ML20101R141
| ML20101R141 | |
| Person / Time | |
|---|---|
| Site: | Fort Saint Vrain  |
| Issue date: | 04/08/1996 |
| From: | Borst F PUBLIC SERVICE CO. OF COLORADO |
| To: | NRC OFFICE OF INFORMATION RESOURCES MANAGEMENT (IRM) |
| References | |
| P-96025, NUDOCS 9604160070 | |
| Download: ML20101R141 (37) | |
Text
_ _ _ _ _ _ _ _
O eubiic service'
- - =.
16805 WCR 191/2; Platteville, Colorado 80651 April 8,1996 Fort St. Vrain l
P-96025 U. S. Nuclear Regulatory Commission ATTN: Document Control Desk Washington, D. C. 20555 Docket No. 50-267
SUBJECT:
Response to Unresolved Item and Inspection Report Concerns (NRC Inspection Report No. 50-267/96-01)
REFERENCE:
NRC Ietter, Scarano to Crawford, dated March 5,1996 (G-96019)
Gentlemen:
This letter provides responses from Public Service Company of Colorado (PSCo) and our decommissioning contractor, the Westinghouse Team (WT), to an unresolved item and four additional NRC concerns identified in the referenced letter. The NRC conducted an on-site inspection of the Fort St. Vrain (FSV) final survey program from January 22-26,1996, and requested that PSCo provide a written response to the following items:
Unresolved Item (267/9601-01): Explahi the method for determining shielded background measurements to ensure that local area background values are not being overestimated (and net activity measurements therefore under-reported).
Concern 1: Determine whether there is a bias in instrumentation response which overestimates the amount of contamination present.
Concern 2:
Determine whether coreactions for "hard io detect nuclides" (HTDNs) should be made to survey results in unaffected areas that exceed 25 percent of the surface contamination limits.
Concern 3: Determine whether investigations of suspect measurements were adequately conducted to justify removing the original measurement from the survey data base.
l Concern 4: Determine whether scan survey coverage percentage in nonsu",pect affected areas should be increased.
150050 9604160070 960408 gi PDR ADOCK 05000267 l
G PDR l
l l
l l
P-96025 April 8,1996 Page 2 PSCo/WT's responses to these items are provided in the attachment to this letter. It is noted that several of these issues are based on differences between survey data taken by the WT/ Scientific Ecology Group (SEG) and data taken by the NRC's consultant, the Oak Ridge Institute for Science and Education (ORISE). The NRC has provided PSCo with a copy of ORISE's report on these measurements, however this report does not include the raw measurement data or calibration information. The attached responses are based on available information and we believe they adequately support FSV Final Survey Procedures currently in use. If, however, a more detailed response is required, this additional information will be needed.
I If you have any questions regarding this information, please contact Mr. Michael H.
Holmes at (303) 620-1701.
Sincerely, hhk(J. Borst 9fk Frederic Decommissioning Program Director J
FJB/SWC cc:
Regional Administrator, NRC Region IV Mr. Robert M. Quillin, Director Radiation Control Division Colorado Department of Public Health and Environment i
l l
i l
Attachment to P-96025 RESPONSE 'ID QUESTIONS RESULTING FROM NRC INSPECTION 96-01 NRC Unresolved Item (267/9601-01):
Explain the methodfor detennining shielded background measurements to ensure that local area background nalues are not being overestimated (and net activity measurements therefore under-reponed). Demonstrate thefollowing:
1 hat background measurements collected by diferent licensee technicians using instruments with diferent detector shields were consistent.
That background wilues applied do not increase as the radioactivity few!
increases, resulting in an underestimate ofnet surwy results.
PSCo/WT Reenanse:
I l
PSCo/WT's response to this unresolved item is provided in two parts:
Part 1 provides a comprehensive evaluation of background shield materials and a comparison of measurements collected by different technicians.
Part 2 discusses the reasons why local area background measurements may in fact i
increase as surface activity increases, due to gamma interaction with the detector, but will not result in an underestimate of net survey results.
These responses are provided in the following pages:
s I
f
i Attachment to P-96025 Page 2 PSCo/WT Response to Unresolved Item 267/9601-01 Part 1
\\
\\
l Evaluation of Background Shield Materials and Comparison Measurements l
l l
l i
l i
i i
.I t
1 4
. = _ -
1 Evaluation of Background Shield Materials and Comparison Measurements Introduction This information is provided to present the experimental testing and analysis performed to evaluate recent comparison survey data and the questions raised during the comparison study l
concerning background measurement protocols. This survey was performed by two independent i
SEG teams using Ludlum Measurements Inc. (LMI) 43-68 gas flow detectors and plexiglass beta shields.
As a result of the comparison survey, a question was raised concerning the appropriateness of plexiglass as a shield material for total surface activity measurements. It is therefore the intent of this review to address this concern as well as to provide an evaluation of the comparison survey data.
l Three specific questions are intended to be answered by this review. These are:
1)
Do photon interactions with a plexiglass shield create a significant background measurement problem?,
2)
Is the thickness of plexiglass used as a beta shield sufficient to adequately shield for FSV beta emitters?, and 3)
Did the comparison survey data agree within statistical allowances, and if not, why not?
These questions were answered by both analytical calculations and experimental testing conducted to evaluate the background measurement protocols used at FSV.
Summary of Results 1.)
Analytical calculations performed to verify the choice of plexiglass as a shield material indicate photon interactions with the shield are not significant.
2.)
The 1/8" thickness of plexiglass used is sufficient to provide adequate beta shielding to correct gross measurements.
3.)
After evaluating the initial and follow-up comparison survey data in consideration of all statistical factors, the results are within agreement (at 95% confidence). There was not a significant difference between plexiglass and steel as a background shield material, which further supports the choice of plexiglass as a shield material' for background measurements.
l l
1 of 19
- 1. Effect of Photon Interactions with a Plexiglass Beta Shield on Background Count Rate The following analysis is performed to evaluate the effects of using a plexiglass shield over the detector window when performing background measurements with the Ludlum Measurements Inc. (LMI) 43-68 detector (rectangular gas flow detector). In question, is whether the shield material causes a significant increase in background from secondary particles that result from photon interactions within the shield.
I Assumntions:
To perform this analysis, the following conditions are assumed:
1)
A 1/8" thick shield of plexiglass is used as the shield material i
1 2)
The shielded detector is placed over a 125 cm area contaminated to a level of 4,000 2
2 dpm/100cm w th Co-60, which results in 5,000 dpm Co-60 under the shield 3)
Each Co-60 disintegration emits two 1.25 MeV photons (i.e., average of 1.173 and 1.332 MeV) 4)
50% of the Co-60 emissions potentially interact with the shield (i.e.,2n geometry) 5)
All photons are considered to be incident upon the shield at average angle of 0 6)
Photon interactions are conservatively assumed to occur uniformly throughout the shield (interaction density will actually decrease through shield resulting in a smaller number of actual compton elec'.rons generated within range of detector / shield surface) 7)
Shield interactions are dominated by compton scattering collisions (i.e., c/p p/p) t 8)
All compton electrons may potentially enter the detector (i.e., moving toward the detector)
Calculations:
Photons incident upon shield:
l l
(5,000 dpm) * (0.5) * (2y/ dis) = 5,000 y/ min Attenuation is determined by use of the following equation:
- % (po N = N, e i
)
2 of 19
1 For a plexiglass shield (with thickness 1/8" or 0.3175 cm) p is 1.2 g/cm) and /p is equal to 2
0.%22 g/cm [1]. This results in the following:
5,000 y/ min
- e - gm22). (1.2). (m5)l = 4883 y/ min or 117 photons / min interact with the shield, which are assumed to produce 117 compton electrons per minute.
