LabSOCS™ vs. Source-Based Gamma-Ray Detector Efficiency Comparisons For Nuclear Power Plant Geometries
|
David Groff CANBERRA Industries, Meriden, CT., USA
INTRODUCTION
The sample geometry modeling and efficiency calibration file generation were performed using the Geometry Composer feature of the Genie 2000 Version 2.0 and Gamma Acquisition and Analysis (GAA) software packages. The resulting sample geometry files and Genie 2000 efficiency calibration files were delivered to Sequoyah NP for testing on a laboratory-based gamma spectroscopy system. This system includes the coaxial IGe detector (serial number 7386) which had been characterized by CANBERRA.
A total of 48 different sample counting geometries were specified for this LabSOCS modeling project. Each of these 48 geometries represents a unique combination of a particular sample container, a particular source matrix, and a specific source-to-detector end cap distance. A total of 16 different sample containers and fixtures weresupplied by TVA. These components were inspected and measured by CANBERRA for accurate determination of container dimensions, absorber thickness values, and source-to-detector distance values needed as part of the modeling process. The critical assumptions and methods used to determine the dimensions and final parameter input required for the LabSOCS models are described in this report.
LabSOCS MODELING
The following sections detail the LabSOCS geometry modeling and efficiency file generation process for the 48 different sample geometries.
SAMPLE CONTAINERS AND MATERIALS
A total of 16 different sample containers and sample matrix materials were used in this study and are listed in Table 1:Sample containers and materials.
| Sample Container Description | Sample Matrix Description |
| 50 mm diameter Falcon Petri disk F & J “C” charcoal filter with plastic case F & J charcoal filter with tall metal case 20 mL Packard liquid scintillation vial (LSC) 0.25 liter GA-MA liquid Marinelli beaker 0.50 liter GA-MA liquid Marinelli beaker 1.0 liter GA-MA liquid Marinelli beaker 4.0 liter GA-MA liquid Marinelli beaker 0.12 liter Alpha wide-mouth Poly bottle 0.25 liter Nalgene wide-mouth Poly bottle 0.50 liter Nalgene wide-mouth Poly bottle 1.0 liter Nalgene wide-mouth Poly bottle 25 cc GA-MA gas sampler 1.24 liter GA-MA gas Marinelli beaker 14 cc glass serum vial Aluminum ring with thin tape layer |
47 mm diameter paper filter Charcoal (carbon and air) Charcoal (carbon and air) 20 mL water 250 mL water 500 mL water 800 mL water 3500 mL water 120 mL water 250 mL water 500 mL water 1000 mL water 25 cc air 1240 cc air 14 cc air Point source |
A set of pre-fabricated acrylic plates and tubes intended to hold these samples at various reproducible positions relative to the detector end cap were provided. These sample positioning components and the six reference counting configurations used by TVA Sequoyah are summarized in Table 2:
LabSOCS MODEL DESCRIPTIONS
Each of the LabSOCS models was designated with a unique number, ranging from Model #01 to Model #48. Each model represents a specific combination of sample container, sample matrix and sample position.
Counting geometries used.
| Position Designation | Acrylic Absorber Thickness | Source-to-Detector Distance |
| No Shelf Shelf 0 Shelf 1 Shelf 2 Shelf 3 Shelf 4 |
No absorber present 5.44 mm (2) 8.01 mm (2) 8.01 mm (2) 1.20 mm (2) 1.20 mm (2) |
0.00 mm (1) 5.44 mm (2) 29.5 mm (2) 97.0 mm (3) 50.1 cm (4) 100.1 cm (4) |
| Notes: | 1 - Used for Marinelli beakers 2 - Dimensions measured with a micrometer (smallest units = 0.001 inches) 3 - Dimensions measured with a metal ruler (smallest units = 0.5 mm) 4 - Dimensions measured with a flexible tape (smallest units = 0.063 inches) |
|
The applicable dimensions listed above were used as part of the parameter input
for the 48 different LabSOCS models, with each model representing a specific
combination of sample container, sample matrix and sample position. For some
models, the total source-detector distance was increased beyond the default”
values listed above, due to additional spacing contributed by the designated
acrylic centering plate or a ortion of the sample container itself. The acrylic
centering plates also contributed some additional photon ttenuation near the
base of the 20 mL LSC vial, 14 cc glass serum vial and the 0.12 liter Alpha
Poly bottle containers. These factors have been included, when appropriate,
in the final LabSOCS models described in the project documentation provided
to TVA. All of the sample configurations involving a water matrix were modeled
twice with LabSOCS, once with an actual water matrix (as appropriate for actual
samples) and again with a “water equivalent” solid epoxy matrix
(used by Analytics to prepare “water-equivalent” radioactive standards).
