An Unattended Verification Station for UF6 Cylinders:
Implementation Concepts and Development Status
L.E. Smitha, K.A. Millerb, J. Garnerc, R. Polandd, B. McDonalda, H. Nordquistb,
J. March-Leubac
a
Pacific Northwest National Laboratory,
902 Battelle Boulevard,
PO Box 999,
Richland, WA, 99352, USA
b
Los Alamos National Laboratory,
Los Alamos, NM, USA
c
Oak Ridge National Laboratory,
Oak Ridge, TN, USA
d
Savannah River National Laboratory,
Savannah River, GA, USA
Abstract. In recent years, the International Atomic Energy Agency (IAEA) has pursued innovative
techniques and an integrated suite of safeguards measures to address the verification challenges posed
by advanced centrifuge technologies and the growth in separative work unit capacity at modern
centrifuge enrichment plants. These measures would include permanently installed, unattended
instruments capable of performing the routine and repetitive measurements previously performed by
inspectors. Among the unattended instruments currently being explored by the IAEA is an Unattended
Cylinder Verification Station (UCVS) that could provide independent verification of the declared
relative enrichment, 235U mass and total uranium mass of 100 percent of the declared cylinders in the
plant, as well as the application and verification of a “Non-destructive Assay Fingerprint” to preserve
verification knowledge on the contents of each cylinder throughout its life in the facility. As the
IAEA’s vision for UCVS has evolved, Pacific Northwest National Laboratory (PNNL) and Los
Alamos National Laboratory (LANL) have been developing and testing candidate nondestructive
assay (NDA) methods for inclusion in a UCVS. Modelling and multiple field campaigns have
indicated that these methods are capable of assaying relative cylinder enrichment with a precision
comparable to or perhaps better than today’s high-resolution handheld devices, without the need for
manual wall-thickness corrections. In addition, the methods interrogate the full volume of the cylinder,
thereby offering the IAEA a new capability to assay the absolute 235U mass in the cylinder and muchimproved sensitivity to substituted or removed material. Building on this prior work, a UCVS field
prototype is being developed and tested. This paper describes potential implementation concepts,
UCVS functions, a preliminary prototype design, and preparations for upcoming field trials.
1
Introduction
The International Atomic Energy Agency’s (IAEA’s) current enrichment-plant safeguards approaches
include attended weighing and nondestructive assay (NDA) of a subset of the plant’s cylinder flow and
inventory, collection of bulk uranium hexafluoride (UF6) samples for destructive analysis, and
environmental sampling for subsequent laboratory analysis. New safeguards measures that are more
effective and cost-efficient than contemporary measures are needed, particularly for modern highcapacity plants [1]. Detection of prominent diversion scenarios could be improved at enrichment
plants if the IAEA could monitor 100 percent of material flows and periodically calculate independent
uranium and 235U mass balances for the facility. However, human and financial resources preclude
continuous inspector presence at the facility to measure all of the material flow, using today’s attended
methods. Further, the portable measurement methods currently used by inspectors have relatively low
accuracy for the assay of relative 235U enrichment, especially for natural and depleted UF6, and no
capability to assay the absolute mass of 235U and total uranium in the cylinder, because of the highly
1
localized nature of the instrument geometry and low-energy gamma-ray signature. The poor accuracy
of today’s cylinder verification instruments necessitates additional safeguards measures, including the
destructive analysis of UF6 samples from select cylinders.
Unattended instruments capable of continuously monitoring material flows, and of performing the
routine and repetitive measurements previously performed by inspectors without additional burden to
operators, are central to the new safeguards approaches being considered by the IAEA [2][3]. One of
the instrumentation concepts being considered by the IAEA is an Unattended Cylinder Verification
Station (UCVS) [3][4]. UCVS units would be located at key intersections of cylinder movement
between material balance areas, or at the operator’s accountancy scales (in order to take advantage of
the facility’s cylinder weighing operations). The station would include technologies for cylinder
identification, NDA of the cylinder contents, video surveillance and data transmission to an on-site
computer or inspectorate headquarters. UCVS units would be owned and operated by the IAEA, but
the data streams could be shared with the operator (e.g., for process control) in conformance with
IAEA requirements for shared-use instruments. A notional UCVS is illustrated in Fig. 1.