The angular differential cross-section for compton scattered photons is given by the following equation [2]:
i 1
32 ' l + cos 0 do _ y'2 2
a (1 - cos 0)2 1
2 1
(1 + cos 0) [1 + a(1 - cos 0)]j dQ 1 + a (1 - cos 0);
2 2
1 s
- s 2
where a is hv/m c, which is 1.25/0.511 or 2.446, and 0 is the scattered photon angle.
o Using the preceding equation, the relative probability at incremented scatter angles was determined (letting Zr 2 = 1). Table 1 lists the calculated probabilities for photon scatter angles o
in increments of 5' (relative probability in column 1, photon scatter angle in column 2). To determine the compton electron scatter angle at each photon angle, the following equation was used [1]:
'0' cot & = (1 + a) tan z es where $ is the average compton electron scatter angle.
The average scatter photon energy, is given by the equation below [2]:
hv hv' =
1 + a (1 - cos 0)
The average compton electron energy is then equal to (1.25 - hv') in MeV.
Using the preceding equations, the compton electron scatter angle, and scatter photon and electron energies are determined for each selected photon scatter angle. Table 1 shows the values calculated for each selected photon scatter angle.
The average photon and electron angles and energies are calculated by a weighted mean of the individual values (i.e., Ipi i pi), where pi s the relative probability of the ith value, V. The V /I i
i results are an average photon scatter angle of 45.0, electron scatter angle of 48.9 and electron energy of 0.387 MeV.
3 of 19
The range of 0.387 MeV electrons in plexiglass is conservatively estimated to be 0.15 g/cm (3),
2 which with an average scatter angle of 48.9*, indicates compton electrons within 0.082 cm of the detector could reach the detector (i.e., cos (48.9 )*0.15/1.2 = 0.082). This results in a correction factor of(0.082)/(0.3175) = 0.26 to account for compton electrons generated in the shield within range of the detector surface. Compton electrons potentially entering the snield then equates to (0.26)*(117) = 30 electrons / min. With a nominal detector efficiency of 20%,
a 40% efficiency is assumed (to correct for the fact that all compton electrons are considered to be moving toward the detector). This results in a 12 cpm background count rate increase, which is below the counting error for a 1 minute background count yielding a result of 400 cpm (as typically observed with the LMI 43-68 detector).
Conclusion-The preceding analysis, although average based, is considered reasonable and sufficiently conservative to provide an estimate of expected photon / shield induced background from licensed i
activity beneath the shield. The result indicates that centamination from licensed material 1
beneath a plexiglass (or closely similar material) would not create a significant increase in background from photon / shield interactions (i.e., only about 3 %). Calculations were specifically i
performed for a 1/8" shield; however, it should be evident that a larger shield thickness would not increase the contribution because once the secondary compton electron range is exceeded additional shield does not result in additional secondary particle detections.
This conclusion is further supported by the experimental data provided later in this review.
4 of 19
i Ttble 1 - Compton Sc:ttering Data For 1.25 M3V Gamms Scatter Photor.
Scatter Photon Scatter Electron Recoti Photon Gompton tiectron Productof Probability Product of Probability Product of Probability Product of Probability l
Angle Probability Angle (degrees)
Angle (degrees)
Energy (MeV)
Energy (MeV) and Photon Angle and Electron Angle and Photon Energy and Electron Energy (relative) 1 000 00 90 0 1260 0 000 0 000 90 000 1260 0000 0 978 5.0 81,4 1.238 0 012 4 890 79.648 1.211 0 011 0 916 10 0 73 2 1.205 0 045 9.162 67.089 1.104 0 041 0.826 15.0 65 6 1.154 0096 12.394 54.200 0 953 0.079 0.722 20 0 58.7 1.089 0.161 14.444 42 405 0787 0.116 0 617 25.0 52.6 1.017 0.233 15 423 32.463 0 627 0.144 0 519 30 0 47.3 0.941 0.309 15.580 24.554 0.489 0.160 0.434 35.0 42.6 0 867 0.383 15.198 18.509 0.376 0.166 0.363 40.0 38.6 0.795 0.455 14.524 14 003 0.289 0.165 0.305 45 0 35.0 0.728 0.522 13.740 10 691 0.222 0.159 0.259 50.0 31.9 0667 0.583 12.964 8.270 0.173 0 151 0.223 55 0 29.1 0.612 0.638 12.265 6.498 0136 0.142 0.195 60.0 26.7 0.562 0.688 11.674 5.192 0 109 0.134 0.172 65.0 24.5 0.518 0.732 11.201 4.220 0 089 0126 0.155 70.0 22.5 0479 0.771 10.845 3.488 0 074 0.119 0.141 75.0 20 7 0.444 0 806 10 595 2.927 0.063 0.114 0.131 80.0 19.1 0.414 0 836 10.441 2.490 0 054 0.109 0.122 85 0 17.6 0.387 0.863 10 370 2.144 0 047 0.105 0.115 90 0 16.2 0.363 0.887 10 370 1 865 0.042 0.102 0 110 95 0 14 9 0.342 0.908 10 431 1.635 0 038 0.100 0.105 100 0 13.7 0.323 0.927 10.544 1.443 0 034 0 098 0.102 105 0 12.6 0.306 0.944 10.702 1.279 0 031 0 096 0.099 110.0 11.5 0.292 0 958 10.896 1.138 0029 0 095 0.097 115.0 10.5 0.279 0.971 11.123 1.013 0.027 0 094 0095 120.0 9.5 0.268 0.982 11.377 0.902 0 025 0 093 O.093 125.0 8.6 0.258 0.992 11.655 0801 0 024 0 093 0.092 130 0 7.7 0.249 1.001 11.954 0.709 0 023 0 092 0.091 135.0 69 0.242 1.008 12.270 0 623 0 022 0 092 0 090 140.0 60 0.235 1 015 12.603 0.543 0 021 0.091 0.089 145 0 5.2 0.229 1.021 12.950 0.467 0.020 0 091 0.089 150.0 4.4 0.225 1.025 13.310 0.395 0.020 0 091 0 088 155 0 3.7 0.221 1.029 13.683 0 325 0 019 0.091 0 088 160 0 2.9 0.218 1.032 14.067 0.258 0 019 0 091 0.088 165.0 2.2 0.215 1.035 14.462 0.192 0 019 0 091 0.087 170.0 1.5 0.213 1.037 14 868 0.127 0.019 0 091 0 087 175 0 0.7 0.212 1.038 15.287 0 063 0 019 0.091 0.087 180.0 0.0 0.212 1.038 15.717 0 000 0 019 0 091 Weighted Average 45.0 48.9 0.863 0.387 Values:
5 of 19
- 2. Evaluation of Beta Shielding Effectiveness of 1/8" Plexiglass Evaluation When performing background measurements for plant surfaces and structures, a 1/8" thick slab of plexiglass is used to shield beta radiation. This corresponds to 381 mg/100cm (i.e.,0.125"*
2 3
2 2.54 cm/"*1200 mg/cm = 381 mg/cm ).
To verify the effectiveness of the shield for the detectable nuclides at FSV, the following analysis has been performed.
