IR 05000275/1985010
| ML17083B552 | |
| Person / Time | |
|---|---|
| Site: | Diablo Canyon  |
| Issue date: | 04/02/1985 |
| From: | Hamada G, Yuhas G NRC OFFICE OF INSPECTION & ENFORCEMENT (IE REGION V) |
| To: | |
| Shared Package | |
| ML17083B551 | List: |
| References | |
| 50-275-85-10, 50-323-85-09, 50-323-85-9, NUDOCS 8504260160 | |
| Download: ML17083B552 (14) | |
Text
U.
S.
NUCLEAR REGULATORY COE1ISSION
REGION V
Report Nos.
50-275/85-10, 50-323/85-09 Docket Nos.
50-275, 50-323 License No.
DPR-80, CPPR-69 Licensee:
Facility Name:
Pacific Gas and Electric Company 77 Beale Street San Francisco, California 94106 Diablo Canyon Units 1 and
Inspection at:
San Luis Obispo County, California Inspection conducted:
February 11-14, 1985 Inspector:
G.
H.
da, Radiatio aboratory Specialist Date Signed Approved By:
G.
P.
u s,
hief Facilit'adiological Protection Section Da e
igned Summary:
Ins ection of Februar 11-14 1985 (Re ort Nos. 50-275/85-10 50-323/85-09)
Areas Ins ected:
Routine announced inspection of radiochemical analysis programs and associated quality assurance procedures and practices.
Inspection also included split sample test analyses involving the Region V
Hobile I,aboratory.
This inspection involved 32 hours3.703704e-4 days <br />0.00889 hours <br />5.291005e-5 weeks <br />1.2176e-5 months <br /> onsite by one inspector.
Results:
No items of noncompliance or deviations were identified.
SS04nOSao 8SO402 PDR
- DOCK 05000275
DETAILS 1.
Persons Contacted AE s,D
"(C.
"R R.
'"W
"R
">> W.
"B
'R
"-W.
"A
~ D
"R
"R.
'D-R.
"<W.
AL Conway, Personnel and General Services Manager Clifton, Training Supervisor Eldridge, NPO gC Manager Giffin, Supervising Engineer Garacci, Senior Systems Analyst Johnson, Foreman, Effluent and Rad-Chem Kaefer, Assistant Plant Manager McLane, Materials and Project Coordination Manager Nanninga, Senior Mechanical Maintenance Engineer O'ara, Senior Chem and Rad Protection Engineer Peterson, ISC General Foreman Powers, Senior Chem and Rad Protection Engineer Rapp, Chairman, OSRG Sisk, Regulatory Compliance Engineer Taggart, Acting Director - equality Support Thornberry, Plant Manager Todaro, Security Supervisor Unger, Chem and Rad Protection Engineer Weinberg, News Services, NPO Wogsland, GO Unit 2 Womack, Engineering Manager
"-Indicates personnel present for the exit interview.
2.
Discussion This inspection involved a review of radio-chemical and radioactivity measurement procedures and practices, and the associated quality assurance programs.
Direct split sample measurement tests were conducted with the Region V Mobile Laborator P
The test results are summarized below.
Table
Reactor Coolant Li uid (Sampled at 3:10 a.m., February 12, 1985)
Nuclide Na-24 Cr-51 tfn-54 Co-58 Co-60 I-131 I-132 I-133 I-135 W-187 Diablo uCi/ml 4.58 E-3 1.05 E-3 1.23 E-4 4'3 E-4 3.88 E-4 3.82 E-4 5.07 E-4 2.01 E-3 NRC uCi/ml 4.46 E-3 2.15 E-4 3.67 E-5 1.67 E-4 1.58 E-5 2.18 E-5 3.42 E-4 3.63 E-4 5.32 E-4 1.87 E-3 Ratio Diablo/NRC l. 03 4.88 3.35 2.59 1.13 1.05 0.95 1.07
"-Agreement
~Ran e
0.80 - 1.25 0.60 - 1.66 O.6O - 1.66 0.75 - 1.33 0.50 - 2.00 0.6o - 1.66 0.6o - 1.66 0.80 - 1.25 0.6o - 1.66 0.80 - 1.25
"See enclosure for agreement criteria.