FIG. 1. Conceptual design of an integrated UCVS that includes unattended NDA instrumentation
(blue panels), camera surveillance and cylinder identification technology.
According to the IAEA, the NDA components of the UCVS will support several measurement
objectives, including: unattended, independent assay of cylinder enrichment (Ecyl) and 235U mass
(M235) for product, feed, and tail cylinders; independent assay of total uranium mass (MU) as a
confidence-building measure on the authenticity of data from operator weighing systems; and the
unattended application, verification, and re-verification of an “NDA Fingerprint” to maintain the
verification pedigree of the cylinder contents and to verify that no partial removal of material has
occurred during the cylinder’s life at the facility [4][5].
The concept of an NDA Fingerprint with UCVS is more fully described in Ref. [3] and aims to
address the challenge of maintaining continuity of knowledge (CoK) on cylinders and their contents
during the cylinder’s lifetime at the facility. CoK is a particular challenge in enrichment plants since
the traditional CoK tool for nuclear material containers, metal or electronic seals, would require very
frequent inspector presence to either emplace or remove seals. The UCVS’s NDA Fingerprint concept
is not equivalent to a traditional seal, but has the potential to periodically confirm, in an unattended
fashion, that the contents of the cylinder are unchanged and therefore, that diversion has not occurred.
From the technical perspective, the NDA Fingerprint is a collection of signatures and attributes that
reflect the key verification parameters for the cylinder. There are a number of observable signatures
that could be used in the creation of the NDA Fingerprint attributes, and generally speaking, they are
the same signatures that are used for the direct assay of Ecyl and M235 (e.g., gamma-ray peak ratios or
neutron emissions). The constancy (or at least predictability, if decay half-lives are a factor) of these
signatures would allow a quantitative check that the key verification parameters for the cylinder are
unchanged between scans.
2
If the potential of the UCVS concept can be realized, such an instrument could significantly enhance
the IAEA’s efficiency in implementing safeguards approaches at large-capacity enrichment plants,
while simultaneously improving effectiveness for deterring and detecting diversion of material from
declared flow. A UCVS could also provide benefits to the operators, by easing and expediting the
release process for product cylinders, and cylinder tracking for process control [3]. Though the
potential of a UCVS system is understood, its field performance and operational viability in a
commercial enrichment facility has yet to be fully tested. Under the auspices of the United States and
European Commission Support Programs to the IAEA, a project has been undertaken to assess the
technical and practical viability of the UCVS concept. In this paper, we describe potential
implementation concepts, a preliminary UCVS prototype design and functional description, and
preparations for upcoming field trials.
2
Potential implementation concepts in enrichment facilities
Building from previous work by the authors and the IAEA, potential concepts of operation for UCVS
units deployed in enrichment facilities have been developed assuming three classes of material and
associated cylinders: 1) unblended product and tail cylinders, 2) blended product cylinders and 3) feed
cylinders. Only unblended product and feed cylinders are discussed here; nominal plant design and
processes are assumed for simplicity but in practice, each safeguarded facility would have its own
unique characteristics.
For unblended Type 30B product cylinders, UCVS tracking would begin as the empty cylinder is
transferred from the Storage Material Balance Area (MBA) to the Process MBA (see Fig. 2, steps 1
and 2). This initial scan would verify that the cylinder is indeed empty by industry standards (i.e.,
some heel material often remains in an empty cylinder). After the product cylinder is filled and
homogenized in the Process MBA, a UCVS scan is performed during the transfer back to the Storage
MBA (steps 3 and 4). In this scan of the full unblended product or tail cylinder, the UCVS would
measure, independently, Ecyl, M235, and MU, and store these data in a way that supports automated
comparison to operator declarations of those parameters. The NDA Fingerprint for each filled cylinder
would also be collected and archived during this scan. Product cylinders would remain in the Storage
MBA until the operator is ready to ship the cylinder off-site. As the cylinder is removed from the
Storage MBA for shipment, the UCVS would confirm the constancy of the NDA Fingerprint since
previous scans, perhaps with review and approval by a remotely located inspector (e.g., at IAEA
headquarters). This automated confirmation process could enable an expedited cylinder release
process for facility operators (steps 5 and 6), when compared to today’s approaches that involve
routine interim inspections and on-site inspector measurements.