A 381 mg/100 cm2 shield of plexiglass is conservatively estimated to stop up to 740 kev beta particles l3]. In other words, only particles with energy greater than 740 kev will be able to penetrate the shield. To determine percentage of beta particles from detectable contamination at Fort St. Vrain that would penetrate this amount of shielding, a decay distribution shape is assumed as shown in Figure 1 below superimposed on a typical (curved line) beta decay distribution (i.e., typical for p' decay, which is decay scheme ofinterest at Fort St. Vrain, a p+
decay distribution would involve a shift to higher energies due to Coulomb interaction with the I
nucleus) [2, 4].
f i
?(D rw a w o. m IN w.aw am/f i
N l
l l
N N
N xx i
N\\NN
/'
i NN
/
sG el
<D i
Figure 1 - Typical and Assumed p-Decay Distributions 6 of 19 f
As seen in Figure 1, the distribution is assumed to be composed of two segments. The first segment assumes a uniform energy probability for beta particle energies up to Ebar. The second segment assumes a triangular shape terminating at Emax. Under each segment, the normalized area is defined to be 0.5, which, in effect, assumes that half the beta particles have energy below Ebar and half above Ebar. To determine attenuated area under the second segment of assumed distribution when Ebar is < 740 kev and Emax is > 740 kev, the following equation is used:
)
P(Ebar)
P(E) = P(Ebar) - (Emax - Ebar)(E - Ebar) l Since area under second segment of assumed distribution (i.e., triangular region) is by definition l
equal to 0.5, solving for area of the triangular region results in the following expression for l
P(Ebar):
I P(Ebar) =
(Emax - Ebar) which by substitution results in the following expression for P(E):
l I
P(E) -
(E - Ebar)
(Enar - Ebar)
(Emax - Ebar)2 Using this equation, attenuated area under second segment of assumed distribution can be determined for a given energy (i.e., 740 kev).
(
Assumptions used to determine attenuation for each decay line of a given radionuclide are as follows:
1)
For beta particle distributions whose Ebar is greater than 740 kev, the fraction of beta i
particles that are attenuated by the shield is determined by:
744 '
% attenuated -
- 50 i
Etcr l
where Ebar is in kev.
2)
Beta decay distributions whose Emax is < 740 kev ard internal conversion and auger electron emissions whose mono-energetic energies are < 740 kev are attenuated 100%
(internal conversion and auger electron emissions < 740 kev are grouped together with a summed abundance).
7 of 19
3)
Internal conversion electron emissions with mono-energetic energies > 740 kev are not attenuated.
4)
For beta particle distributions whose Ebar is less than 740 kev, but whose Emax is greater than 740 kev, the fraction of beta particles that are attenuated by the shield is determined by:
I i
(
4 I
(Emax-740)
(#-
- )
% attenuated = 50 + 100* 0.5 -
(Em- &r)2 m-2 1
where Ebar and Emax are in kev.
Total attenuation for a given nuclide is given by weighted averaging of all individual decay lines of the nuclide (i.e., weighted by individual line abundances). Total attenuation for Fort St. Vrain nuclide mix is determined by weighted averaging of individual nuclide total attenuations where weight is based on fraction present in the average nuclide mix.
The following table lists specific radionuclides identified in " detectable" contamination at Fort St. Vrain and % attenuated for each decay line of the nuclide and total effective attenuation for the nuclide. These radionuclides are the values and fractions identified in the sample set used to determine SGLVs. Total attenuation for Fort St. Vrain contamination by 1/8" thick plexiglass is estimated at 99.4%.
Conclusion The 1/8" thick plexiglass shield is adequate to shield beta emitters encountered at FSV.
Attenuation was calculated to be over 99% for the average FSV nuclide composition as well as over 99% for key nuclides (i.e., Cs-137 and Co-60). The effectiveness of plexiglass to shield FSV beta has also been evaluated by experimental tests. The following section presents the results of this testing which included additional measurements by two SEG teams using both the normal plexiglass shield and a 0.2" thick steel shield.
'I 8 of 19
Shielding Effectiveness of 1/8" Plexiglass Shield l
l Nuclide Average Mean #
Ebar (kev)
Emax (kev)
% Beta Overall %
t Fraction of Beta per Line per Line Attenuated Attenuated I
of Nuclide Particles per Line per Nuclide l
per Decay l
Co-60 0.62000 1.000 95.8 (100 %)
317.9 (100 %)
100.0 99.9 l
1460 (O.12 %)
1460 (O.12 %)
0.0 Sr-90 0.00234 2.000 195.8 (Sr-90,100%)
546.0 (Sr-90) 100.0 67.3 934.8 (Y-90,100%)
2283.9 (Y-90) 34.5
{
Cs-134 0.00746 1.015 23.1 (27.4 % )
88.5 (27.4 %)
100.0 99.8 123.4 (2.48 %)
415.1 (2.48 %)
100.0 210.1 (70.1 %)
657.9 (70.1 %)
100.0 758.4 (0.22 %)
758.4 (0.22 %)
0.0
< 740 (1.28%)
< 740 (1.28%)
100.0 Cs-137 0.15500 1.174 156.8 (94.6 % )
511.6 (94.6 %)
100.0 99.3 415.2 (5.4 %)
1173.2 (5.4 % )
83.7
< 740 (17.4%
< 740 (17.4%)
100.0 Eu-152 0.19900 1.424 47.5 (1.78%())
176 (1.78 %)
100.0 98.1 l
112.5 (2.4 %)
385 (2.4 %)
100.0 221.8 (13.6 %)
696 (13.6 %)
100.0 227 (0.23 %)
710 (0.23 %)
100.0 295.3 (0.29 %)
889 (0.29 %)
96.9 364.8 (0.89 %)
1064 (0.89 %)
89.3 j
535.6 (8.44 %)
1475 (8.44 %)
70.1 l
< 740 (l14.4%)
< 740 (114.4%)
100.0 Eu-154 0.01480 1.838 68.8 (27.9 % )
247.4 (27.9 %)
100.0 96.8 86.9 (0.77 %)
306.1 (0.77 %)
100.0 91.7 (0.149 %)
321.2 (0.149 %)
100.0 100.9 (1.58 %)
349.8 (l.58 %)
100.0 119.8 (O. I 17 %)
407.4 (O.117 %)
100.0 129.3 (0.281 %)
435.7 (0.281 %)
100.0 168.3 (0.188 %)
548.6 (0.188 %)
100.0 175.7 (36.5 % )
569.4 (36.5 %)
100.0 l
224.5 (0.64 %)
703.2 (0.64 % )
100.0 229 (0.245 %)
715.4 (0.245 %)
100.0 276 (17.4 % )
839.2 (17.4 %)
98.4 327.5 (2.0 %)
970.7 (2.0 % )
93.6 400.4 (0.29 %)
1151.5 (0.29 %)
85.0 587.4(0.24 %)
1596 (0.24 %)
64.0 695 (l1.4 % )
1843.9 (l1.4 % )
53.8
< 740 (83.6%)
< 740 (83.6%)
100.0 Tc-99 0.00168 1.000 84.6 (100 % )
293.6 (100 %)
100.0 100.0 Overall Attenuation (%) =
99.4 9 of 19
l
- 3. Experimental Testing l
Review of Comparison Survey Data l
Comparison measurements were previously taken (at time of NRC visit) at four locations by two survey teams. A summary of the corrected results is provided in the table below.
l SEG Team 1 SEG Team 2 Difference l
Result (dpm/100cm )
(dpm/100cm )
2 2
2 (dpm/100cm )
l Battery Room 729.8 828.2 98.4 i
l Lube Oil Room
-133.7
-82.5 51.1 Rx Bld - #2500 2199.5 2476.5 277 Rx Bid - #5000 4334.7 4212.7 122 l
To determine if the measured differences are statistically significant, means testing can be l
performed using the following equation:
1 l
E
~5 l
2 g,
2 2
3 8
__1_ + _2
\\
n n
l When this is calculated for each of the survey locations surveyed for the comparison study the l
following values for t and corresponding probability are obtained.
l I
Survey Location Calculated t value Probability
- Battery Room 0.964 37 %
Lube Oil Room 0.676 51 %
Rx Bld - #2500 1.944 7%
Rx Bld - #5000 0.942 38 %
l Probability (as shown) is the probability that the observed difference or greater between data sets would be observed even if taken from l
the same population (i.e.,95% confidence requires that the probability be 2 5%).
l l
As shown in the table, comparison data at each location was in statistical agreement (at 95%
confidence). These values were calculated using the sample standard deviation listed with the mean net result for all but the Rx Bld - #2500 location. For the Rx Bld - #2500 location, an initial calculation yielded a t value of 2.579 and corresponding probability of 1.9%. However, the initial calculation (i.e., using the sample standard deviation of the net results) did not allow for background measurement uncertainty (because the sample standard deviation associated with 10 of 19
~ ~ -
=
1 the net results are each corrected with the same background value).
