Experience Gained and New Correction Techniques Developed based on the Mixed Waste Focus Area NDA System Capability Evaluation Project
|
Dorothy Davidson, Marcel Villani
CANBERRA Industries
Abstract
Canberra Industries participated in the EM30 sponsored Mixed Waste Focus Area (MWFA) NDA System Capability Evaluation Project in late 1997 and early 1998, providing three NDA trailers for evaluation: Segmented Gamma Scanner (SGS) with isotopics capability, High Efficiency Passive Neutron Counter (HENC), and a multiple germanium detector Gamma NDA and Isotopics System (IQ3). The project evaluated the performance of these three systems as well as NDA systems provided by other commercial vendors and DOE funded facilities. The tests included 11 surrogate drums and 21 real waste drums generated at RFETS and stored at INEEL. These drums represented the range of matrices and gram loadings that INEEL will be required to assay to characterize their waste for WIPP certification.
The demonstration project provided an opportunity to assay several difficult matrices and source configurations that challenged the capabilities of each of the NDA systems. The data from this project have provided valuable experience, which is not routinely available to a commercial vendor, in the measurement problems associated with assaying matrices such as sludges and MSE salts, as well as unexpected source configurations and isotopic ratios. This paper covers a discussion on experience gained, validating both the successes and limitations of each of these systems as well as defining the need for test data to evaluate new assay correction and data integration techniques to improve the overall precision and accuracy of the assay results.
The results obtained with these instruments demonstrates the need for the integration of results from both the passive neutron and quantitative gamma/isotopics system in order to achieve the best possible characterization data over the range of conditions encountered.
1. Introduction:
In late 1997 and early 1998, the Mixed Waste Focus Area (MWFA) sponsored an NDA demonstration project at INEEL. The purpose of the demonstration project was to evaluate the capabilities and determine deficiencies of commercially available mobile NDA systems to facilitate resource allocation to areas requiring development. Canberra Industries participated in the project with three mobile NDA systems:
- A Segmented Gamma Scanner (SGS) with an additional LeGe detector for plutonium isotopics measurements,
- A High Efficiency Passive Neutron Counter (HENC) with Add-A-Source matrix correction and multiplicity capability, and
- A multiple detector, shielded, gamma assay system with transmission correction and plutonium isotopics capability (IQ3).
The demonstration project consisted of performing eight replicate counts on a group of 32 drums. Of the 32 drums, 11 were surrogates designed to represent the measurement conditions found at INEEL and 21 were real waste drums originally generated at the Rocky Flats Environmental Technology Site (RFETS).
2. Measurement Technique for Each System as Submitted:
2.1 SGS
The
SGS utilizes a collimated, high purity germanium detector that is mounted
on a vertical drive. The detector performs gamma assay measurements
on a rotating 55 gallon (200 liter) drum by measuring and quantifying
the material present in 9 vertical segments. An external transmission
source on the opposite side of the drum is used to measure the average
density of each segment of the drum to correct for absorption of the
gamma radiation in the matrix.
A second germanium detector, optimized for low energy and high resolution (LeGe), is mounted in a fixed position with a wide field-of-view collimator. The LeGe detector collects a single average spectrum of the drum that is then analyzed using the Multi-Group Analysis (MGA) software to obtain the isotopic ratios of the plutonium as well as ratios to Am-241, U-235, and Np-237. MGA was originally developed by Dr. Ray Gunnink at Lawrence Livermore National Laboratory for safeguards applications. Dr. Gunnink later enhanced MGA for waste applications under a consulting contract for Canberra.
For normal assays, the quantified value of the Pu-239 isotope is corrected for unresolved matrix or self-attenuation errors using a differential peak absorption technique. This corrected assay result is then applied to the plutonium ratios obtained from the MGA measurement to obtain the total plutonium result. Any other directly measured gamma emitters from the quantitative assay such as U-238 are then added to the result.
2.2 HENC
The HENC is a 4p passive neutron coincidence counter, optimized to provide a flat response throughout the assay cavity. The counter utilized 113 He-3 proportional tubes embedded in a High Density Polyethylene moderator to provide a detection efficiency of 31% for Pu-240 neutrons. While the counter is normally operated in the multiplicity data acquisition mode utilizing a Canberra 2150 multiplicity shift register, the software allows either multiplicity analysis or standard coincidence counting algorithms to be used.