FIG. 2. Conceptual overview of how an unblended product cylinder could be verified and released
from an enrichment facility using a UCVS.
The UCVS-based verification of feed cylinders could leverage the specific nature of natural feed
material. For example, assuming natural—not recycled—uranium feed material, the isotopic content is
known to high precision and accuracy, and the contents of the cylinder are expected to be highly
homogenized. It should be sufficient, therefore, to simply ensure that each new feed cylinder has the
attributes of “typical” natural-uranium feed material entering the facility. In practice, this means that
an attribute monitor like the NDA Fingerprint could provide all of the necessary quantitative data for
3
verification of the feed material entering the facility. Minor isotopic content, for example 234U or 232U
could also be verified using the NDA Fingerprint and compared to operator declarations based on
mass spectrometry. Periodic destructive analysis on cylinder samples taken by the IAEA could ensure
that an assumption of 0.711 percent enrichment (and minor isotope concentrations) for feed cylinders
remains valid.
3
Preliminary UCVS prototype design
A conceptual schematic of a UCVS field prototype is shown in Fig. 3. The radiation sensors and
associated pulse-processing electronics are located in the Collection Node enclosure. Other data
acquisition components would be located in a nearby cabinet, along with ancillary components that
include an industrial computer, uninterruptible power supply, power distribution, support for
surveillance cameras, and devices for secure communications. The central facility server (not shown)
would serve as the collection point for data from UCVS units and all other instrumentation at the
facility (e.g., cylinder identification, load cells, online enrichment monitors and additional
surveillance).
FIG. 3. Notional architecture for a field-prototype UCVS that includes a Collection Node (in this case,
co-located with the facility’s accountancy scale) and nearby cabinet that provides support for remote
monitoring by the IAEA. Blue indicates tamper-indicating measures, either enclosures or conduit.
In recent years, research and development (R&D) programs around the world have advanced the stateof-the-art in technology areas relevant to the UCVS concept; this project seeks to consolidate and
capitalize on that prior work. A summary of relevant prior work is provided here.
3.1
NDA methods
The IAEA has identified two candidate NDA methods for UCVS, both of which were developed under
support from the U.S. Department of Energy’s (DOE’s) Next Generational Safeguards Initiative: the
Hybrid Enrichment Verification Array (HEVA) being developed by Pacific Northwest National
Laboratory (PNNL), and the Passive Neutron Enrichment Meter (PNEM) being developed by Los
Alamos National Laboratory (LANL).
PNNL’s HEVA is a hybrid cylinder assay technique that utilizes an array of NaI(Tl) spectrometers to
simultaneously measure the direct 186-keV signature from 235U and via high-energy gamma rays
induced by neutrons in 56Fe and the NaI(Tl) itself, the total neutron emission rate from the cylinder
[7][8]. The 186-keV signature provides an unambiguous measure of Ecyl. Under assumptions of known
234
U/235U behaviour in the plant, the total neutron signal can be calibrated to total M235 in the cylinder.
Proof-of-principle field prototypes have consisted of an array of NaI(Tl) spectrometers surrounded by
special collimators composed of iron and polyethylene that also serve to enhance the high-energy
gamma-ray signal. The current HEVA design consists of multiple 7.5-cm x 7.5-cm NaI(Tl)
spectrometers coupled to Canberra Osprey digital photomultiplier tube bases (Fig. 4).
4
FIG. 4. Left: HEVA detector module to be used in the UCVS prototype. Right: cutaway showing the
NaI(Tl) crystal (green) and surrounding collimator consisting of neutron-moderating (polyethylene)
and (n,γ) conversion material (56Fe).
LANL’s PNEM employs polyethylene-moderated 3He neutron detectors to measure the singles and
doubles neutron count rates from the cylinder [9][10]. The singles counts come primarily from the 234U
and under an assumption of known 234U/235U behaviour, and allow determination of 235U mass, a
method utilized by the Uranium Cylinder Assay System deployed by the operator at a Japanese
enrichment plant [11]. PNEM extends beyond singles neutron counting to utilize the coincidence (i.e.,
doubles) neutron signature that arises from induced fission in 235U. The PNEM hardware consists of
two polyethylene-moderated detector pods, each containing 12 3He tubes at a pressure of 10 atm (Fig.