When background uncertainty was included, the results of the previous table were obtained. The total uncertainty of each comparison result would also include the uncertainty of the efficiency determination, but including this uncertainty was not necessary to show agreement between the comparison data.
Additional Testino Despite the understandable differences with the initial comparison survey data, re-survey of the initial comparison locations was performed by two independent SEG survey teams. Two additional locations of higher surface activity were also selected for evaluation. Each team surveyed each location using both the normal plexiglass shield (381 mg/cm ) and a 0.2" thick 2
slab of steel (about 4,000 mg/cm ). Unshielded and shielded measurements were also taken one 2
meter from the location surface to provide additional information. Contact background readings (i.e.,6 measurements) were taken in the immediate area surrounding the survey location. The results of these surveys are provided in the attached tables.
i Measurements were also taken by placing the shields (including a third thin metal shield) over i
a Tc-99 (pure beta emitter) and a Co-60 gamma source to evaluate the shield materials.
Although desired to also evaluate a Cs-137 source, this nuclide also emits gamma and, as was i
noted by the Co-60 source response, would not provide any meaningful information as the nuclides gamma penetrates the shields.
Conclusion Significant differences in survey data between plexiglass and steel shields was not observed.
Comparisons between Team 1 and Team 2 survey results were also acceptable. As shown on each table, a 95% confidence allowed difference was calculated by allowing for counting
{
uncertainty in background, gross and efficiency measurements. Efficiency uncertainty was 1
included for these measurements because some measurements were also taken at higher surface activity locations and as activity increases the absolute counting uncertainty decreases but the absolute uncertainty due to efficiency increases. With two comparisons at each location (i.e.,
one with steel shielding and one with plexiglass shielding) 12 total comparisons have been made.
Only one comparison (i.e., the Battery Room with steel shields) was slightly outside of 95%
confidence. With 95% confidence, a failure rate of 1 in 20 is expected. Therefore, a 1 in 12 failure rate is not surprising (actually the odds are greater that there would be a 1 in 12 failure than 0 in 12).
]
The evaluation of the shields effectiveness to shield Tc-99 beta indicated that significant beta does not penetrate the shields. Evaluations with the Co-60 source resulted in an expected decrease in count rate with increasing shield thickness.
1 l
l 11 of 19 l
SEG MEASUREMENT COMPARISON t
~
SEG Team 1 IAy N Meam.
Unstuoided Meas. Stueided Meas. (piexiglass) Stueided Meas. (plexiglass)
Shseeded Meas. (steel)
Stueided Meas. (steel)
C: __ - M 1888 Contact (com) 1 meter (cpm)
Contact (qwn) 1 maler(com)
Contact (com) 1 meter (cpm)
Battery Room 436 344 276 236 248 288
- n " ;-
')
424 332 284 284 252 288 420 376 224 280 240 240 440 316 276 256 260 288 368 368 272 264 292 200 404 312 264 296 296 248 412 352 536 392 432 316 416 300 Sed Dev 40 1 30 8 21 6 21 7 23 7 36 0 Mean 430 8 340 8 2660 269 3 264 7 258 7 Efliuency 0225 Not Result with Plexiglass Shield (dpmf100an2) 5860 Not Result with Steel Shield (dpmf100cm2) 590 7
~
SEO Team 2 N Mees.
Unstuelded Meas. Shiekled Meas. (plexigloss) Shielded Meas. (plexiglass)
Simeeded Meas. (steel)
Shielded Meas. (steel)
E___
' 1188 Contad (qun) 1 meter (qun)
Conted (com) 1 meter (com)
Contact (com) 1 meter (cpm)
[
Battery Room 636 408 352 396 336 288 l
(Background) 636 384 388 280 400 348 480 428 300 300 344 292 556 384 384 336 316 292 I
612 392 312 364 340 316 528 348 360 304 300 256 I
604 408 L
548 456 576 420 532 428 l
Std Dev 51.3 30 2 36.5 43.9 34.1 30 8 Mean 570 8 4056 349.3 330.0 339.3 298.7 EfEciency 0.219 Net Resun with Plexiglass Shield (dpmf100cm2) 809 0 Not Renas wi8: Steel Shield (dpmf100cm2) 845 5 c-_--, in : - a f
Beta Shield Plexiglese Steel f
Team 1 586 0 590 7 Team 2 809 0 8455 Ddierence 223.1 254 8 95% De-sta 238 7 249 4 12 of 19
SEG MEASUREMENT COMPARISON i
SEG Team 1 UnsNeided Mees.
UnsNeided Meas. SNeided Meas. (plexigless) Stuelded Meas. (piexiglass)
SNeided Meas. (steel)
Stueided Meas. (steel)
CL
'1099 Contad (cpm) 1 meter (cam)
Contact (com) 1 meter (qun)
Contact (qwn) 1 meter (cpm) l Lube Oil Room 436 304 200 256 212 252 468 340 224 256 208 232 i
492 304 260 292 260 184 472 356 288 272 192 248 428 324 320 304 252 220 444 316 320 312 260 264 400 336 404 320 496 352 I
400 340 l
Sid Dev 36 8 18.5 36 8 24 2 30 1 28.7 Mean 444 0 329 2 282 0 282 0 230 7 233 3 Efikaency 0225 Net Result wilh Plexigloss SNeed (dpmI100cm2)
-10 0 Not Result wilh Steel SNeid (dpmI100cm2) 167 8 I
k SEG Team 2 Unshielded Meas.
UnsNeided Mees. Sb ars.,e Meas. (plexiglass) SNeided Mess. (plexiglass)
SNeided Meas. (steel)
Stueided Meas. (steel)
[
C:
__1108
_CMact (com) 1 meter (com)
Conlad (com) 1 meter (qwn)
Contact (cpm) 1 meter (cpm)
^
Luhe Oil Room 508 436 340 284 312 300 480 380 356 308 328 364 508 376 344 280 296 316 536 380 268 184 292 212 516 408 336 328 304 352 496 444 404 344 280 316 624 436 i
532 408 l
468 452 548 380 Sid Dev 43.7 30.1 43.7 56.8 16.7 53 8 Mean 521.6 4to0 341.3 288 0 302 0 3100 EfEcsoncy 0219 Net Rosadt with Plexiglass SNeid (dpmI100cm2)
-150 5 Not Resut wi8: Steel Shield (@mI100tzn2)
-43 3 r-
.- _ 1 --w:
Bela Shield Plexquises Steel Team 1
-10 0 167 8 Team 2
-150.5
-43.3 OsNerence 140.5 211.1 95% Dette 262.2 275 3 13 Of 19
SEG MEASUREMENT COMPARISON SEG Teesa 1 Unsheelded Meas.
Unshselded Meas.
SNeided Meas. (plexigloss) Sheelded Meas. (plexigloss)
Shelded Meas. (sieel)
Sheelded Meas. (steel)
C-n. - -f1000 Conlad (com) 1 maler (com)
Contact (cper.)
1 meter (@m)
Contact (gm) 1 meter (cpm)
Rm Bid Lv 9 (2800 1032 476 420 352 344 348 880 484 304 328 308 324 1076 360 324 264 288 360 912 372 308 412 296 296 1068 376 380 372 344 364 980 432 312 368 328 308 968 412 1004 336 888 408 972 460 Std Dev 69 5 51 0 47 6 50.1 24 2 28 2 Mean 9780 411.6 341.3 349 3 3180 333 3 EfEcsoncy 0225 Net Resus with Plexiglass Shield (dpm/100an2) 1877.7 Net Result wNh Sleet Shield (dpmI100cm2) 17560 i
SEG Team 2 Unshielded Meas.
