ML20023A528
| ML20023A528 | |
| Person / Time | |
|---|---|
| Site: | 05000000, Point Beach |
| Issue date: | 07/25/1977 |
| From: | Shao L Office of Nuclear Reactor Regulation |
| To: | Eisenhut D Office of Nuclear Reactor Regulation |
| Shared Package | |
| ML19302F297 | List:
|
| References | |
| FOIA-82-309 NUDOCS 8210070205 | |
| Download: ML20023A528 (2) | |
Text
, _ - - _ _ _
i UNITED STAI ES I
[#.
.\\
NUCLEAR REGULATORY COMMISSION
/
}
WASHINGTON. D. C. 20555 t
y &{
JUL 2 51977 MEMORANDUM FOR:
D. G. Eisenhut, Assistant Director for Operational Technology, DDR FROM:
L. C. Shao, Chief, Engineering Branch, D0R
SUBJECT:
POINT BEACH UNIT fio.1 - STEAM GEf;ERATOR TUBE LEAK The Engineering Branch, Division of Operating Reactors, has made an inquiry about' the recent tube leak incident at Point Beach Unit No.1 and has reviewed the latest inspection results contained in the Annual Operating Report for 1976.
Following are a summary of findings and our comments:
1.
The latest leak started on June 21, 1977, and by June 24, 1977, the rate increased to 200 GPD (o.14 GPM) in generator-1B.
The leakage y
rate continued to increase to 500 r,PD (0.35 GPM), within next 24 i
hours, which is the limit alloEed b Q chnical Specifipation.
W W M 2.d& W D 2.
The reactor was then shutdown to locate the source of the leak.
It was found that the leak came from a plug.
The plug was then manually repaired by welding.
3 Subsequent to the repair of the leaky plug, the steam generator was hydro-tested at a secondary-to-primary differential pressure of 800 psi.
4 During the hydro-test, two additional leaks were found; one at another plug 0 one drop per minute, and other from an unplugged tube 0 one drop per three minutes.
These two tubes were plugged and the unit was returned to power operation. ~
5.
On November 10, 1976, during an inservice inspection of the reactor vessel, a loose tube plug was found and recovered from the bottom of reactor vessel.
6.
The manner by which the small leak from an unplugged tube was discovered resembles very much the situations experienced at Turkey Point Unit 4 and Surry Unit 1 for leaky dents.
7.
The 1976 Annual Operational Report showed that, in S/G l-A (Unit 1-A Generator), 23 row 1 tubes were examined and 15 of them show dent signals up to 9 mils at top support intt..: actions.
In steam generator 2-B, row 1 tubes were not examined.
- Overall, 142 out of 170 tubes on the hot leg side of S/G l-A examined have dent indications, and 102 out of 102 on the hot leg side and 93 out of 102 on the cold leg side in S/G 2-B examined showed denting.
821007020S 820910 PDR FOIA LONG82-309 PDR
" h O m t e W h h fir m s h e h ie w e-e:n e w *n+ m - +
t o
o JUL 2 51977 8.
The maximum indication was about 10 mils, and that the larger dent indications were predominantly on the hot leg sides of all six support plate intersections.
We have also checked with Region III l
inspector to find out whether or not inspection for flow slot hourglassing was made.
He indicated that the licensee had made inspection through the handhole and found no visible hourglassing of flow slots.
9.
Based on the inspection results summarized above, the tube denting in Point Beach Units 1 & 2 can be considered moderate.
Althouah we do not see an ir*,ndiate safety concern. the conditions at these two l
units warrant a clnwr watch.