All of the sample configurations involving an air matrix were modeled twice
with LabSOCS, once with actual air as the material (as appropriate for actual
samples), and again with an “airequivalent” polystyrene bead matrix
(used by Analytics to prepare “air-equivalent” radioactive standards).
This comparative modeling was done to evaluate the expected magnitude of variation
in efficiency for the water
vs. “water-equivalent” and air vs. “air-equivalent”
sample configurations.
The Geometry Composer option of the CANBERRA Industries Genie 2000 Version 2.0 and Gamma Analysis Version 2.0A software packages was used to create the 48 specific models for this project. Two special materials were defined and added to the Materials Library file: “Charcoal” as 100% carbon (mass percentage) with a default density = 0.59 g/cc, and “smltdair” as 93.6% polystyrene and 6.4% air (mass percentages) with a default density of 0.03 g/cc.
LabSOCS SAMPLE MODELS
A summary list of all models created for this project is provided in Table 3.
Geometry models that were created.
| Model Number | Sample Container Description | Shelf | Geom Code | Sample Matrix Description |
| 01 |
Polystyrene Falcon Petri dish Polystyrene Falcon Petri dish Polystyrene Falcon Petri dish F&J “C” (plastic) charcoal filter F&J “C” (plastic) charcoal filter F&J “C” (plastic) charcoal filter F&J (metal case) charcoal filter F&J (metal case) charcoal filter F&J (metal case) charcoal filter 20 mL Packard LSC vial 20 mL Packard LSC vial 20 mL Packard LSC vial 20 mL Packard LSC vial 20 mL Packard LSC vial 20 mL Packard LSC vial 0.25 mL GA-MA Marinelli beaker 0.25 mL GA-MA Marinelli beaker 0.50 mL GA-MA Marinelli beaker 0.50 mL GA-MA Marinelli beaker 1.0 liter GA-MA Marinelli beaker 1.0 liter GA-MA Marinelli beaker 4.0 liter GA-MA Marinelli beaker 4.0 liter GA-MA Marinelli beaker 0.12 liter Alpha poly bottle 0.12 liter Alpha poly bottle 0.25 liter Alpha poly bottle 0.25 liter Alpha poly bottle 0.50 liter Alpha poly bottle 0.50 liter Alpha poly bottle 1.0 liter Alpha poly bottle 1.0 liter Alpha poly bottle 25 cc GA-MA gas sampler 25 cc GA-MA gas sampler 25 cc GA-MA gas sampler 25 cc GA-MA gas sampler 25 cc GA-MA gas sampler 25 cc GA-MA gas sampler 1.24 liter GA-MA gas Marinelli 1.24 liter GA-MA gas Marinelli 14 cc glass serum vial 14 cc glass serum vial 14 cc glass serum vial 14 cc glass serum vial Aluminum ring with thin tape layer Aluminum ring with thin tape layer Aluminum ring with thin tape layer Aluminum ring with thin tape layer Aluminum ring with thin tape layer |
0 1 2 0 1 2 0 1 2 0 0 1 1 2 2 0 0 0 0 0 0 0 0 1 1 1 1 1 1 1 1 0 0 1 1 2 2 0 0 0 0 2 2 0 1 2 3 4 |
MPF0 MPF1 MPF2 CHF0 CHF1 CHF2 WCF0 WCF1 WCF2 D200 D200 D201 D201 D202 D202 L250 L250 L500 L500 LM10 LM10 LM40 LM40 P121 P121 P251 P251 P501 P501 PLB1 PLB1 G250 G250 G251 G251 G252 G252 GM10 GM10 S140 S140 S142 S142 PSG0 PSG1 PSG2 PSG3 PSG4 |
47 mm paper filter 47 mm paper filter 47 mm paper filter “Face-loaded” charcoal “Face-loaded” charcoal “Face-loaded” charcoal “Face-loaded” charcoal “Face-loaded” charcoal “Face-loaded” charcoal 20 mL water 20 mL epoxy 20 mL water 20 mL epoxy 20 mL water 20 mL epoxy 250 mL water 250 mL epoxy 500 mL water 500 mL epoxy 1000 mL water 1000 mL epoxy 3500 mL water 3500 mL epoxy 120 mL water 120 mL epoxy 250 mL water 250 mL epoxy 500 mL water 500 mL water 1000 mL water 1000 mL water 25 cc air 25 cc simulated air 25 cc air 25 cc simulated air 25 cc air 25 cc simulated air 1240 cc air 1240 cc simulated 14 cc air 14 cc simulated air 14 cc air 14 cc simulated air Point Source Point Source Point Source Point Source Point Source |