Unshselded Meas. SW Mees. (plexigloss) SW Meas. (pioxigloss)
Shaoided Meas. (sleel)
Shielded Meas. (sieel) f C;
f1100 Contact (com) 1 meter (com)
Contact (com) 1 meter (cpm)
Contact (@m) 1 meter (mm)
Rm Bid Lv 9 (2500 1096 392 504 424 380 424 920 500 428 364 364 316 984 460 440 384 392 380 l
1032 520 444 352 452 416 l
1188 472 404 404 372 372 1052 480 392 432 344 344 1020 496 I
1076 464 f
1996 384
[
1092 484 I
Sid Dev 72.9 44 4 39.3 32 3 37.0 41.4 Mean 1055 6 465 2 435 3 393.3 384 0 375 3 Eflicaency 0219 Net Resus with Plexiglass Shield (dpmI100cm2) 1456 8 Not Reste th Sleet Shield (@mI100cm2) 1607 8 Bels Shield Plexigloss Steel Team 1 1677.7 17560 Team 2 1456 8 1607 8 Defterence 220.9 148.1 95% DeNa 360 3 336 2 14 of 19
SEG MEASUREMENT COMPARISON SEG Team 1 p Meas.
UnsNaMad Meas.
Stuelded Meas. (piexigloss) ENeeded Meas. (piexigloss)
SNeeded Meas. (steel)
Stueided Meas. (sieen C;--
' 1999 Contad (qwn) 1 meter (cpm)
Coniad (cpm) 1 meter (cpm)
Coniad (cpm) 1 meter (cpm)
Rx Bad Lv 9 (-Sk) 1600 368 332 380 356 316 i
1520 384 340 420 332 256 l
1600 356 412 300 292 300 l
1432 352 408 392 340 308 j
1848 384 388 356 340 316 1500 364 432 376 308 340 1620 364 1472 480 1660 436 1768 368 Std Dev 101.3 40 8 40 8 40 5 23 6 27 9 Mean 1582.0 385 6 385 3 370.7 3280 3060 Ef6caency 0225 Net Result with Plexigloss SNeid (cipmf100an2) 3888 8 Not Res A with Sleet Shield (dpmI100cm2) 38680 SEG Team 2 UnaNaMad Meas.
Unab= Mare Mees. SNeided Mees. (plomigloss) SNeided Meas. (pissigloss)
Sheoided Meas. (steel)
Steelded Meas. (steel)
C-
' 1988 Contact (com) 1 meter (com)
Conlad (cpm) 1 meter (qwn)
Contad (cpm) 1 meter (qwn) p Rx Bid Lv 9 (-Sk) 1624 452 384 416 464 336 1688 420 432 364 368 360 1648 368 444 408 380 368 t
1784 160 480 380 500 356 1676 504 432 352 384 344 1788 524 464 352 380 372 1540 464 1812 352 1736 528 l
1628 484
{
Sid Dev 87.0 60 4 33 0 27.9 55 2 13 9 Mean 1692 4 455 6 439 3 378.7 412.7 3560 EfEcnoncy 0219 Net Result wim Plexigions Shield (dpmf100cm2) 3788.4 Not Reese wim Sleet Shield (@m/100an2) 3829 3 r-Beta Shield Piemagines Sleet 1
Team 1 3888 8 38680 1
Team 2 3788 4 3829 3 DiNorence 99.6 38 6 i
95% Deita 449.7 4240 15 of 19
SEG MEASUREMENT COMPARISON l
SEG Team 1 UnsNeided Meas.
UnsheMed Meas. SW Meas. (plexiglass) SNeided Meas. (plexiglass)
Sheided Meas. (steel)
Sheided Meas. (steel)
En v 1999 Contact (cpm) 1 meter (qwn)
Contact (cpm) 1 meter (com)
Contact (qwn) 1 meter (cpm)
Ra Bad Lv 9 (-10k 3132 524 404 304 316 296 2940 392 312 332 368 344 3036 396 312 364 248 384 2916 408 408 332 360 280 3064 400 380 312 308 312 3056 500 348 280 324 328 3168 384 3124 460 2988 420 3136 396 Std Dev 86.6 49 4 43 3 28 8 43 0 37.1 I
Mean 30560 4280 360 7 320 7 320 7 3240 Efficency 0225 Net Result wilh Plexiglass Shield (dpm/100cm2) 8997.4 Net Result with Steel SNeH (dpm/100cm2) 9134 9 P
SEG Team 2 Unsholded Meas. Unsheided Meas. SW Meas. (plexiglass) SNeided Meas. (plexiglass)
SNeeded Meas. (steel)
Sheided Meas. (steel)
Download 1100 Contact (com) 1 meler(qwn)
Contact (cpm) 1 mescr (cpm)
Contact (com) 1 meter (com)
Rm Bid Lv 9 (-tek 3316 460 364 328 468 320 t
3052 480 416 476 380 300 I
3592 384 424 328 396 348 3232 404 436 372 424 272 3364 552 472 428 376 376 I
f 3100 508 372 368 464 356 3200 444 3168 528 3208 476 3336 476 I
Std Dev 154.3 51.9 40 5 58 4 40 9 38 7 Mean 3256 8 471 2 414 0 383.3 418 0 328.7 EIEciency 0219 Net ResuN with Plexigloss Shield (dpmI100cm2) 9575.7 Not Resut wWe Steel Shield (@mI100cm2) 9524 5 Compensons Beta Shield Pen - amma Steel Team 1 8997.4 9134 9 Team 2 9575.7 9524 5 Dillerence 578.3 389 6 95% Dema 737.5 7326 16 of 19
SEG MEASUREMENT COMPARISON SEG Team 1 Unstuelded Meas.
UnsNeided Meas. Sheelded Meas. (plexigless) Stueided Meas. (piemiglass)
Shneided Meas. (sieel)
Shielded Meas. (steel) 0; --
-f1999 Contad (com) 1 meter (cpm)
Contact (cpm) 1 meter (com)
Contact (com) 1 meter (com)
Rm Bid Lv 5 (-18h 5028 404 412 320 280 376 5132 496 336 380 364 328 5252 400 320 348 320 272 5168 420 312 332 380 356 5156 428 344 344 428 332 5064 420 408 408 408 304 5116 416 4952 392 5108 408 5160 484 Sid Dev 83.2 35.1 43 8 32.7 55 3 36 9 Mean 51136 426 8 355 3 355 3 363 3 3280 Eh.cf 0225 Nel Result with Ple-e.s Sheeld (dpm/100cm2) 16332 3 Net ResuN with Steel Shiekt (dpm/100cm2) 16299.1 SEG Team 2 IW Meas.
Unshielded Mees.
SNeided Meas. (plexigions) Shielded Meas. (plexiglass)
Sheelded Meas. (sieel)
Sheelded Meas. (steel)
C_-.
" 1100 Conted (com) 1 meter (cpm)
Contact (com) 1 meter (com)
Contact (com) 1 meter (com)
Rm Bad Lv 9 (-15k 4792 504 352 404 464 348 5072 520 420 444 424 332 5104 396 368 348 424 332 4812 440 392 304 448 336 4896 512 392 388 388 392 5056 400 428 412 484 516 5228 456 5120 488 5104 480 5196 412 Std Dev 152.7 46 9 29 2 50 0 34 0 72.3 Mean 50380 460 8 392 0 383.3 438 7 3760 j
Eh4 0219 Net Result with Ple-W Shield (dpm/100cm2) 18162.7 Not Resut wipi Steel Shield L.."100cm2) 15955 7 C.
