ML19322D213
| ML19322D213 | |
| Person / Time | |
|---|---|
| Site: | Crane  |
| Issue date: | 06/29/1978 |
| From: | Rogers L BABCOCK & WILCOX CO. |
| To: | Lawyer L, Geoffrey Miller, Ohanlon J METROPOLITAN EDISON CO. |
| References | |
| TASK-TF, TASK-TMR SOM-II-166, NUDOCS 8002100059 | |
| Download: ML19322D213 (11) | |
Text
Y V
P 0306 *
(, (,, bc,go T
\\M, [.. Wi Lsos F
~~ tal T G ( tdsss,u k P d ' """ * ~
o M
t"'.!'.@- tzs
'cuMewsh o A'.Ti oN chs /7s s
1 ca. su Weu e.o Locu pop Ty au 4~
wa cu s
.. _. me PLCASI EZFLY TO H: ATTACH D 3.f.F.
-,. m.,, ~, -
+ v u.,,...-
.....s TECHN CAL SU:TCRT, i;333 LYNCHB'JRG, VA.
/)/
A g.,%
onioiwr:..co ot wei s snow wes ime anty to
_,,,, /
4 5 a'.v enn N TO =
_ _ - -,-=
t g,P' dbb MTEf(VY W '
fW/bjNa:..... ----- !~ t ^'. !-
- A,
- * { {'
O OItvb m a nIIiR l \\.
y, v aa,.
(
Y
' N * ('t,*U..I (
4 ; O (*_
,) l *3 [
q,j \\ J*,
- [,n) (,'.[A 3 g,. ), &l.{[,a q }QQ Q g ;
Q,
[
=~
g.
%s u.j.gg RETURN THIS COPY TO SEf4 DER 8 0 0210 06f
[
C')
-.eo.Cd % m Babcock &Wilcox
%, cener.uon c,ou, P.O. Box 1260. Lyncheurg. Va. 24505 Telephone:(804)384 5111 June 29,1978 SOM-II-166 RD4-I-362 Mr. L. L. Lawyer Manager, Generation Operations Metrop611 tan Edison Company Post Office Box 542 Reading, ?A-19603 Mr. G. P. Miller Station Superintendent Metropolitan Edison Co=pany Post Office Box LSO Middletown, PA 17057 Mr. J. P. O 'Hanlon Superintendent, Unit I Metropolitan Edison Cc:pany Post Office Sox 480 Middletown, PA 17057 Subj ect: Recc mended Actions for Suspected Loose Parts in the Reactor Coolant Systcm Gentle =en:
The attached contains standard reco=mendations to the B&W Nuclear Stess Supply Systes operating utility. It consists of i==ediate actions to be taken should the loose parts =onitor ir.dicate a loose part in the Reactor Coolant System.
These reco=mendations apply regardless of the supplier of the loose parts monitors.
If you have any further questions, please do not hesitate to contact me.
Very truly yours, C
L. C. Rogers Site Operations Manager LCR/Sn4/ bay l
The Babcock & Wacox Company / Estabhshed 1867
O O
L. L. Lawyer G. P. Hiller 6/29/78 J. P. O'Hanlon cc:
L. R. Pletke W. H. Spangler G. K. Wandling J. G. Herbein R. M. Klingaman J. B. Logan J. L. Seelinger J. T. Janis J. D. Phinney G. A. Kunder bec:
D. G. Culberson (SIP #6/229) t l
0 9
l.
O O
RIC0!EIDID ACTIONS FOR A SUSPIETED LOOSE PARI' Due' to the serious consequences of potential material damage in the event of a loose part inside the Reactor Coolant System, it is stron61y recommended that a plan of action be developed by each utility to cover this situation.
Quick and decisive action by the operations personnel is required in the event of a suspected loose part to place the unit into a safe condition to avoid the potential of several months of repair.
Isose Parts Monitoring (LPM) Systems supplied by Bf.W are designed and calibrated to detect unusual noises. aboves the. normal. system background.
The Loose Parts Monitoring System is set to alarm for detected noises having the characteristics of metal-to-metal impacts. Regardless of the supplier of the Loose Parts Monitoring System, B&W recommends that i= mediate action be taken to determine the validity of the alarm. The alarm should be assumed to be the result of a loose part until proven to be otherwise.
If the alarm cannot be invalidated, the plant should be shutdown, cooled down, and placed into the decay heat removal mode to minimize the da= age that could be sustained due to the presence of a loose part(s).
i The types of questions that must be answered to determine the validity of the alarm include but are not limited to:
i 1.
What were the plant operations i==ediately prior to the Loose Parts Monitoring alarm? Did anything change abruptly or in a manner that
'could have caused the noise or that could have dislodged a loose part?
2.
