenance operations may have taken place near to CMS equipment.
Response
f 5.
PNPS procedures 2.2.122, Scram Discharge Volume Level Monitoring,and 2.2.15, Daily Surveillar.ce Log, have been implemented or revised to comply with the requirements listed in the first paragraph of Action Request 5 of I.E.Bulletin 80-17, Supp. 4 However, we request that the manual check frequency called for in the second paragraph be the l
same as that called for in I.E.Bulletin 80-17, Supp. 3 which required a once per shift check. Unless informed otherwise, we will consider our proposed once per shift frequency to be acceptr_ble.
The periodic surveillance tests delineated in the bulletin will be performed at the interval specified as follows:
A&B: The surveillance for response and power output of the transducer is continuous and automatic with an alarm in the control room to warn of system degradation.
t a
COLTON EDCON COMPANY Mr. Boyce H. Grier t
February 4,1981 Page Nine i
A,B,&C: Procedures are being prepared and are scheduled to be in place by February 28, 1981.
The proposed surveillance for 5 A,B & C will be quarterly with the periodic injection requirenent conducted once per cycle starting with the scheduled 1981 Refuel Outage. The proposed surveillance frequencies were also discussed during the telecon of January 5,1981, (ref. B ).
Unless informed otherwise, we will consider our proposed frequencies to be acceptable.
ITEM 6.
Operating Procedures Develop procedures for operation, periodic testing and calibration of the CMS and for repair or replacement when system design specifications are exceeded.
Develop procedures for the calibration and use of the hand held UT device in the event of a malfunctioning CMS.
Notify the NRC before changing the established CMS alarm level setpoints.
Response
6.
Operations procedure 2.2.122 Scram Discharge Volume Level Monitoring was in effect December 3, 1980 and procedure 2.1.15 Daily Surveillance log has been revised to include a daily check and verification c' oper-abil,ity of the CMS.
In the event the CMS malfunctions, due to ALARA considerations, the hand-held UT devices will not be utilized; however, the sensors in-stalled prior to the CMS installation have been left in place and will be monitored manually using procedure TP 30-68 which will be revised and incorporated into a permanent procedure.
The procedures developed in response to Item SC will contain a require-ment to notify the NRC concerning set point changes.
In addition to the information supplied above, please take note of the following:
Revised response to I.E.Bulletin 80-17, Suco. 1 In our response to the I.E. Bulleting 80-17, Supp.1, Action Item 2d, we stated that the time required to fill the instrument volume from scram initiation to full scram on the RPS level switches was 96 seccnds for the automatic scram.
Subsequent review of the data revealed the time should have been 93 seconds.
=
COOTON EDCON COMPANY Mr. Boyce H. Grier February 4,1981 Page Ten i
i l
We believe this information adequately addresses the concerns of the subject i
bulletin. Should you have any additional questions or concerns, please do not t
hesitate to contact us.
i i
Very truly yours,
[
-i
/f / (' bl
?l f
,L//;
t Commonwealth of Massachusetts)
County of Suffolk
)
i Then personally appeared before me A. Victor Morist, who, being duly sworn, did state that he is Manager-Nuclear Operations Support of Boston Edison i
Company, the applicant herein, and that he is duly authorized to execute and file the submittal contained herein in the name and on behalf of Boston Edison Company and that the statements in said submittal are true to the best of his knowledge and belief.
7 l
///
t My commissio..n e.xpires:
/Am
/M,i p' ~o g<e i
g..
.,T. _,.._. __
Notary Public# /
j L
l i
cc: Director, I
Office of Inspection and Enforcement l
U.S. Nuclear Regulatory Commission Washington, D.C.
20555 l
i f
i s
v I
e l
i
-