- a:
L' Shield Plongstess Steel Team 1 16332.3 16299.1 Team 2 16162.7 15955 7 Deflerence 169 6 343 4 95% Deus 1049 4 1069 7 1
17 of 19
Source and Background Aleasurements from 2.4 nCi Tc-99 Source Shield Alaterial Background (cpm)
Shielded Source (cpm)
Plexiglass (1/8")
384 20 388 11 Steel (0.2")
365 19 363 t 11 Sheet Metal (0.08")
348 19 377 11 Source and Background hieasurements from 1 Cl Co-60 Source Shield Alaterial Background (cpm)
Shielded Source (cpm)
Plexiglass (1/8")
283 17 9,291 i 56 Steel (0.2")
293 17 6,168 45 Sheet Metal (0.08")
322 18 10,104 i 58 18 of 19 1
l Summary of Results l
1.)
Analytical calculations performed to verify the choice of plexiglass as a shield material l
indicate photon interactions with the shield are not significant.
l 2.)
The 1/8" thickness of plexiglass used is sufficient to provide adequate beta shielding to i
correct gross measurements.
3.)
After evaluating the initial and follow-up comparison survey data in consideration of all l
statistical factors, the results are within agreement (at 95% confidence). There was not a significant difference between plexiglass and steel as a background shield material, which further supports the choice of plexiglass as a shield material for background measurements.
References 1.
Attix, F.H., " Introduction to Radiological Physics and Radiation Dosimetry", John Wiley
& Sons, New York,1986.
2.
Knoll, G.F., " Radiation Detection and Measurement", John Wiley & Sons, New York, 1989.
l l
3.
Lederer et.al., " Table of Isotopes" 7th edition, John Wiley & Sons, New York,1978.
4.
Krane K.S., " Introductory Nuclear Physics", John Wiley & Sons, New York,1988.
)
l l
i l
i j
1
(
19 of 19
1,
]
Attachment to P-96025 l
Page 22
}
i i
PSCo/WT Response to Unresolved Item 267/9601-01 i.
Part 2 4
Evaluation of Local Area Background Variations As Surface Activity Incmases a
e 1
i J
i 1
i
Attachment to P-96025 Page 23 Evaluation of Local Area Background Variations As Surface Activity Increases Where gamma emitting radionuclides are present, the local area background may increase as surface activity increases :lue to gamma interaction with the detector. However, this will not result in an underestimate of net survey results because surface activity is quantified in terms of the detector response to beta radiation.
On average, one disintegration of FSV surface activity will emit one beta particle (slightly more than one
" beta" when conversion electrons are included). Tc-99, a pure beta emitter with an average energy of 85 kev and having an intensity of 1, is used as the calibration source.
Therefore, the number of disintegrations per unit time is correctly deteririned by quantifying the number of beta particles emitted per unit time from the surface.
When considering gamma radiation as a component of local area background, three possibilities related to interaction exist: (1) gamma radiation passes through the shield and the detector without interaction, (2) gamma radiation passes through the shield without interaction, but does interact with the detector, and (3) gamma radiation interacts with the shield.
1.
For instances where gamma radiation passes through the shield and the detector without interaction there is no effect on the shielded or the unshielded measurement due to photon interaction. This would be the ideal condition when quantifying surface activity based on the response of the detector to the beta radiation. For this condition, there would be no effect on the net measurement result.
2.
For local area background measurements collected with the detector shield in place, gamma radiation passing through the shield without interaction, but interacting with the detector, would cause a slight increase in the detector response to the local area background (gamma efficiency is approximately 1 to 2 percent). This condition would be identical to subsequent measurements collected without the detector shield in position. For this condition, there would be no effect on the net measurement result.
3.
For local area background measurements collected with the detector shield in place, gamma radiation interacting with the shield such that secondary radiation interacts with the detector, but the gamma radiation does not interact with the detector, a slight increase in the detector response to the local area background
l l
l Attachment to P-96025 Page 24 l
l could occur. For this condition, there would be no significant effect on the net measurement result.'
Evaluations were also performed to determine if using different detector shields would cause significantly different responses. The results of measurements collected by two survey teams (i.e., Team 1 and Team 2) were used for this testing. This evaluation included the effect of photon interactions with plexiglass (the material used for detector shields). Results of these evaluations, and a comparison of the Team 1 and Team 2 measurement results are included in Part 1 of this response discussion.
In summary, the evaluation indicates that local area background measurements collected using steel shields were generally slightly lower than local area background measurements collected using plexiglass shields.. This is as expected since a greater fraction of gamma radiation is shielded by the steel. The reduction in the local area i
background measurement caused by the use of a steel shield would amplify any high bias caused by using the plexiglass shield due to the larger differential in the gamma flux for subsequent measurements collected without the shield in position.
Overall, there was not a statistically significant difference in net results collected using the plexiglass or steel shields due to the low detection efficiency for gamma radiation and the minimal amount of secondary radiation detected.
Therefore, background measurement protocols and the use of plexiglass as the background shield are appropriate.
t l
l l
8 De fraction of the photons interacting with the shield, and the effect of any secondary radiation produced i
as a result of this photon interaction has been evaluated and deternuned to be approximately 3 percent for contanunation levels at the SGLV. This is less than our typical counting error ofi 10 percent (1 standard j
deviation). Refer to Part 1 of this response discussion for the details of this evaluation.
i j
i l
Attachment to P-96025 j
Page 25 1
NRC Concern 1:
}!.
Detennine whether there is a bias in instrumentation response which overestimates the amount of contamination present.
i PSCo/WT Resoonse:
The basis for this concern is a side-by-side measurement comparison performed by PSCo i
and ORISE at FSV. At the time of the comparison, several concerns were expressed by PSCo related to the method of efficiency determination, the method used to assign i
background values, and the calibration parameters used by ORISE. At the conclusion l
of the comparison measurements, information related to the above, and the individual gross measurement results were requested by PSCo in order to adequately evaluate the relative responses of the instrumentation used for the side-by-side measurement l
comparison. In the absence of this information, only a limited evaluation of the relative response can be performed. This evaluation focused on the effects of differing source-to-i detector geometry used during efficiency determination.
I The method of efficiency determination used by PSCo includes positioning the source at i
l a distance of approximately 1/8" from the detector; and evaluating the response across i
3 regions of the sensitive area. (e.g., the heel, the center and the toe of the detector)
]
Positioning the source at a distance of approximately 1/8" from the detector is intended j
to more closely match the counting geometry encountered during field measurements.
j (e.g., face plate thickness prevents survey surfaces from contacting the screen, irregular surfaces, etc.). Evaluating the response across 3 regions of the sensitive area is intended to define the efficiency of the LMI 43-68 detector over the entire detector area, and to i
ensure uniform response within the established bounds.
)
It is understood that the method of efficiency determination used by ORISE includes i
positioning the source on contact with the detector at approximately the midpoint of the sensitive area. This method should result in a greater solid angle of detection than the average solid angle over the entire detector area, and would not evaluate the response of the detector across the entire sensitive area. PSCo evaluated the effect on the detection efficiency by using each of the above methods.
Results of testing indicate that the detection efficiency can be increased by as much as 10% (relative) by placing the source on contact with the detector at approximately the midpoint of the sensitive area. (e.g., a 20% efficiency becomes 22%). This would cause ORISE measurement results to be biased low by as much as 10% relative to the PSCo measurement results.
1 Attachment to P-96025 Page 26 To determine if the differences in the methods used for efficiency determination explain the discrepancy between PSCo and ORISE measurement results, the results of the side-by-side measurements were reviewed. For the two locations where licensed surface activity was present, the ORISE measurement results were about 18% lower than the PSCo measurement results. Therefore, although measured efficiency differences would explain about half of this difference, there is apparently additional bias present.
l Although limited comparison data was taken, review of the side by side measurement results revealed that ORISE detectors consistently yielded a lower response than PSCo detectors. The lower response was observed for all measurements (i.e., average gross, i
background and net results). The lower results reported by ORISE is indicative oflower detection efficiency than PSCo detection efficiency, although ORISE detectors are assumed to be more efficient. In order to determine the cause of lower response by ORISE detectors, information concerning ORISE instrument parameter settings (e.g.,
plateaus / operating voltages, cable lengths, threshold settings, etc.) would be needed.