Was the plant undergoing a significant evolution at or near the time when the alarm occurred (reactor trip, turbine trip, feedvater flov
~
transient, etc. )? Could the evolution attribute to the noise or cause a loose part to become dislodged?
3.
Do other plant instruments indicate abnormal readings that may be related to the noise?
k.
Are other Loose Parts Monitoring. channels alarming? Do the relative magnitudes of the signals from the various Loose Parts Monitoring channels indicate the location of the noise?
b the ansvers to the above questions confirm a loose part or fail to promote justification to ignere the alern, the plant ad=inistrative procedures should l
call. for (a) notification to key plant personnel of the presence of a poten-i tial loose part and (b) the timely shutdown of the plant. The manner in which the reactor is shutdown (i.e., orderly,. reactor trip, reactor and reactor coolas. pump trip, etc.) s,hould be dictated by the potential conse-quences of the particular alarm situation.
l
_1_
O O
~
REC 0K4 ENDED ACTIONS FOR A SUSPETED LOOSE PART The preservation of plant. integrity st.;'id take precedence over data sequisition in all cases; however, during the evaluation period, it is desirable that the following data be taken to provide a historical record of the alarm for action, justification, and later comparisons and analyses:
- 1.. Log all alam conditions, high or lov, on an alarm record sheet (see Attachment 1).
2.
Reset the alarm and log the tape-recorder didital counter.
3 It requi. red, adjust the amplifier gain for the maxi =um output without overload (adjust so average, signal is 20 ho percent full scale).. Tag the new gain setting and tape digital counter.
h.
For a lov alar = condition, defeat the low alarm option of the specific slarmed module.
5.
If any individual high alam occurs within 30 minutes of a previous high alarm, the shift supervisor and technical staff engineer should be notified.
A Upon notification of the Loose Parts Monitoring Syste= alar =s, the shift supervisor or the technical staff engineer should review the Loose Parts !bnitoring System alam record. The locations of the alarmed sensors should be cross-referenced with the plant conditions at the time of the alarm.
7.
Review the automatically recorded tapes and dete=ine the peak-to-
"G" levels of the recorded impacts. The operability and cali-rate.on of the tape auto start system should also be periodically checked.
8.
Use the audio =odule, oscilloscope, or spectrum analyzer to characterize the signals frc= the alarmed sensors. The points of interest should include metal-to-metal impact noises, the delay time matrix if more I
than one sensor indicates the noise, and the amplitude of the i= pacts (i.e., in "G's").
1
..w 9
Evaluate the cause of the alarm:
a.
Electrical: Most of the electrical noises are found to be periodic in nature and usually have individual wave forms with spike shapes and no amplitude decay.
b.
Mechanical,and if so, what is the probability that it was caused by nor=al plant operation?
4 t
O O
REOMME!!DD ACTIONS FOR A SUSPECTE LCCSE PART
- e.
Loose Part: Take immediate action to preclude further damage and contact MW.
- 10. Log the sensor gain setting, if changed, on the alarm record sheet to allow accurate correlatis with the recorded tape.
e 8
4 e
e S
e e
J 4
D e
e 9
3-m
--y
. we
-m w'-=--
7' v
~
o o
ATTACF24E!.T I e
EXAMPLE LOOSE PARTS M0!iITCRING SYSTE'4 ALARM RKORD SEEIT TIME OF ALARM CHANNELS ALARMG RDERKS REORDD BY OPERATOR ACTIONS, SENSOR xy.TH/ DAY / HOUR /MIIiviE FIRST OUT/OTHUIS ALARME GAIN & TAPE DIGITAL COUNTE4 SIGITATURE 9
.e l
G j
Q i
I l
1
' ' ') '/% )
'/W1, /k3
'jg
- S $ !A I4bM SITE INSTRU ION NO.
/
THE BABCOCK & WILCOX COMPANY DISTRIBUTION 8 M 3 M " '4 /7/'l 3,6,r,'up,SM, POWER GEllERATI0tt GROUP
- > 1A 12 M nr srO.wr., x iS r;m er g
Distribution f
DNMd N.- 78 --.
APPROVED.
From W. E. Wiison
/ f,e. W A T [ g h du/..{ g 6/2//F R. P. Williamson u s eas.s Cust.
File No. NSS-3,4,5,6,7,8,9, DPCO, Met Ed, JCP&L, FPC, AP&L, SMUD, CPCO, TECO or Ref.
11,12,13,14 / T3.72
~
Recomended Actions for Suspected Loose Parts in the RCS June 14, 1978 lm.i..,......,..................,.Distribution i
C. D. Russell R. C. Luken J. T. Janis C. E. Mahaney i
L. C. Rogers F. R. Faist
]
G. T. Fairburn The attached contains recommendations to utility customers for the immediate actior.s to be taken should the loose parts monitor indicate a loose part.
These reco:stendations apply regardless of the supplier of the LPM.
Please transmit the attached to your customers.