5). Data acquisition and analysis are based on pulse processing electronics from Precision Data
Technology, a Canberra JSR-12 shift register and the IAEA Neutron Coincidence Counting software.
FIG. 5. Left: PNEM detector pods, each containing 12 3He tubes. Right: PNEM detector pods
deployed for 30B cylinder measurements.
Previous, separate field measurement campaigns by PNNL and LANL provided early insight into the
ability of the HEVA and PNEM methods to assay Ecyl and M235 [8][10] but no side-by-side testing of
the methods to provide comparative findings. In 2013, under a bilateral agreement between the U.S.
Department of Energy and the European Atomic Energy Community (Euratom), the HEVA and
PNEM methods were tested concurrently on both Type 30B and Type 48 cylinders, in an operating
enrichment plant [12][13]. This study at URENCO’s enrichment plant in Almelo, Netherlands
provided important findings regarding the suitability of the methods for the assay of for Ecyl and M235
(but not MU). The results of the study indicate that the HEVA and PNEM unattended cylinder NDA
methods are capable of assaying relative enrichment with a precision comparable to today’s highresolution handheld devices and meeting the International Target Values [14], without inspector
presence. In addition, both methods are capable of interrogating the full volume of the cylinder,
enabling the verification of 235U mass. It was also preliminarily demonstrated that by hybridizing two
different and complementary radiation signatures, the HEVA and PNEM methods are capable of
detecting 234U/235U ratios outside the calibrated (i.e., typical) range and therefore, to indicate the
presence of UF6 material with non-natural origins [12][13].
5
3.2
Cylinder identification and surveillance
R&D programs in IAEA Member States and Member State Support Program projects have been
pursuing cylinder identification technologies for a number of years. Among the candidate technologies
are image-based optical character recognition, image-based feature matching, active and passive
radiofrequency identification, and laser-based systems [15][16]. While the UCVS field prototype
developed in this project could ultimately provide a platform for the testing and evaluation of mature,
field-ready cylinder identification technologies, it is outside the current scope to integrate and test
those technologies. For the current project, the UCVS prototype will optically capture images of the
valve end of each cylinder, including the nameplate, for the purposes of cylinder identification. The
camera system will be the IAEA’s Next Generation Surveillance System (NGSS), which is now being
deployed globally [17]. The use of the NGSS in a UCVS field prototype will provide the IAEA with
additional experience in the use of coupled radiation-surveillance systems and information about the
potential roles of surveillance in the UCVS concept.
3.3
Software
Several relevant software development activities are underway at the IAEA, and the UCVS project
will leverage and align with these activities as appropriate. For example, the IAEA has developed a set
of requirements for implementing data security and remote transmission in unattended systems called
RAINSTORM (Real-time and Integrated Stream-Oriented Remote Monitoring) [18]. The IAEA has
provided a sample implementation of software meeting these requirements and is collaborating with
the UCVS team to implement and test that software in the field prototype. In addition, the IAEA is
collaborating with Euratom to develop a unified data analysis software package, called iRAP, for the
analysis and review of unattended instruments. The basis for this new software is Euratom’s Central
RADAR Inspection Support Package (CRISP) [19]. The IAEA and Euratom are collaborating with the
UCVS team to integrate and evaluate iRAP. Finally, the IAEA’s development of the On-Line
Enrichment Monitor (OLEM), under the auspices of the United States Support Program and led by
Oak Ridge National Laboratory, has produced a software architecture for an unattended monitoring
system with modules for data acquisition, data analysis, data security and remote reporting—code and
experience that is highly relevant to the UCVS project [20].
A guiding principle of the UCVS software architecture will be the need for maintainability and
sustainability by the IAEA. The UCVS software will be modular (Fig. 6), written in C# for a Windows
operating system, utilize common standardized data formats (e.g., the ANSI N42.42 file format), and
will use standard open-source or commercial libraries to the extent possible.
