ML20087A759
Text
,
4
.)
p ~g e.,
e
~'
V'
( ')
UNITED STATES '
)
ATOMIC ENERGY COMMISSION DIVISION OF COMPLIANCE 201 645 l
l*
REGION I 970 SROAD STREET NEWARK, NEW JERSEY 07102 February 4, 1970 R. T. Carlson, Senior Reactor Inspector Region I, Division of Compliance
,1 3
i JERSEY CENTRAL POWER AND LIGHT COMPANY (0YSTER CREEK-1)
C0 REPORT NO. 219/70-1 Only recently has JC been involved in grab sampling of the off-gas system.
It seemed that GE made no great effort to train JC personnel or furnish this type information until the activity increase in stack effluent on November 25, 1969.
Ross said that their new chemist has learned the off-gas sampling counting technique from observing GE personnel.
Relative to the increase in stack activity, I spent some time attempting to correlate primary coolant activity levels with the occurrence but was unable to make sense of the data due to varying power levels, intermittent operation of the cleanup system, etc.
For information, following is data extracted from a paper (which your have probably read before) by N. R. Horton titled Analytical Methods for W Evaluating the Radiological Aspects of the GE BWR, presented at the HP So-ciety Symposium, January 1969:
Dresden I Operating Data July 1969
'May 1963 Jan 1964 Mar 1965 Power MWT 620 620 530 375 Off-gas (uCi/sec, 30 min holdup) 300 2000 4000 60,000 Approx number defective fuel 2
5 6
15 ident assemblies 28 suspect 5000 8300 9,000 Average Burnup MWD /T Type Cladding Zr Zr+SS Zr+SS Zr+SS
{
9508070162 950227 PDR FOIA DEKOK95-36 PDR j
.l
.?
g
/
~C (j
')
. Stack Monitoring.
I have spent considerable time in trying to arrive at some " confidence level" for the stack monitor calibrations, not only for OC-1 but the data I accumulated will afford comparison with other GE BWR's.
At present, the off-gas grab sample calibration is a "one-point" calibra-tion at about 3500 uCi/second stack release rate.
Ross said they plan to i
accumulate additional data now that JC personnel can perform the sampling and counting. The assignment of 100 uCi/second stack release for one count per second on the monitor is in agreement with GE data and assumes this factor is constant over the range of the monitor. For the calibration with liquid Cs-137 performed by Kaulback this factor (uCi/sec per 1 cps) was not constant and varied from 1000 uCi/sec per 1 cps at a monitor indication of 6 eps; 400 uCi/sec per 1 cps at a monitor indication of 400 cps; to 100 uCi/sec per 1 cps at a monitor indication of 20,000 cps.
This means that at low counting rates (the background was about 5 cps) the change of I cps in count rate would mean a significant change in stack release.
If we consider the above as calibration errors then we must add the insunment errors. For a ratemeter a plus or minus 2% of full scale meter accuracy means, in terms of count rate, pbs or minus (about) 30% of actual count rate for a seven decade log scale, which is the type on the stack monitor and off-gas recorders.
You can see in the recorder trace attached to the report, that the pen trace, according to where it is on the chart, spans about 20 cps, due to a relatively short meter time-constant. Taking the average of the trace and using the oeviation of the trace makes the readings about:
a.
for the off-gas calibration: 3000 i 1000 uCi/sec b.
for the Cs-137 calibration:
24,000 t 4000 uCi/sec Assuming a continuous annual average release rate at these levels the total activity released during one year would be:
a.
for the of f-gas calibration: 95,500 1 31,500 Curies / year b.
for the Cs-137 calibration:
760,000 i 126,000 Curies / year.
The reason I included the above figures is that Kaulback totals the stack re-leases by making an " eyeball integration" of the tack monitor recorder trace and the errors over a period of one year could be at least that stated above.
1
l
(.
s 3-In summary, and based on the data available, I think the off-gas cali-bration is probably the more accurate of the two.
The calibration of
~
the GE reactor in India was reported as about 80 uCi/sec.
I will con-tinue to check with Caphton on the status of stack releases and we will eventually have some additional data from NMPC.
5' f
a s
4
' % / / bb ',hYn Leo flIh, s
ot am Radiation Specialist 1
_ - _ _ _ _ _