Although one might expect specific parameter differences between detectors to be accounted for in the calibration, this cannot be confirmed unless specific differences are j
known and tested. Additionally, differences will not be accounted for when measuring licensed material if different calibration techniques are used (e.g., threshold setting, source position (s), source construction, back scattering correction, etc.).
By reporting higher measurement results than ORISE, PSCo measurement results are conservative (relative to ORISE) assuming that ORISE measurement results are not biased low significantly. This condition will not result in inappropriate classification and/or inappropriate conclusions regarding suitability for release for unrestricted use.
If further explanation is required, or to allow more efficient evaluation of future comparison measurements, specific details concerning comparison instrument parameters and calibration techniques are required.
PSCo side-by-side measurement comparisons with GPUN were also evaluated. Better agreement was obtained between PSCo and GPUN than was observed between PSCo and ORISE. Raw measurement results were very consistent between PSCo and GPUN.
When corrected for efficiency, GPU results were about 6 to 7 percent lower than PSCo results. This minor discrepancy is within PSCo efficiency determination tolerances (i 10 percent) and shows PSCo survey methods to be more conservative. For information, the GPUN Comparison Data is attached to this response.
Based on the above evaluation, PSCo does not consider that there is a bias in PSCo instrumentation response which overestimates the amount of contamination present.
l I
Attachment to P-96025 Page 27 GPUN Comparison Data TURBINE BUILDING - LEVEL 5 BATTERY ROOM - CONCRETE FLOOR LWon Sample Count GPUN I SEG GPUN [ SEG GPUN l SEG Number Number Time cts /15 sec/125 cm2 cpm /125 92 dpm/100 cm2 Battery Room 1
15 127 146 508 592 1822 2310 Battery Room 2
15 167 155 AaB 620 2396 2420 Battery Room 3
15 160 143 640 572 2296 2232 Battery Room 4
15 155 137
~ 020 548 2224 2139 Battery Room 5
15 157 154 6f6 616 2253 2404 Battery Room 6
15 139 152 556 608 1995 2373 Battery Room 7
15 146 145 584 580 2095 2263 Battery Room 8
15 138 155 552 620 1880 2420 Battery Room 9
15 170 135 680 540 2439 2107 Battery Room 10 15 159 174 636
"'M
??n?
2716 AVERAGE 182 180 607 599 2178 2334
)
Pye1 1
7
-.-_.---.~-._...----.-- -- ---.-. - - -
I Attachm:nt to P-96025 Page 28 GPUN Comparison Data 1
i 1
l ELECTRICAL WAREHOUSE -#14 CONCRETE FLOOR e
i j
Location Sample Count GPUN SEG GPUN [ SEG GPUN l SEG
~
Number Number Time cts /18 sec/128 cm2 cpm /128 cm2 4 --W106 cm2 j
Electai Shop 1
15 N/T 205 N/T 820 N/T 3200 j
Electral Shop 2
15 N/T 194 N/T 776 N/T 3028 i
Eleciical Shop 3
15 N/T 202 N/T 808 N/T 3153 Electdcal Shop 4
15 N/T 212 N/T 848 NIT 3=
E!+7/ai Shop 5
15 N/T 201 N/T 804 N/T 3138 Electncal Shop 6
15 N/T 189 N/T 756 N/T 2950 i
Electrical Shop 7
15 212 188 848 752 3M2 2935 l
Elecidcal Shop 8
15 201 172 804 688 2884 2685
{
Elecidcal Shop 9
15 170 158 680 632 2439 2466
}
Eleciical Shop 10 15 195 188 780 752 2798 2935 j
Elecrai Shop 11 15 180 183 720 732 2M3 2857 Elecrai Shop 12 15 187 187 748 748
'*m3 is ;
l EiedraiShop 13 15 193 187 772 748 4tiv i-1 E!6Grai Shop 14 15 197 184 788 738 2627 2 erd j
E16cralShop 15 15 194 176 776 704 ~
2784 2747 i
{
E!+7f aiShop 16 15 196 192 784 768 2813 2997 l
Electdcol Shop 17 15 174 183 696 732 2497 2857 Electrai Shop 18 15 215 169 860 676 W5 2aM i
Electral Shop 19 15 169 168 676 672 2425
_M H Electdcal Shop 20 15 169 168 676 672 2425
'm M i.'
AVERAGE 189 188 788 741 2718 2892 i
NIT = NOT TAKEN, SEG AND ORISE SURVEY OF THESE POINTS HAD 1
l BEEN COMPLETED PRIOR TO OUR ARRlVAL AT ELECTRICAL WAREHOUSE #14. THESE POINTS HAD NOT BEEN MARKED, HENCE GPUN COULD NOT PERFORM A REPLICATE SURVEY l
AT THESE LOCATIONS.
l i
i i
J I
Pa9e2 j
l
{
Attachm:nt to P-96025 Page 29 GPUN Comparison Data i
REACTOR BUILDING - LEVEL 3 RESIN CHANGE OUT AREA - CONCRETE FLOOR j
Leon Sample Count GPUN SEG GPUN I SEG GPUN I SEG
{
Number Number Time cts /15 sec/125 cm2 cpm /125 cm2
)
d sin /100 cm2 j
RESIN CHANGE OUT AREA 1
15 364 326 1456 1304 5223 wJe i
}
RESIN CMANGE OUT AREA 2
15 942 1008 3768 4032 13517 15735 l
RESIN CHANGE OUT AREA 3
15 421 400 1684 1600 6041 62 4 RESIN CHANGE OUT AREA 4
15 324 311 1296 1244 We 4a55 RESIN CHANGE OUT AREA 5
15 333 366 1332 1464 4778 5713 i
RESIN CHANGE OUT AREA 6
15 392 419 1568 1676 5625 Su]
RESIN CHANGE OUT AREA 7
15 398 378 1592 1512 5711 5900
)
RESIN CHANGE OUT AREA 8
15 1515 1481 Anaa 5924 21740 23118 l
RESIN CHANGE OUT AREA 9
15 439 465 1756 1860 6300 7259 I
RESIN CHANGE OUT AREA 10 15 336 337 1344 1348 4822
'Nm l
RESIN CHANGE OUT AREA 11 15 487 467 1948 1868
- 5 72sc l
RES!N CHANGE OUT AREA 12 15 2137 1989 BEAa 7956 3P"e5 31048
}
RESIN CHANGE OUT AREA 13 15 490 521 1960 2"
7031 8133 l
RES!N CHANGE OUT AREA 14 15 349 407 1396 1628
==
RESIN CHANGE OUT AREA 15 15 1449 1350 5796 m353 I
2Urv3 21073 RESIN CHANGE OUT AREA 16 15 711 662 2844 2648 10203 103M RESIN CHANGE OUT AREA 17 15 346 335 1384 1340 4965 5229 RESIN CHANGE OUT AREA 18 15 961 1018 SS" 4072 13790 15891 RESIN CHANGE OUT AREA 19 15 694 665 2ito 2aan
==
jnw i
RESIN CHANGE OUT AREA 20 15 326 323 1304 1292 4678 m2 i
i l
AVERAGES 671 641 2683 2644 9624 10324 i
i l
I l
l l
f i
t I
Page 3 1,
i I;
Attachment to P-96025 Page 30 NRC Concern 2:
Detennine whether correctionsfor "hard to detect nuclides" (HTDNs) should be made to survey results in unafected areas that exceed 25 percent of the surface contamination limits.
PSCo/WT Paenanw:
PSCO does not believe that the FSV Final Survey Plan and procedures should be revised to account for HTDNs when contamination is found in unaffected areas.
