Photonuclear-based Detection of Nuclear Smuggling in
Cargo Containers
J.L. Jones, K.J. Haskell, J.M. Hoggan, D.R. Norman, and W.Y. Yoon
Idaho National Engineering and Environmental Laboratory
Abstract. The Idaho National Engineering and Environmental Laboratory (INEEL) and the Los Alamos National
Laboratory (LANL) have performed experiments in La Honda, California and at the Idaho Accelerator Center in
Pocatello, Idaho to assess and develop a photonuclear-based detection system for shielded nuclear materials in
cargo containers. The detection system, measuring photonuclear-related neutron emissions, is planned for
integration with the ARACOR Eagle Cargo Container Inspection System (Sunnyvale, CA). The Eagle Inspection
system uses a nominal 6-MeV electron accelerator and operates with safe radiation exposure limits to both
container stowaways and to its operators. The INEEL has fabricated custom-built, helium-3-based, neutron
detectors for this inspection application and is performing an experimental application assessment. Because the
Eagle Inspection system could not be moved to LANL where special nuclear material was available, the response of
the Eagle had to be determined indirectly so as to support the development and testing of the detection system.
Experiments in California have successfully matched the delayed neutron emission performance of the ARACOR
Eagle with that of the transportable INEEL electron accelerator (i.e., the Varitron) and are reported here. A
demonstration test is planned at LANL using the Varitron and shielded special nuclear materials within a cargo
container. Detector results are providing very useful information regarding the challenges of delayed neutron
counting near the photofission threshold energy of 5.5 – 6.0 MeV, are identifying the possible utilization of prompt
neutron emissions to allow enhanced signal-to-noise measurements, and are showing the overall benefits of using
higher electron beam energies.
providing real-time, radiographic imaging of cargo
containers. The inspection system uses a nominal 6MeV electron accelerator and operates within
operationally safe radiation exposure limits to both
container stowaways and to its operators.
INTRODUCTION
The
Idaho
National
Engineering
and
Environmental Laboratory (INEEL) and Los Alamos
National Laboratory (LANL) have proposed an
enhanced technology for cargo container inspections.
This proposed concept integrates radiographic
imaging with a photonuclear-based nuclear material
detection. A demonstration test with special nuclear
material and a cargo container is planned in CY02 at
LANL. To support this concept, a test campaign was
undertaken during April 16-19, 2002 in La Honda,
California to match the operational performance of
the Idaho National Engineering and Environmental
Laboratory (INEEL) Varitron accelerator to that of
an Advanced Research and Application Corporation
(ARACOR) Eagle accelerator. The Varitron, shown
in Figure 1, is a transportable, custom-built,
operationally-flexible, selectable-energy (2-12MeV), electron accelerator1,2. The ARACOR Eagle,
shown in Figure 2, is a mobile, cargo inspection unit
FIGURE 1. The INEEL Varitron system.
CP680, Application of Accelerators in Research and Industry: 17th Int'l. Conference, edited by J. L. Duggan and I. L. Morgan
© 2003 American Institute of Physics 0-7354-0149-7/03/$20.00
947
DU target while Detector #1 was located at about 45degress off beam-axis and forward the DU target.
Between these two detectors, Detectors #2 and #3
were equally spaced with the position of #3 closest to
Detector #0. LANL deployed another helium-3
detector assembly that is 2.4-m long, 1.5-m tall, and
0.15-m wide. (See left portion of Figure 3). The
LANL detector was positioned parallel to the beam
centerline at a distance of 1.34-m and opposite the
INEEL detectors. It had its closest end positioned
such that it was aligned with the front face of the
ARACOR shield assembly.
The primary goal of this effort was to identify a
Varitron operating condition that matched the
delayed neutron emission response of a “nominal” 6MeV ARACOR Eagle-like accelerator. This paper
provides neutron detection measurements using the
INEEL detectors. All delayed neutron data are
FIGURE 2. The ARACOR Eagle.
acquired in the time interval ranging from 4.95 to
19.9 ms after each accelerator pulse. All prompt
neutron data are acquired during 0.156 to 4.91 ms
after each accelerator pulse. Specific prompt and
delayed neutron acquisition intervals can still be
optimized.
FIGURE 3. Experimental configuration of the
neutron detectors, the depleted uranium target,
and the ARACOR “nominal” 6-MeV
accelerator and modified shield assembly.
EXPERIMENTAL CONFIGURATION
ARACOR provided a “nominal” 6-MeV
accelerator encased in a proprietary Eagle-like,
photon shield assembly (See white “box” in the
upper right of Figure 3). To support another on-going
project, ARACOR has modified the actual Eagle
collimator assembly design to include a cylindrical
hole having a 10-cm diameter opening. The hole is
axially centered on the electron beam centerline.