We recommend that the liconsee be requested to submit an inspection program for next scheduled inspection for our review.
h
]
/ /
V_
n
/\\
v'd L. C. Shao, Chief Engineering Branch Division of Operating Reactors cc:
R. Stuart J. Guibert R. Cudlin F. Almeter B. D. Liaw P. Wagner 3
a i
&~1.i2:i;, @&QG:-Eg@L&Qi?:=@Mengf.fyj%g9MMM=U?b?-G2#G2@ML f
[paarc'4 o /
o UNITED STATES 7J s'~
g
- q
't NUCLEAR REGULATORY COMMISSION f.x, @, j WASHINGTON, D. C. 20555 j
y S ' f* b E
'%.,'[.. #
AUG 181977 u'
MEMORANDUM FOR:
R. S. Boyd, Director, Division of Project Management, NRR R. J. Mattson, Director, Division of Systems Safety, NRR H. R. Denton, Director, Division of Site Safety and Environmental Analysis, NRR FROM:
Victor Stello, Jr.,, Director, Division of Operating Reactors, NRR
SUBJECT:
INFORMATION MEMORANDUM NO. 3 FAILURE OF RADIATION MONITORS Identification of a significant leak in a steam generator tube at Point Beach was delayed for many minutes due to the fact that two process monitors failed to respond as expected for this type event.
Indications showed that the condenser air ejector discharge monitor, which detects activity in the steam system resulting from steam generator tube leaks, spiked momentarily causing an alarm but then went downscale when the monitor saturated (Tracerlab Model MD-12C G. M. detector).
Control room personnel were misled in assuming that the event causing the alarm had terminated since saturation of the monitor clouded the evidence that primary to secondary leakage had occurred.
(Portable radiation monitoring units later estab-lished the actual system reading at 1R/hr at contact with the air ejector filter.) During this time, the steam generator blowdown monitor (Tracerlah Model MD-5B Scintillation Detector) also caused a spike and appears to have saturated when it also went downscale.
However, because of low sample flow past the monitor, there was a long delay in causing the response of the monitor to perform its intended function of isolating the blowdown lines during the incident.
As a result of the instrument malfunctions, a small amount of radioactivity was released to the environment which could have been reduced by automatic control features or operator action.
The air ejector monitor, which employs a Geiger counter as the radiation detector, was tested and showed that the output signal of the G. M. tube had saturated the counting system.
Tests showed that the G. M. tube, which provides pulses having a small charge and a rate of occurrence related to the rate of intercepted ionising event was suddenly intercepted by a very large number of ionising events (i.e., high radiation dose rate from activity on the filter adjacent to the monitor).
Consequently, instead of
Contact:
S. Block, EFB/ DOR X 28066 x
-i.-m
s_
_ generating discrete numbers of small charged pulses for which the system was designed, a much larger current was produced.
The elec-tronic circuit ry employed for this monit oring syst em precluded t his current from reaching t he count rate meter (i.e., a syst em high l
voltage blocking condenser, also blocked the d.c. current to the pre-amplifier).
Thus, the count rate meter which, at the instant of high exposure triggered the high radiation alarm, now started to if read downscale.
It thus appeared to the control room operator that
- C.
the event which had caused the alarm had terminated.
The Steam Generator Blowdown Monitor, whose intended function is to isolate the blowdown and sampling lines, did not indicate a radio-active discharge due to the long delay time because of primary to secondary mixing and transport time of the liquid to the detector, j
The alarm and valve trip thus did not respond fast enough and had still not responded before manual shut-off was effected on the steam generator blowdown.
Investigation also showed that there was little or no flow through this off-line monitor.
As a result of the failure of the two improperly designed process moritors whose function was to warn personnel of steam generator tube failure, radioactivity was released to the off-site environment and the reactor was operated with a degraded primary / secondary boundary for a significant period of time.
The total airborne activity released from the time of tube failure until reactor start-up is estimated to be about 30% of,3 pe technical specification annual average release rate t
of airborne Xe equivalent radioactivity (about 3755 Ci).
The total liquid regeased during this ceriod was about 1 Ci of S,Y activity and l3TXe,133m I and 13 concentrations were 68, 5 and 110 Ci of H.
Tne 1.2% respectively of 10 CFR 20 limits.
These releases would have been substantially higher had the condenser been subsequently lost as a heat sink.