For unaffected areas where licensed material is not expected, the guideline values (GLV) are the levels included in Regulatory Guide 1.86, Termination of Operating Licenses for Nuclear Reactors, which are 5,000 dpm/100 cm (average) and 15,000 dpm/100 cm 2
2 (maximum). The FSV Final Survey Plan requires the following for unaffected areas:
Surveys will be conducted with instruments having an MDA less than 25% of the 2
GLV (1,250 dpm/100 cm ),
An investigation will be performed if greater than 10% of the measurements exceed 25% of the GLV (1,250 dpm/100 cm), or if any one individual 2
measurement indicates the presence of licensed material in excess of 50% of the 2
GLV (2,500 dpm/100 cm ),
J If as a result of the investigation, greater than 25 percent of the measurements are verified to exceed 25% of the GLV (1,250 dpm/100 cm), or if any one 2
individual measurement is verified to indicate the presence of licensed material i
in excess of 50% of the GLV (2,500 dpm/100 cm ), the area will be reclassified j
2 as affected.
In affected areas, site-specific guideline values (SGLV) have been established which account for the presence of HTDNs. The FSV SGLVs are 4,000 dpm/100 cm (average) 2 2
and 12,000 dpm/100 cm (maximum). In affected areas, an investigation is performed for individual measurements in excess of 75 percent of the SGLV (3,000 dpm/100 cm ),
2 The only effect of using the existing, uncorrected GLVs for unaffected survey units is slightly different action levels. Using corrected release limits for unaffected areas would 2
result in total surface activity reclassification action levels of 2,000 dpm/100 cm 2
i (individual measurement) and 1,000 dpm/100 cm (Ereater than 25% of the i
measurements), versus the current requirements of 2,500 dpm/100 cm2 (individual 2
measurement); and 1,250 dpm/100 cm (greater than 25% of the measurements).
1 i
Attachment to P-96025 Page 31 PSCo considers that the current action level of 2,500 dpm/100 cm (50% of the GLV) 2 for reclassification of unaffected areas is sufficient for the following reasons:
Licensed material is not expected in unaffected areas; The action level provides sufficient margin to ensure that areas do not exceed the 2
SGLV of 4,000 dpm/100 cm ;
2 The 25% GLV (1,250 dpm/100 cm ) action level for 10% of the measurements i
e in unaffected areas is less than the 75% SGLV (3,000 dpm/100 cm ) action level 2
for 10% of the measurements for suspect affected areas; The 50% GLV (2,500 dpm/100 cm') action level for individual measurements in unaffected areas is less than the 100% SGLV (4,000 dpm/100 cm ) action level 2
for individual measurements for suspect affected areas; and Revising the action level to 2,000 dpm/100 cm to account for HIDNs would 2
require revisions to procedures, training plans, and numerous release records that are nearly complete. PSCo does not consider that this effort is justified, 1
?
l i
i l
Attachment to P-96025 Page 32 NRC Concern 3:
Determine whether innstigations of suspect measurements were adequately conducted tojustify removing the originalmeasurementfrom the surwy data base.
PSCo/WT hmane:
l PSco believes that the documentation describing the investigation and disposition of the l
measurement results in the referenced survey packages is adequate. In all instances, the investigation states that the initial result could not be duplicated, that the initial result would be removed from the measurement set used for statistical evaluation, and that the additional measurements would instead be included for statistical evaluation.
FSV Final Survey Plan for Site Release, Section 3.8.11, Investigation, describes the general considerations for conducting an investigation, and for the disposition of measurement results depending upon the outcome of the investigation.
L For instances where investigative actions include the collection of additional fixed point measurements, a scan survey is performed at the initial measurement location and surrounding surfaces to identify the presence of elevated activity. The additional measurements are then collected at the initial measurement location and from the surrounding surfaces, ensuring that a measurement is also collected from any location of elevated activity identified during the scan.
The initial measurement result of final survey may be removed from the measurement set used for statistical evaluation in the event that the result cannot be duplicated at the same survey measurement location, appears to be an anomaly, or investigative actions determine that the result is unlikely to be due to licensed material. For such instances, the initial measurement result is not simply excluded from the Final Report, rather the initial measurement result is referenced in the text of the Investigation section of the Final Report. The additional measurements collected during the investigation are then i
substituted for the initial result, included in the statistical evaluation, and are considered by PSCo to be most representative of the final condition.
i l-i
Attachment to P-96025 Page 33 NRC Concern 4:
Detennine whether scan surwy cowrage percentage in nonsuspect afected areas should be increased.
PSCo/WT Resnonse:
The minimum requirements for final survey of unaffected survey units is defined by the FSV Final Survey Plan for Site Release as follows:
... a scan of approximately 10% of the accessible surface area comprisingfloors and mils below 2 meters,....
l For survey units 11500 A minimum of 30 measurement locations square meters For survey units > 1500 A minimum of 1 measurement location for square meters each 50 square meters surveyed."
The minimum requirements for final survey of non-suspect affected survey units is defined by the FSV Final Survey Plan for Site Release as follows:
... non-suspect affected survey units above 2 meters...
For survey units 1600 square meters A minimum of 30 measurement locations For survey units >600 square meters A minimum of 1 measurement location for each 20 square meters l
surveyed."
For non-suspect affected survey units, the minimum frequency established for fixed point measurements of total surface activity is 1 measurement per 20 square meters. This is implemented at FSV through the performance of a scan survey over an area of not less than 1 square meter at each measurement location to identify the location of highest residual activity, at which the fixed point measurement is then collected. This practice results in a minimum scan fraction of 5% of the surfaces included in non-suspect affected survey units.
4 i
1 Attachment to P-96025 Page 34 PSCo believes that the minimum scan fraction established in the FSV Final Survey Plan for Site Release for non-suspect affected survey units is adequate based on the following additional considerations.
1.
For unaffected survey units, there is no requirement to perform scan surveys on the upper walls and ceiling surfaces as is required for non-suspect affected survey units which are by definition, comprised of upper wall and ceiling surfaces. Instead, the approach for unaffected survey units involves the selection of those surfaces to be included in the scanned fraction from those surfaces having the highest potential for residual activity. (e.g., floor and lower wall surfaces) 2.
Similarly, in affected areas, the floor and lower wall surfaces ate considered as having the highest potential for residual activity and are classified as suspect affected. These surfaces receive 100% scan survey coverage, which is far in excess of the minimum scan coverage required for unaffected survey units.
l 3.
For non-suspect affected survey units, FSV has implemented the practice of collecting the fixed point measurements at the locations of highest activity identified within the scanned fr-Jon. This serves to ensure that locations where significant residual activity exists will be included in the measurement set for comparison against the Administrative Action Levels.
4.
FSV has established Administrative Action Levels for individual measurement results, and for the average of the measurement results collected from within a given survey unit. These action levels are based on a fraction of the site-specific guideline value established for affected survey units at FSV. Measurement results in excess of the action levels serve to initiate investigative actions. Investigative actions often include additional scan survey and fixed-point measurements which result in an increase in the scan fraction in excess of the minimum requirement for instances where residual activity is present.
I 5.
Prior to final survey, the surfaces above 2 meters within affected areas are routinely scanned during the performance of Characterization and/or Remediation surveys. These surveys are performed where the potential I
exists for residual activity (e.g., the identification of significant activity on the floor and lower wall surfaces, the identification of airborne j
radioactivity, the incidence ofleaks or spills, etc.). The extent of the scan i
survey is dependent upon the magnitude of the residual activity found on l
the floor and lower wall surfaces, and upon the results of the scan survey j
performed on upper wall and ceiling surfaces. In the event that residual 4
a
I Attachment to P-96025
[
Page 35 activity is identified on upper wall and ceiling surfaces at levels which approach the action levels for final seney, the scope of the Characterization survey is increased. Although the minimum scanned fraction is currently not formally defined for Characterization and/or Remediation surveys, and is dependent upon the results of these scan surveys, these surveys also serve to increase confidence that residual activity above 2 meters has been adequately characterized.
i l
I l
I
-