Table 1 gives the results for the first test for reactor coolant liquid analysis.
The data indicate that although good agreement was achieved for those nuclides with activities sufficient to provide good counting statistics, nuclides with concentrations near the detection limit range for this particular measurement, showed poor agreement.
Diablo performed a recount of this sample using a one hour interval instead of the 10 minutes used earlier for the first count.
These results are summarized in Table 2.
Table
Reactor-Coolant Li uid (Recount)
(Sampled at 3:10 a.m., February 12, 1985)
Nuclide Cr-51 Mn-54 Co-58 I-131 Diablo uCi/ml 2.03-E-4 5.64 E-5 1.58 E-4 3.54 E-5 NRC uCi/ml 2.15 E-4 3.67 E-5 1.67 E-4 2.18 E-5 Ratio Diablo/NRC 0.94 1.54 0.95 1.62 Agreement
~Ran e
0.60 - 1.66 o.6o - 1.66 0.75 - 1.33 0.60 - 1.66 The data indicate that not only was agreement achieved for those nuclides which showed poor agreement earlier, but I-131, which was not detected in the first count, was identified this time Table 3 lists the results for a second sampling of reactor coolant liquid.
Table 4 lists the data for the corresponding fission gases stripped from the liqui Table
Reactor Coolant Li uid (Sampled at 9:00 a.m.,
February 12, 1985)
Nuclide Na-24 Mn-54 Co-58 I-131 I-133 W-187 Diablo uCi/ml 2.37 E-3 4.46 E-5 3.24 E-5 3.98 E-5 3.24 E-4 7.13 E-4 NRC uCi/ml 2.25 E-3 4.24 E-5 6.43 E-5 3.58 E-5 3.09 E-4 6.93 E-4 Ratio Diablo/NRC 1.05 1.05 0.50 1.11 1.05 1.03 Agreement
~Ran e
0.80 - 1.25 0.60 - 1.66 0.75 1.33 o.6
- 1.66 0.80 - 1.25 0.75 - 1.33 Table
Reactor Coolant Stri ed Gas Nuclide Diablo uCi/ml NRC uCi/ml Ratio Agreement Diablo/NRC
~Ran e
Xe-133 Xe-135 2.10 E-4 4.55 E-4 2.17 E-4 5.31 E-4 0.97 0.86 0.75 - 1.33 0.75 - 1.33 Except for Co-58, for which poor counting statistics was obtained, the results agree reasonable well for both sample categories.
The good agreement achieved for the gas splits is significant because of the potential for large erro'rs in gas stripping and sampling.
Table
Nuclide Cr-51 Mn-54 Fe-59 Co-58 Co-60 Diablo uCi/ml 1 '6 E-6 3.51 E-6 4.06 E-7 4.63 E-5 2.46 E-6 NRC uCi/ml 1.37 E-6 3.41 E-6 6.66 E-7 4.51 E-5 2.35 E-6 Ratio Diablo/NRC 1.21 1.03 0.61 1.03 1.05 Agreement
~Ran e
0.4
- 2.5 0 '5 -
1 '3 0.6o - 1.66 0.80 1.25 0.75 - 1.33 A liquid waste sample was obtained to test for large sample geometry.
Because concentrations of radionuclides in liquid waste normally are quite low and often are at or below the sensitivity limit of the measurement system, the common practice is to use large samples (usually 1 liter) to enhance detection sensitivity.
The results in Table
indicate adequate agreement for this category.
Because neither charcoal cartridge nor particulate filter samples with sufficient activity to measure could be obtained, the licensee was asked to analyze NRC's charcoal =cartridge and particulate filter calibration standards.