LabSOCS CUSTOMIZED BEAKER FILES
For some of the models listed in the previous table, customized beaker files were created to accurately define the inner and outer wall contours, the material(s), and density value(s) of the container. These files were created using a standard text editor and stored with a *.bkr file extension to allow selection as one of the “complex beaker” templates in the Geometry Composer window. The template file names, corresponding container types, and models using each of these beaker shapes for parameter input are summarized in Table 4:
Sample container files.
| File Name | Container Description |
| TVA20cc.bkr | 20 mL Packard Polypropylene LSC (including acrylic plate). Model #10 - 15] |
| TVA120mL.bkr | 0.12 liter Alpha Polystyrene bottle (including acrylic plate). Model #24 - 25] |
| TVA250mL.bkr | 0.25 liter Nalgene wide-mouth bottle (LDPE). [Model #26 - 27] |
| TVA500mL.bkr | 0.50 liter Nalgene wide-mouth bottle (LDPE). [Model #28 - 29] |
| TVA1l.bkr | 1.0 liter Nalgene wide-mouth bottle (HDPE). [Model #30 - 31] |
| G-130G.bkr | 1.24 liter GA-MA gas Marinelli beaker (Polystyrene). Model #38 - 39] |
| 14ccvial.bkr | 14 cc glass serum vial (including acrylic plate). [Model #40 - 41] |
Copies of these text files and the additional complex beaker files named 130G.bkr and 430G.bkr distributed with the standard CANBERRA software were provided in the final project documentation. The 130G.bkr file represents a 1.0 liter GA-MA Marinelli beaker with Polypropylene walls, used for Models #20 and #21. The 430G.bkr file represents a .0 liter GA-MA Marinelli beaker with Polypropylene, used for Models #22 and 23.
For each of the 48 models listed previously, a detailed description of the parameter values used to define the dimensions and material composition of the container, sample matrix, acrylic shelf absorber layer (if present), and source-to-detector distance is provided in the final project documentation. A written description of each model is provided, followed by the Geometry Composer report and printed copy of the *.GIS text file for that model. The naming convention for the Geometry Composer *.GEO files and corresponding *.GIS files used throughout this project is as follows:
| Model # | *.GEO File Name | *.GIS File Name |
nn |
TVA_nn.GEO |
TVA_nn.GIS |
LabSOCS EFFICIENCY CALIBRATION PROCESS
For each of the sample models described in the previous section, the LabSOCS Version 4.0 software was used to generate a set of mathematically calculated efficiency values for a specified set of energy values. The energy range specified by TVA for all routine gamma spectroscopy measurements at Sequoyah NP was 45 keV to 2000 keV. A customized energy list was created and stored as a text file named TVA1.txt for use in all LabSOCS efficiency alculations for this project. This energy list included 16 energy values ranging from 45 keV to 2000 keV, with appropriate corresponding percent uncertainty values ranging from 10% at low energies to 4% at high energies.Prior to perform in the final LabSOCS modeling using the Geometry Composer, the detector characterization file named 7386.par (created by CANBERRA for TVA’s serial number 7386 coaxial germanium detector) was copied to the Genie2k\isocs\data\Dcg folder on the CANBERRA personal computer used to run the software. This detector characterization file was then used for all LabSOCS modeling and efficiency calculations performed during this project.