4
Field trials
Previous field campaigns with HEVA and PNEM were based on proof-of-principle prototypes
operating in an attended mode and relatively small numbers of cylinders. Further study of the NDA
signatures and methods, longer field trials and large, diverse cylinder populations, along with refined
versions of HEVA and PNEM hardware and software will enable a more definitive evaluation of NDA
capabilities relevant to unattended cylinder verification. An integrated prototype that has been
fabricated based on IAEA’s preliminary user requirements will allow evaluation of operational factors,
maintainability and estimated life cycle cost.
The UCVS field prototype will be initially deployed at a Westinghouse fuel fabrication facility in
South Carolina, USA in early 2015. At that facility, Type 30B cylinders with enrichments ranging
from natural to approximately 5 percent will be measured over the span of several months and
tentatively, several hundred cylinders. The UCVS team will use the data and findings to inform the
continued refinement of the analysis algorithms, the integrated software package, and the UCVS
implementation concept. Depending on the results of the initial field trial, the UCVS prototype could
be deployed at an operational enrichment plant where ideally, the field testing would include the
measurement of hundreds of Type 30B and 48Y cylinders spanning a full range of enrichment and
including atypical cylinders that are likely to challenge the NDA methods and aid in the continued
evolution of implementation concepts, algorithms, and reporting practices.
6
FIG. 6. Conceptual schematic of a modular UCVS software architecture design that includes data
acquisition modules for PNEM, HEVA, and the NGSS; occupancy detection; and facilitates “separate
but parallel” development, and testing of candidate analysis algorithms. Data security and remote
transmission software consistent with IAEA’s remote monitoring requirements will be used to support
off-site data transmission (e.g., raw data, state-of-health [SOH]data).
5
Next steps
The next steps in the UCVS project include continuing preparation for the field trial with the facility
operator, refinement and testing of the NDA hardware and data acquisition software, software
integration and testing, design and fabrication of a portable mechanical support system for cylinders
and the UCVS components, and testing of the integrated prototype prior to deployment at the
Westinghouse facility. To complement the field trials and address one of the UCVS performance
targets, PNNL and LANL will perform a modelling-based study of the HEVA and PNEM sensitivity
to diversion scenarios in which a portion of the enriched material in the interior of a cylinder has been
replaced with depleted or enriched material. This effort will build on prior work by the two
organizations [10][21] but utilize a common set of diversion scenarios and metrics. The findings of the
UCVS project will inform an IAEA decision regarding the viability of the instrument in terms of
performance, operator acceptability and lifecycle cost.
6
Acknowledgments
Funding for this work has been provided by the U.S. National Nuclear Security Administration’s
(NNSA) Office of Nonproliferation and International Security (NA-24) and the Next Generation
Safeguards Initiative, and the U.S. Support Program (USSP) to the IAEA. The authors are appreciative
to Chris Orton of the NNSA and Joseph Carbonaro of Brookhaven National Laboratory and the USSP
for their support and guidance of this project. The authors would also like to thank James Ely of the
IAEA for his oversight of the UCVS development, and Peter Schwalbach and James Morrissey of
Euratom safeguards for their advice and collaboration.
REFERENCES
[1]
[2]
[3]
COOLEY, J.N., Model Safeguards Approach and Innovative Techniques Implemented by the
IAEA at Gas Centrifuge Enrichment Plants. 2007 INMM Annual Meeting Proceedings.
LEBRUN, A.R. et al. Improved Verification Methods for Safeguards Verifications at
Enrichment Plants. 2009. Proceedings of the ANIMMA Conference.
SMITH, L.E., A.R. LEBRUN, and R. LABELLA. Potential Roles of Unattended Safeguards
Instrumentation in Centrifuge Enrichment Plants. Journal of Nuclear Materials Management
42(2013).
7
[4]
[5]
[6]
[7]
[8]
[9]
[10]
[11]
[12]
[13]
[14]
[15]
[16]
[17]
[18]
[19]
[20]
[21]
8
Unattended Cylinder Verification Station, SP-1 Task Proposal. IAEA Department of
Safeguards. 2013.
Preliminary User Requirements for an Unattended Cylinder Verification Station. IAEA
Department of Safeguards. 2013.