The results are summarized in Tables 6 and Table
NRC Charcoal Cartrid e Standard Nuclide Co-57 Co-60
~
Y-88 Cd-109 Sn-113 Cs-137 Am-241 Diablo uCi/ml 7.50 E-4 1.44 E-2 7.23 E-4 4.24 E-2 3.82 E-3 1.53 E-2 6'2 E-3 NRC uCi/ml 9.21 E-4 1.60 E-2 7.67 E-4 4.14 E-2 4.38 E-4 1.55 E-2 6.83 E-3 Ratio Diablo/NRC 0.81 0.90 0'4 1.02 0.87 0.99 0.91 Agreement
~Ran e
0.75 - 1.33 0'0 - 1.25 0.60 - 1.66 0.80 - 1.25 0.60 - 1.66 0.60,-
1.66 0.80 - 1.25 Table
NRC Filter Pa er Standard Nuclide Co-57 Co-60 Y-88 Cd-109 Sn-113 Ce-139 Am-241 Diablo uCi/ml 9.63 E-4 1.57 E-2 7.39 E-4 3.89 E-2 5.26 E-4 2'0 E-4 6.65 E-3 NRC uCi/ml 8.95 E-4 1.54 E-2 6.58 E-4 3.86 E-2 4.36 E-4 2.56 E-4 6.56 E-3 Ratio Diablo/NRC 1.08 1.02 1.12 1.01 1.21 1.09 1.01 Agreement
~Ran e
0.80 - 1.25 0.80 - 1.25 0.60 - 1.66 0.80 - 1.25 0'0 - 1.66 0.75 - 1.33 0.80 - 1.25 While the results indicate adequate agreement for both categories, agreement for the charcoal cartridge was achieved only after the results had been recalculated using the average of the counts obtained for each face of the cartridge.
This technique tends to smooth out the differences in efficiency, which is a function of deposition depth of the activity in the cartridge, and appears to give better results even when calibrations have been performed with standards that are not identical in geometry with that of the sample being measured.
The licensee has indicated that "front face" counting will be used primarily for screening purposes and that samples with measurable activity will be analyzed by counting both "faces" of the cartridge.
While it can be concluded that adequate agreement was achieved for all categories tested, several areas in the gamma spectroscopy procedure'will require special attention.
Because of the multichannel analyzer memory limitations, Diablo uses a two thousand channel (2K) format instead of 4K for gamma spectroscopy.
As the activity spectrum becomes more complex (as it will with more operating experience)
the chance for the occurrence of anomalies in the peak strip routine increases because more information is be'ing stored in a relatively small memory space.
The licensee is aware of the potential problem this situation could create and action has already been initiated to review several possible options and based on this information to upgrade the existing system to permit routine operation at the 4K levels
Another area that could require some consideration concerns measurement sensitivity.
Although specific lower limit of detection (LLD) limits are not prescribed when measuring reactor coolant activity (as it is for effluent activity), it is implicit that certain sensitivity criteria must be met.
Diablo has a license condition that requires activities of greater than 10 minutes half life in reactor coolant be quantified.
A corollary condition requires that these analyses be performed within two hours after sampling.
If the analysis is performed towards the end of the two hour limit, considerable decay takes place for several of the more significant nuclides.
For example, Xe-138 with half life (t~g) of 14 minutes, Rb-88 (tq = 17.8 minutes)
and Rb-89 (t~q = 15 minutes) will be left with only a few percent of its original activity after about 90 minutes from the time the sample is taken.
This requires that adequate sensitivity considerations be factored into the measurement process should a situation such as this arise.
The licensee is aware of this potential problem and will use appropriate procedures such as larger sample, longer counts and/or early counting to meet this requirement.
I 3.
Exit Interview ll Inspection findings were discussed with licensee personnel indicated in paragraph l.
Inspector concerns relative to measurement sensitivity and potential peak anomalies were included in this discussio <<$
Enclosure Criteria for Acce tin the Licensee's Measurements Resolution Ratio
<4
8
-
16
-
51 200 200 0.4 0.5 0.6 0.75 0.80 0.85 2.5 2.0 1.66 1.33 1. 25 1.18 Com arison 1.
Divide each NRC result by its associated uncertainty to obtain the resolution.
(Note:
For purposes of this procedure, the uncertainty is defined as the relative standard deviation, one sigma, of the NRC result as calculated from counting statistics.)
2.
Divide each licensee result by the corresponding NRC result to obtain the ratio (licensee result/NRC).
3.
The licensee's measurement is in agreement if the value of the ratio falls within the limits shown in the preceding table for the corresponding resolution.