With the desired *.GEO file opened in the Geometry Composer window, the Efficiency|Generate efficiency data points option was selected from the menu bar. This action generated the required set of energy/efficiency/ uncertainty data triplets to be used for the final efficiency calibration file. These data triplets were stored in a file named TVA_nn.ECC for each model, where nn is the same two-digit number present in the TVA_nn.GEO file used in Geometry Composer.
For each model, the following steps were then performed:
1. The appropriate *.ECC file was used to generate the final efficiency calibration results, as follows. A Gamma Acquisition and Analysis (GAA) window was launched, and a pre-existing CAM file datasource opened in the GAA window. The Calibrate|Efficiency|By ISOCS|LabSOCS option was selected from the GAA window menu bar. The desired *.ECC file was then selected as the data input file.
2. The traditional “Efficiency” option (counts per gamma) was selected as the appropriate LabSOCS efficiency calculation factor for all models.
3. When the “Calibrate by ISOCS/LabSOCS:Efficiency Results” dialog box was displayed, the Show action button was used to display the Dual, Empirical and Linear efficiency curves. The order of the polynomial for the efficiency curve type was modified as necessary to achieve the best curve fit.
4. From the “Calibrate by ISOCS/LabSOCS:Efficiency Results” dialog box, the Report action button was used to generate a one-page report of the LabSOCS efficiency results. The “Geometry Description” field for each of these reports has the format TVA_nn, where nn is the model number. Each of these reports was included in the final project report.
5. From the “Calibrate by ISOCS/LabSOCS:Efficiency Results” dialog box, the Store action button was used to save the results as a standard Genie 2000 efficiency calibration file in the Genie2k\Calfiles folder. The naming convention used to store these files is TVA_nn.CAL, where nn is the same twodigit model number in the corresponding *.GEO, *. GIS and *.ECC file names. The “Eff.Geom.ID” field for each *.CAL file is identical to the “Geometry Description” field in the report of LabSOCS efficiency results, i.e., TVA_nn, where nn is the two-digit model number.
6. From the “Calibrate by ISOCS/LabSOCS:Efficiency Results” dialog box, the Finish action button was used to close the dialog box and return to the GAA window.
7. The *.CAL file created in Step 5 was then opened as a CAM file datasource
in the GAA window. The Calibrate|Efficiency show option was
selected from the menu bar, and the Print action button used
to generate printed plots of the appropriate efficiency curve type. These plots
were included in the final report.
Note: These curve plots were printed from the GAA window with the TVA_nn.CAL
datasource opened to insure that the datasource file name included on the plot
would match the actual TVA_nn.CAL file name to avoid possible confusion when
reviewing these plots at a later time.
SOURCE-BASED EFFICIENCY CALIBRATION PROCESS
Equipment Utilized
The detector utilized in these source-based measurements was the same CANBERRA coaxial IGe detector used in the LabSOCS efficiency calculations, serial number 7386. The Peak-to-Total calibrations were performed using a CANBERRA Model S-PTC Peak-to-Total Calibration source set. A set of mixed gamma efficiency calibration sources were purchased from Analytics, Inc. and were fabricated from the late 2001 NIST source batch.
Software Utilized
The detailed isotopic information from the Certificates of Calibration for the mixed gamma efficiency calibration standards from Analytics was entered using the Genie 2000 Certificate Editor Version 2.1. The Genie 2000 GAA was used to acquire the spectra and perform the efficiency calibration calculations. Finally, the LabSOCS and source-based efficiency results were compared using a custom Visual Basic program written by Greg Landry of CANBERRA Industries. This program, the CANBERRA Empirical Efficiency Point Calculator Program Version 2.0, was modified to calculate the LabSOCS/Source-based efficiency ratios for each of the mixed gamma energies.