Despite the shield modification, the goal to match the
delayed neutron output was not compromised.
Positioned on the beam centerline, and relative to the
accelerator’s photon source, was an air ionization
chamber at 1-m and a 4.4-kg depleted uranium (DU)
target at 1.05-meter. INEEL deployed four
cylindrical, neutron detectors (see lower right portion
of Figure 3) located at 1-m from the DU. Each
INEEL detector weighs about 12 kg and consists of a
76-cm long, 2.54-cm diameter, helium-3 tube
individually encased in a custom moderator and
shield configuration. These detectors have been
specifically designed for pulsed, photonuclear
applications.
Specifically, Detector #0 was
positioned at 90-degrees from the beam axis at the
“NOMINAL” 6-MEV EXPERIMENTS
Many tests using both X-ray ON and OFF
operations were performed over the three-day test
campaign. Results showed the average neutron
background at La Honda, California to be 0.11 ±
0.03 counts per second for each INEEL detector
using the delayed neutron acquisition window
identified above. Since the prompt count is
dependent on the specific experimental configuration
and accelerator operation, no average prompt neutron
background was determined.
Using the experimental configuration described
above, an operational assessment was made to
optimize the electron beam energy for a 50-Hz, 3 µs
pulsed operation of ARACOR’s “nominal” 6-MeV
948
accelerator to obtain a maximum delayed neutron
output from the DU target. Maximizing the
magnetron power and iteratively reducing the
accelerator’s electron gun filament voltage from 0.85
to 0.67 resulted in the maximum delayed neutron
output. This optimized performance remained within
the Eagle’s operational requirements. The
corresponding photon dose rate was about 80 R/min
at one meter.
VARITRON Matching Experiments
With the exception of a symmetric reflection of
the INEEL and the LANL neutron detectors about
the beam centerline, a similar experimental
configuration was established using the INEEL
Varitron instead of the ARACOR accelerator. This
test configuration is shown in Figure 4. No photon
shielding was used forward of the Varitron’s
bremsstrahlung converter.
With this operational condition determined,
several key characterization experiments were
performed with the DU still in position. These
experiments involved the addition of a 30-cm long
beryllium (Be) bar (0.93 kg) or a 20-cm long heavy
water (D2O) container (1.8 kg) inside the hole of the
modified ARACOR shield assembly. Beryllium and
heavy water (D2O) were chosen to help
characterize/identify any differences in the
bremsstrahlung produced by these accelerators at
photon energies less than the DU photonuclear
threshold energy (~5.8 MeV). Specifically, Be and
D2O were selected since they have very low
photoneutron threshold energies: 1.7 MeV and 2.2
MeV, respectively. Each material was separately
positioned at the ARACOR’s electron/photon
converter and aligned with the beam centerline. The
delayed and prompt neutron detection results for
these cases are given in Tables 1 and 2, respectively,
along the corresponding one-sigma uncertainty.
FIGURE 4. Matching experimental configuration
using the INEEL Varitron.
After several iterations it became clear that while
the delayed neutron response from DU could be
matched, it was not possible to completely match the
photon dose at a meter. It appears that the electron
capture efficiency during the acceleration process in
the two accelerator waveguides is different. Hence, it
was decided to match the DU target delayed neutron
response and see what the additional Be and D2O
material responses would indicate. Tables 3 and 4
present the net delayed and total prompt neutron
count rates, respectively, for this series of tests.
Figure 5 presents the energy spectrum of the
Varitron’s accelerated electrons for this operational
condition having an average beam current of 3.3 µA.
A measurement of the beam current of the Eagle-like
system was not available.
Table 1. Net delayed neutron counts per second for the
INEEL detectors positioned at one meter from the DU
target and using the ARACOR “6-MeV” accelerator.
Detector Number with Error
Case
#0
#3
#2
#1
DU Only 1.36 ± 0.11 1.22 ± 0.11 1.06 ± 0.10 1.00 ± 0.10
DU + Be 0.13 ± 0.04 0.15 ± 0.05 0.12 ± 0.04 0.11 ± 0.04
DU+ D2O 0.67 ± 0.08 0.62 ± 0.08 0.57 ± 0.08 0.55 ± 0.07
Table 2. Total prompt neutron counts per second for the
INEEL detectors positioned at one meter from the DU
target and using the ARACOR “6-MeV” accelerator.
