NWIS signatures for confirmatory measurements with B33 trainers

Nuclear weapons identification system (NWIS) signatures have been used successfully to confirm that B33 trainer parts, shipped from military bases to the Oak Ridge Y-12 Plant, were, as declared by the shipper, nonenriched uranium. The B33 was a gun-assembled weapon consisting of two components. Verification was accomplished by comparing signatures for B33 trainer parts with signatures for mock-ups made with depleted uranium packaged at the Y-12 Plant in M102 containers. Measurements with a normal production war reserve (WR) Component 2 part and calculations for a normal WR Component 1 part showed the very high sensitivity of the frequency-domain signatures to the presence of enriched uranium. Some measured frequency-domain signatures were greater than a factor of 100 different for the WR units. These verifications alleviated the criticality safety and safeguard concerns for material returned to the Y-12 Plant for dismantlement. This work in 1993 was part of the weapons dismantlement program at the Y-12 Plant, with hardware partially funded by the U.S. Department of Energy (DOE) Office of Arms Control and Non-Proliferation through the National Security Program Office of the Y-12 Plant. The weapons dismantlement program dismantles nuclear weapons components and trainers, safely stores fissile materials, and disposes of the wide variety of materials in an environmentally acceptable manner, meeting all applicable DOE, state, and federal regulations. These verifications were conducted in a timely, reliable manner and produced no false positives for the 512 verifications (263 containers with Component 1 parts and 249 containers with Component 2 parts). Measurement times were 10 min each for time- and frequency-analyzer verifications. As many as 32 verifications were performed in one day (normal eight-hour shift) and 111 in one week. Some of the signatures were different from those of the reference units, but not sufficiently different to be those for enriched uranium components. All of the anomalous signatures (those that differed significantly from the reference) were explained once the M102 containers were opened and the parts removed. The verification measurements revealed that the weights of eight Component 1 parts were not as declared by the shipper. These deviations resulted in the Y-12 Plant receiving more Component 1 parts than declared by the shipper. The Y-12 Plant, through the DOE, Oak Ridge Operations Office (DOE-ORO), resolved these differences with the Albuquerque Operations Office and related shipping personnel. For one Component 1 part and one Component 2 part, the uranium was not in the usual location in the container. The Component 2 part was not at the bottom of the container, and the Component 1 part, which normally had an aluminum spacer under the part at the bottom of the container, was itself at the bottom of the container. The measurements also showed that alternative fixtures were used in the M102 containers for the Component 2 parts. Ten Component 2 parts showed anomalously high values of the coherence between the source and detectors, identifying increased transmission of particles through the parts and suggesting less nonenriched uranium in the containers. This finding was confirmed when the containers were opened and it was found that many holes had been drilled partway into each part, removing 4.5% of the mass of uranium. These verifications demonstrated the use of NWIS signatures for identification of nuclear weapons mock-up parts. The sensitivity of some of the signatures to small changes in packaging and others to mass of uranium inside the container was also demonstrated. Signatures for mock-up units usually require more measurement time than normal WR units because there is no significant fission-induced part of the signal and because only directly transmitted, scattered, and secondary particles reach the detectors. Thus, measurement and identification of normal production weapons components would be easier for NWIS. This was the first use of NWIS in a nonresearch environment and indicated the desirability of automated operation. NWIS signatures have been demonstrated to be adequate for shipper-to-shipper confirmatory measurement within DOE and between the U.S. Department of Defense (DOD) and DOE. A nonintrusive use of NWIS signatures was demonstrated, which would allow the use of this method by foreign nationals at DOE or DOD facilities. A field-deployable system, supported by the DOE Office of On-Site Systems within the Office of Research and Development (NN-20) and the Y-12 Plant, is briefly discussed.