Corrective Action After examination of the liquid and gaseous radiation monitoring system and procedures for their use with respect to the characteristics of the incident, the following corrective actions were taken:
1.
An electronic modification was installed in the air ejector effluent monitoring system to preclude the downscale deflection of the count rate meter whenever high radiation levels are encountered.
The modification consists of an oscillator circuit installed in parallel with the circuit pre-amplifier so that when the G. M. tube saturates (i.e., generat.es d.c. current), the oscillator will be set in a mode that will keep the count-rate indicating device upscale.
K
~
.5 cr
.m
_m
. Changes in the steam generator blowdown monitor consist of:
2.
An increase in the sample flow and path past the present a.
detector by increasing the sa=ple line size. This will enhance the surveillance of the flow.
1 Installation of a redundant strap-on monitor to the liquid
.a}$
b.
I discharge pipe of the generator blowdown tank common dis-charge line.
High radiation signals from either monitor qqg{
a) or b) will result in automatic isolation of the steam blowdown effluent.
DOR has evaluated the cause of the electronics failure and finds that the corrective action taken by the licensee with respect to the electronics modification is acceptable. However, we would recommend that if a G.M.
counter is to be used in an effluent or process monitoring system, non-saturating tubes should be considered. That is, those G.M. tubes that can operate in a current mode (an integrating tube) at very high counting These tubes are also made to withstand the bombardment they receive rates.
from high ion densities.
The G.M. tubes normally used for monitoring low radioactivity levels of effluents (i.e., non-integrating) would disin-tegrate rapidly and tube failure might rerult if they were continuously operating in a current mode.
By suitable choice of electronics, the out-put current of G.M. integrating tubes can be made proportional to the radiation intensity and can be measured with a microammeter which should be in parallel with a more sensitive pulse counter.
In any event, these options should be reviewed to assure that radiological monitoring systems employing G.M. counters will not cause downward deflection of the count rate meter when operating in a current mode or become degraded due to continuous bombardment of ions within the counter under anticipated opera-l Additionally, when tional occurrences and during and following an accident.
l off-line process monitors are used, sample flow and path to the monitoring system should be designed to ensure that the detector is continuocsly A monitoring monitoring a representative sample of the effluent stream.
system that has an isolation function should be designed so that failure of the monitor will cause the isolation function to be transferred to a parallel system.
In summary, when considerating the range of normal operations, anticipated operational occurrences and accidents, the range capacity of the monitoring system should cover these extremes either by use of a single or dual monitor-DOR is requesting that IE examine this problem on all operating ing array.
PWR *s.
)
Victor Stello /Jr., Director Division of Operating Reactors cc:
See following page
7 y
'r g
! 'J
.p "
l cc:
E. Case jk-,
l p[e?
l S. Hannuer Deputy Directors, NRR Assistant Directors, NRR DOR Branch Chiefs
- F;l
- n a r W. Kreger 13 J. Collins jdn D. Bunch
@/7 DOR Section Leaders
(,
L. Nichols i
H. Thornburg DISTRIBUTION Central File EEB Reading J. Guibert R. Cudlin f
r#
"2 N
A(
bk c
I
'D^
AD/GTN..
ya D@e s
p.
. QR..
uy EEB/0T/p0R EEB/0T/29R.
E E '
.),.
2, j-
_3...
+..p. 4
..y y..
SBlock p ch
/rtega g
..;e s.
DEisenhut VS]llo
- y 8/3/77
/!-//77 S/f/77 8/ f
/77 E //5' / 77
[7;{h,h e,,.
'-!*t4 D un e. aeva ne.wrNr emiNTiNe or recti t o7e - eas.ed
'3#jg NKC FORM 3tb (9 4) NRCM O240
% =c.
..,_Et.
_ _ _ _ _ _ _ _ _ _ _ _ _ _ _ _.. _.. _ _ _. _ _.. _.. _ _ _ _ _ _ _. _ _ _. _ _