The
system includes an Add-A-Source module to provide measurement correction for
matrix and moderator effects. The Add-A-Source (AAS) correction technique, which
was developed by LANL, uses a small 252Cf source on the outside of
the drum to correct for the matrix effects on the plutonium inside of the drum,
improving the assay result. The AAS is positioned at multiple positions along
the side of the drum.
The passive neutron assay measures the Pu-240 effective (combination of spontaneous fissions from Pu-238, Pu-240, and Pu-242) in the drum. Since the SGS system had been through the program prior to the HENC, and the IQ3 system was scheduled for participation after the HENC, the plutonium isotopics results from the SGS system were used to provide measured isotopics for the HENC.
2.3 IQ3
The IQ3 system utilizes a six inch (15cm) 4p low background steel shield with three coaxial germanium detectors for performing quantitative gamma assay measurements and three LeGe detectors for measuring the plutonium isotopics. Three transmission sources are mounted opposite the coaxial detectors to perform the same type of transmission measurements described above. The quantitative measurements are determined both by summing the transmission corrected result for each detector, and by summing the spectra and performing a quantitative assay based on the average density of the drum.
Results from the quantitative measurements and plutonium isotopics measurements are combined in the same manner as described for the SGS system.
3.0 Experiences:
Prior
to this demonstration project, most of Canberra’s experience with these
NDA systems had been limited to measurements of standards in either benign or
homogeneous matrices, and mobile characterization contracts of CH-TRU waste
matrices and plutonium gram loadings that were well within the known measurement
limitations for each system. Prior to starting the demonstration project, it
was recognized that many of the drums to be assayed would be outside of the
administrative limits for one or more of these systems, therefore, results would
be suspect.
The drums assayed as a part of the demonstration project included a wide variety of matrices including: an empty drum, combustibles, MSE salts, metals, plastics, and organic and inorganic sludges. The gram loadings ranged from less than 0.1 g Pu to levels approaching the 200 g limit. Several of the surrogate drums were intentionally built to simulate difficult measurement conditions.
The order of participation in the demonstration project was the SGS in October/November 1997, the HENC in December 1997/January 1998, and the IQ3 in February/March 1998. After submitting results for each system to the scoring group for the project, Canberra received a preliminary report on which drums had passed and which had failed for precision and accuracy. The criteria used for Pass/Fail were those defined for the PDP measurements for benign matrices.
Based on the preliminary report, Canberra re-evaluated the data to try to find measurement problems that were missed in the initial data review. For the IQ3 system there were obvious benefits from identifying weaknesses in our data review process prior to reviewing the IQ3 data. As a result of the re-evaluation process at Canberra, several improvements in algorithms and data review techniques were developed.
4.0 Measurement Problems encountered:
4.1 SGS
Initial assay results from the SGS system followed the standard assay technique described above. Obvious problems encountered in the initial data review included:
- Gram loadings below the expected sensitivity of the SGS system based on the count times used.
- Drum densities above the acceptable range for transmission correction
- A couple of drums with very high Am-241 levels that were causing pile-up and peak shape problems for the coaxial detector. (The LeGe detector used a thin Cd absorber to reduce the pileup effects from the 59 keV gamma from the Am241, the coax detector did not have an absorber).
After the initial scoring results were returned, a second review of the data identified the following additional problems:
- For some drums the Am-241 and Pu were not commingled. The MGA code assumes that all material is commingled, therefore, the software did not calculate a proper ratio for the Am-241 content. In cases where the higher energy lines from the Am-241 were directly measured in the quantitative gamma assay, the assay results were compared to the MGA calculated Am value. If significant disagreement occurred, the quantitative assay results were used.
- Occasional transcription errors between the analysis software and a separate spreadsheet which was used to combine results from the MGA and the quantitative assay in order to calculate values necessary for reporting to WIPP. Transcription errors were found in all three systems. To eliminate this problem, new total measurement uncertainty software was developed to directly extract the measurement values from the original assay files thus eliminating the need to manually enter data.