SMITH, L.E., et al. Viability Findings for Unattended Verification of UF6 Cylinders using the
Hybrid Enrichment Verification Array. Submitted to Journal of Nuclear Materials Management.
2014.
SMITH, L.E., E.K. MACE, and A.C. MISNER. Signatures and Methods for the Automated
Nondestructive Assay of UF6 Cylinders at Uranium Enrichment Plants. IEEE Transactions on
Nuclear Science 57(2010).
JORDAN, D., L.E. SMITH, J. KULISEK, B. MCDONALD, and E. MACE. Hybrid Enrichment
Assay Methods for a UF6 Cylinder Verification Station: FY13 Final Report. 2013. PNNL23164, Pacific Northwest National Laboratory, Richland, Washington.
MENLOVE, H.O., M.T. SWINHOE, and K.A. MILLER. A More Accurate and Penetrating
Method to Measure the Enrichment and Mass of UF6 in Storage Cylinders Using Passive
Neutron Self-Interrogation. 2010. INMM Annual Meeting Proceedings.
MILLER K.A., H.O. MENLOVE, M.T. SWINHOE, and J.B. MARLOW. Technical
Assessment of the Passive Neutron Enrichment Meter (PNEM) for an Unattended Cylinder
Verification Station (UCVS). 2013. LA UR 13-22911, Los Alamos National Laboratory.
MILLER, K.A., et al. The Uranium Cylinder Assay System for Enrichment Plant Safeguards.
Journal of Nuclear Materials Management 39(2010).
SMITH, L.E., K.A. MILLER, E.K. MACE, D.V. JORDAN, C.R. ORTON, C.D. RAEL, J.
MORRISSEY, and R. VELDHOF. A Study of Candidate NDA Methods for Unattended UF6
Cylinder Verification: Action Sheet 40 Final Report. 2014. PNNL-22954, Pacific Northwest
National Laboratory.
MILLER K.A., L.E. SMITH, C.D. RAEL, E.K. MACE, C.R. ORTON, P. SCHWALBACH, J.
MORRISSEY, P. DE BAERE, T. VISSER, and R. VELDHOF. Joint Field Trial Considerations
for UF6 Cylinder Assay Technologies at an Enrichment Facility. 2014. INMM Annual Meeting
Proceedings.
ZHAO, K., et al. International Target Values 2010 for Measurement Uncertainties in
Safeguarding Nuclear Materials. 2010. IAEA STR-368, IAEA Department of Safeguards.
YAO, J., et al., Automatic laser-based identification for UF6 cylinders. 2010. Machine Vision
and Applications.
DURBIN K., E. WONDER, and M. WHITAKER. NGSI’s UF6 Cylinder Monitoring Project
Update: Technology Survey and Evaluation. 2014. INMM Annual Meeting Proceedings.
MOESLINGER, M., et al. The IAEA’s XCAM Next Generation Surveillance System. 2010.
Proceedings of the Symposium on International Safeguards, Preparing for Future Verification
Challenges.
MORGAN, K., C. BRUNHUBER, J. REGULA, H. NORDQUIST, V. POPESCU, S.
RECHBERG, and S., I. NAUMANN. Real-time And INtegrated STream-Oriented Remote
Monitoring (RAINSTORM) Interface. 2012. INMM Annual Meeting Proceedings.
SCHWALBACH P., A. SMEJKAL, E. ROESGEN, and T. GIRARD. RADAR and CRISP ‐
Standard Tools of the European Commission for remote and unattended data acquisition and
analysis for nuclear safeguards. 2006. Proceedings IAEA Symposium on International
Safeguard.
YOUNKIN J.R., J.R. GARNER, N.C. ROWE, and L.E. SMITH. A Robust and Flexible Design
for Gas Centrifuge Enrichment Plant Unattended Online Enrichment Monitoring. 2012. 9th
International Conference on Facility Operations-Safeguards Interface.
KULISEK J.A., D.V. JORDAN, E.K. MACE, B.S. MCDONALD, and L.E. SMITH. Hybrid
Enrichment Verification Array: Investigations of the High-Energy Gamma-Ray Signature
Origin and Use for Partial Defect Detection. 2014. INMM Annual Meeting Proceedings.