Any questions or requests for additional infomation should be directed to W. E. Wilson.
WEW/RPW/cs cc:
J. D. Phinney W. H. Spangler H. A. Baker
-~
L. R. Pletke REVIF.,WED7 FOR ACCURACY l
/
R.
r n - Midland gnavnggning f,,j,,
zc DATE (
W
- N Ig[
DATE: /
l rt R. J. Baker 7. R. Skillman G
J. C. Deddens P. E. Perrone D. cords Center Re F. Hallman l
l l
[..'
o o
~
RECOMMENDED ACTIONS FOR A SUSPECTED LOOSE PART Due to the serious consequences of potential material damage ir the event of a loose part inside the RCS, it is strongly recommended that a plan of action be developed by each utility t'o cover this situation.
Quick and decisive I
action by the operations personnel is required in the event of a suspected loose part to place the unit into a safe condition to avoid the potential of several months of repair.
Loose Parts Monitoring (LPM) Systems supplied by B&W are designed and calibrated to detect unusual noises above the normal system background.
The LPM system is set to alam for detected noises having the characteristics of metal-to-metal.
impacts.
Regardless of the supplier of the LPM system, B&W recommends that immediate action be taken to determine the validity of the alam.
The alarm should be assumed to be the result of a loose part until proven to be other-wise.
If the alarm cannot be invalidated, the plant should be shutdown, cooled down, and placed into the decay heat removal mode to minimize the damage that could be sustained due to the presence of a loose part(s).
The types of questions that must be answered to determine the validity of the alarm include but are not limited to:
1.
What were the' plant operations immediately prior to the LPM alarm? Did anything change abruptly or in a manner that could have caused the noise or that could have dislodged a loose part?
2.
Was the plant undergoing a significent evolution at or.near the time when the alarm occurred (reactor trip, turbine trip, feedwater flow transient, etc.)? Could the evolution attribute to the noise or cause a loose part to become dislodged?
3.
Do other plant instruments indicate, abnormal readings that may be related to the noise?
4.
Are other LPM channels alaming? Do the relative magnitudes of the signals from the various LPM channels indicate the location of the noise?
If the answers to the above questions confim a loose part or fail to promote justification to ignore the alarm, t'.1e plant administrative procedures should call for (a) notification of key plant personnel of the presence of a potential loese part and (b) the timely shutdown of the plant. The manner in which the reactor is shutdown (i.e., orderly, reactor trip, reactor and RC pump trip, etc.) should be dictated by the potential consequences of the particular 11 arm situation.
The preservation of plant integrity should ta e precedence over data acquisition in all cases, however, during the evaluation peri u it is desirable that the following data be taken to provide a historical record of the alarm for action, justification, and later comparisons and analyses:
1.
Log all alarm conditions, high or low, on an alarm record sheet (see Attachment 1)
O
()
l
^
2.
Reset the alarm and log the tape recorder digital counter.
3.
If required, adjust the amplifier gain for the maximum output without overload (adjust so average signal is 20-40 percent full scale).
Tag the new gain setting and tape digi'tal counter.
4.
For a low alarm condition, defeat the low alarm option of the specific alarmed module.
S.
If any individual high alam occurs within 30 minutes of a previous high alam, the shift supervisor and technical staff engineer should be notified.
6.
Upon notification of LPMS alarms, the shift supervisor or the technical staff engineer should review the LPMS alarm record. The locations of the alamed sensors should be cross referenced with the plant conditions at the time of the alam.
1 7.
Review the automatically recorded tapes and detemine the peak-to-peak "G" levels of the recorded impacts.
auto start system should also be periodically checked.The operability and calibration 8.
Use the audio module, oscilloscope, or spectrum analyzer to characterize the signals from the alarmed sensors.
8 The points of interest should include metal-to-metal impact noises, the delay time matrix if more than one sensor indicates the noise, and the amplitude of the impacts (i.e., in "Gs").
I 9.
Evaluate the cause of the alarm:
r A) Electrical; Most of the electrical noises are found to be periodic in nature and usually have individual wave forms with spike shapes and no amplitude decay.
E.
~
B)
Mechanical and, if so, what is the probability that it was caused by Il normal plant operat'on?
C) Loose Part; Take 'mmediate action to preclude further damage and contact
.: l B&W.
I:
10.
Log the sensor gain stetting, if changed, on the alarm record sheet to allow 4
accurate correlation with the recorded tape.
]
7
.i"$
e t
i
o 9.-
o
- >=uM - 1 EIAMPLE LOOSE PARTS MONITORING SYSTEM ALARM RECORD SHEET T1wE OF ALARM CHANNELS ALARMED REMARKS RECORDED BY j Operator Actions, Sensor
- t h /D.1y /liou r / Minute First Out/Others Alarmed: Gain & Tape Digital _Cpunter' Signature I
1 e
.