4.2 HENC
Initial assay results from the HENC system were impacted by procedural and physical limitations. The procedural problems were due to the constraints necessary to complete the assay of all drums within the time allotted for the demonstration.
- The high altitude at INEEL results in a high cosmic background. The assay protocol developed to meet the demonstration project schedule did not allow an adequate time for background counts.
- The large moderating effects of the sludge drums were beyond the administrative limits for the counter requiring manual data review. (An attempt was made to characterize the background effects of the various matrices at INEEL, but this was not permitted under the scope of the demonstration project)
- Fluctuations in the Pu isotopics results (from the gamma system) for some of the drums caused significant precision problems for the combined neutron results.
- Inadequate count times for a couple of drums caused them to fail to achieve the required precision. (The count to MDA feature in the software was not used so that a fixed counting schedule could be applied. Instead, the maximum count time was fixed at 30 minutes per assay during the day and one hour maximum overnight to achieve the required throughput.)
After the initial scoring results were returned, a second review of the data identified the following additional problems:
- On one drum a second review of the data indicated that the U-238 content (>3 kg), identified from the SGS assay, was a significant contributor to the passive neutron signal. This was not automatically accounted for in the software.
4.3 IQ3
The IQ3 has significantly lower detection levels than the SGS system, therefore, drums with lower gram loadings did not cause a sensitivity problem. At the other end of the measurement range, there were a few drums with activities that significantly exceeded the administrative limits for the IQ3 system and could not report results due to significant pile up and peak shape distortion problems.
5.0 Fixes and New Techniques:
As a result of the lessons learned from this project, Canberra has adopted several changes and new techniques that are being applied to the present mobile NDA systems and will be utilized in future product developments. Some of these were techniques that were already known to be important, but were not able to be covered under the schedule or the provisions of the demonstration project.
5.1 For the Gamma systems these enhancements include:
- Cd liners will be used over all detectors to improve the ability to assay drums with higher gram loadings or with significant Am-241 content. An exception is a measurement situation where the goal is to obtain the lowest possible detection levels.
- Assays on some high density matrices such as sludges will utilize a summed spectrum/average matrix density analysis rather than attempting to utilize transmission correction. This was shown to produce better results because the transmission source data is inadequate and the matrix can be assumed to be relatively uniform.
- Data review techniques have been improved to look for indications of disassociated nuclides such as U-235, U-238 and Am-241 where the MGA ratios may be suspect. Some indicators of this would include separation dates which are significantly older than the known material age based on process knowledge, and significant variations between the assay and MGA calculated results for the isotope in question. A final technique will evaluate, qualitatively, variations in the ratios between the Pu and Am or U for the different segments of the drum.
- New software has been developed to automatically transfer data between the different analyses in order to eliminate the problems associated with transcription errors.
5.2 For the HENC improvements include:
- Implementation of a background correction technique for drums containing large matrix mass. This technique allows correction of the change in cosmic-ray background due these materials. This approach requires reasonable background matrix measurements to be made at a site before it can be implemented for that facility.
- Implementation of a multiplicity based background reduction technique to improve the precision and accuracy of the measurement.
- For assays where the plutonium isotopics measurement has significant errors, and a standard type of material is expected based on process knowledge, a set of default isotopics will be used for the plutonium isotopics. If measured Am-241 results are available this will be applied if there is a significant deviation from the expected default isotopics.
Finally, reviewing both the gamma and the neutron assay data, provides a significant improvement in the overall drum assay capability. Each of the techniques has advantages and limitations. However, by using two complementary techniques (passive neutron and gamma spectroscopy), significant improvements in the overall precision and accuracy can be achieved in many cases. In some cases the additional information provides evidence, through an expert review, to know that there are potentially significant errors in one or both of the results.
6.0 Summary
This demonstration project has been an extremely valuable tool for a commercial company, such as Canberra Industries, to obtain data on difficult matrices and assay conditions in order to improve the quality of the products and services that are offered. While Canberra’s current NDA techniques meet the regulatory requirements for many waste types, further enhancements have been identified that require additional research and development funds. These enhancements will improve the accuracy and precision of assays for more difficult waste types found throughout the DOE complex.
|
| Top of Page |