Peak-to-Total Calibration
The Peak-to-Total calibration (PTC) was performed per the Peak-to-Total section of the Genie 2000 Operations Manual (March 2001) using the CANBERRA Model S-PTC Peak-to-Total Calibration Source Set. The CSC factors were calculated using the Cascade Summing Correction section of the Genie 2000 Operations Manual.
Source-based Efficiency Calibration
Each mixed-gamma standard was counted on the appropriate fixture shelf using the 20,000 net counts in each certificate peak criteria. Then, the efficiency calibration calculations were performed using the GAA Calibrate|Efficiency|By Certificate File option from the GAA menu bar and the report printed.
Efficiency Comparison Method
The efficiency data set to be analyzed was limited to those geometries with identical matrices which assured comparison consistency. The worksheet – example contained in Appendix I – methodology was used to organize and process the data as follows:
1. The LabSOCS efficiencies and efficiency uncertainty values were entered on the worksheet.
2. The source-based efficiencies and efficiency uncertainty values were entered on the worksheet.
3. The CSC factors were entered for the appropriate energies of 88Y and 60Co and those efficiencies divided by the CSC factor.
4. The uncertainty associated with the CSC correction process was calculated using 5% of the CSC corrected efficiency value for the 88Y and 60Co values only.
5. The total source-based efficiency uncertainty was calculated by summing the source-based efficiency uncertainty and the CSC uncertainty value for the 88Y and 60Co values only.
6. The ratio of the total source-based efficiency uncertainty and the source-based efficiency was calculated for each mixed gamma energy.
7. The ratio of the LabSOCS efficiency value to the source-based CSC corrected efficiency value was calculated for each mixed gamma energy.
8. The total uncertainty associated with each ratio calculation was calculated using the equation in Table 2-Uncertainties of ANSI N42.14 for the ratio of two quantities and associated uncertainties.
9. The ratio value ±1.96 times the calculated ratio total uncertainty from Step 8 was then compared to unity (ratio = 1) using an agreement plot generated from an Excel spreadsheet for each mixed gamma energy.
The above steps were repeated for each geometry comparison.
COMPARISON RESULTS
The agreement plots for several representative geometries included in this study are contained in Appendix II. An examination of each agreement plot demonstrates that all of the 95% confidence intervals (ratio value ±1.96 times the calculated ratio total uncertainty) contain the agreement value of unity. An alternative method of stating this agreement is that the hypothesis that the 95% confidence intervals did not contain the agreement value was rejected in every case.
SUMMARY
This study has demonstrated that the LabSOCS efficiency calibration technique will produce efficiency values which, when corrected for cascade summing effects, will agree with source-based efficiency calibrations for a wide variety of sample and container types which support power plant process, radcon, radwaste and effluent operations. Using the LabSOCS efficiency calibration method will reduce costs associated with purchase, maintenance and disposal of physical sources. In addition, the LabSOCS technique, using the Geometry Composer, will enable count room personnel to produce assay-grade measurements of unique sample/ matrix/container samples such as oil, soil, gravel and certain biological samples presented to the count room for analysis.
REFERENCES
1. F. Bronson & R. Venkataraman, “Validation of the Accuracy of the LabSOCS Mathematical Efficiency Calibration for Typical Laboratory Samples”. 46th Annual Conference on Bioassay, Analytical and Environmental Radiochemistry, November 11-17, 2000, Seattle, WA.
2. “ANSI Standard for Calibration and Use of Germanium Spectrometers for the Measurement of Gamma-Ray Emission Rates of Radionuclides”, ANSI N42.14-1999, Table 2, Page 22.
3. D. Groff, “LabSOCS Geometry Modeling and Efficiency Calibration File Generation for TVA Sample Fixtures”, December 6, 2001.
|
| Top of Page |