Detector Number with Error
Case
#0
#3
#2
#1
DU Only
6.2 ± 0.2
5.8 ± 0.2
4.6 ± 0.2
4.3 ± 0.2
DU + Be 465.7 ± 2.0 455.3 ± 2.0 404.7 ± 1.8 362.5 ± 1.7
DU+ D2O 572.8 ± 2.2 581 ± 2.2
544 ± 2.1
FIGURE 5. The Varitron’s electron beam energy
spectrum (normalized) used to match the delayed
neutron emission response of the ARACOR Eagle.
511 ± 2.1
949
SUMMARY
Table 3. Net delayed neutron counts per second for the
INEEL detectors positioned at one meter from the DU
target and using the INEEL Varitron.
Using the measured delayed neutron output for a
4.4-kg depleted uranium target at one meter from an
acelerator’s bremsstrahlung source, the operation of
the INEEL Varitron was successfully matched to the
operation of an ARACOR Eagle. Relying on the
high operational repeatibility of the Varitron and its
transportability,
a
detection
technology
demonstration at LANL is now possible using an
Eagle-like accelerator operation, a cargo container,
and special nuclear materials.
Detector Number with Error
Case
#0
#3
#2
#1
DU Only 1.16 ± 0.10 1.23 ± 0.11 1.10 ± 0.10 1.35 ± 0.11
DU + Be 0.30 ± 0.06 0.36 ± 0.06 0.29 ± 0.06 0.45 ± 0.07
DU+ D2O 0.62 ± 0.08 0.56 ± 0.07 0.69 ± 0.08 0.70 ± 0.08
When using the Eagle-matched operation, the
uncollimated Varitron was able to provide a
maximum net delayed neutron count rate between
1.16 to 1.35 cts/s per INEEL detector (See Table 3).
Total prompt neutron emissions for this same
scenario ranged from 5.2 to 7.1 counts per second
per detector, indicating that prompt counting, if
differentiated from the structural prompt responses
and used in conjunction with imaging, might also be
used to assist in identifying nuclear materials.
Additional attention will focus on assessing the
benefits of prompt neutron counting and higher
energy electron beam operations.
Table 4. Total prompt neutron counts per second for the
INEEL detectors positioned at one meter from the DU
target and using the INEEL Varitron.
Detector Number with Error
Case
#0
#3
#2
#1
DU Only
7.1 ± 0.2
6.2 ± 0.2
5.2 ± 0.2
5.8 ± 0.2
DU + Be 391.1 ± 1.8 303.0 ± 1.6 236.2 ± 1.4 206.4 ± 1.3
DU+ D2O 535.5 ± 2.1 432.0 ± 1.9 349.4 ± 1.7 327.7 ± 1.7
For the “DU Only” cases, note the similar delayed
and prompt neutron detection responses of the four
INEEL detectors with their corresponding responses
for the ARACOR data shown in Tables 1 and 2.
Note, only 24 R/min was produced at one meter with
an open-end bremsstrahlung converter indicating
some difference in the shape of the bremsstrahlung
output for these two machines. Similar trends are
seen when one compares the detection responses
with the addition of the D2O material. In this case,
one notices that the prompt response appears to
slightly decrease with increasing detector-toconverter distance. This is because the Varitron did
not use a massive shield forward of its converter (as
with the ARACOR accelerator). However, a match
is not observed when the lowest energy threshold
material (Be) is added. In this latter case, the
Varitron’s summed (all detectors) delayed neutron
results is nearly 2.7 times that of the ARACOR
accelerator, but shows correspondingly lower prompt
neutron detection responses. No satisfactory
explanation of this anomalous behavior has been
determined, but it must be at least related to the
photon/neutron scattering within the Be and the
ARACOR shield assembly and the correspondingly
lower Varitron photon dose output.
ACKNOWLEDGMENTS
The authors acknowledge ARACOR, Inc. for
providing the use of the La Honda, CA. test facility
and the Eagle-like accelerator, the Idaho Accelerator
Center for supporting technology development
research, LANL for the collaboration support, and
the Department of Energy (DOE) Office of
Nonproliferation Research and Engineering. The
work was performed under DOE Idaho Operations
Office Contract DE-AC07-99ID13727.
REFERENCES
1. Jones, J.L., et al., “Pulsed photoneutron Interrogation:
The GNT Demonstration System,” WINCO-1225,
INEEL Formal Report, October 1994.
2. Van Ausdeln, L.A., Haskell, K.J., and Jones, J.L., “A
Personal Computer-based Monitoring and Control
System for Electron Accelerators,” Proceedings of the
2001 Particle Accelerator Conference, June 18-22,
2001, Chicago, Illinois.
`
950