MEETING WITH IAEA TO DEMO URANIUM DATABASE AND

S E C R E T UNVIE VIENNA 000571

NOFORN

C O R R E C T E D C O P Y (CLASSIFICATION)

SIPDIS

ISN/RA MONGIELLO; NA-241 O'CONNOR; LLNL HUTCHEON; AFTAC FOR
CHARLES BRENNAN

E.O. 12958: DECL: 12/17/2034
TAGS: AORC KNNP IAEA ENRG TRGY
SUBJECT: MEETING WITH IAEA TO DEMO URANIUM DATABASE AND
IDAVE SEARCH ENGINE

Classified By: AMBASSADOR GLYN DAVIES FOR REASONS 1.4 (B) AND (D)


1. (S) Summary: At a December 9 meeting, IAEA Deputy
Director General for Safeguards (DDG) Heinonen reaffirmed the
IAEA's long-term willingness to participate with the U.S. in
a joint effort to maintain a database of global signatures of
uranium yellowcake (and other forms of uranium).
Representatives from the Departments of State and Energy and
Lawrence Livermore National Lab (LLNL),accompanied by
MsnOffs, met with Heinonen to demonstrate the uranium
sourcing database and accompanying internet-based search
engine, the Discriminate Analysis Verification Engine
(iDAVE). The demo, presented by LLNL's Ian Hutcheon and
Martin Robel, consisted of a walk-through of how the iDAVE
database works followed by a discussion of future cooperation
between the IAEA and LLNL on sample analysis and data
interpretation. Responding to Heinonen's positive reaction
to the presentation, DepCounselor encouraged the IAEA to
maximize the number of samples taken during safeguards
inspections (especially in locales such as Iran, Syria, and
other states in which the U.S. would not otherwise have data)
and indicated a U.S. willingness to fund analysis of as many
samples as the IAEA is able to collect. Next steps in the
joint project will be a return visit by LLNL experts in
January/February 2010 timeframe to establish IAEA
connectivity so that select safeguards inspectors will be
able to query iDave directly. Finally, Heinonen reaffirmed
his previous request to treat with discretion the knowledge
of the close U.S.-IAEA collaboration on this database, given
that the Agency is providing the USG with information on the
origins of samples that it would not share with other member
states. End Summary

--------------
iDAVE - On the Brink of Going Live
--------------


2. (S) During the December 9 meeting with Olli Heinonen
and Maxim Penkin, LLNL's Ian Hutcheon opened by noting that,
of the approximately 130 samples the IAEA has sent to LLNL
since January 2007, roughly two-thirds have been analyzed and
the results returned to the Agency. In return, the IAEA has
provided source location information for roughly half of the
130 samples. The internet-based search engine which operates
on the uranium sourcing database, iDAVE, should be ready to
'go live' within four-to-eight weeks, or before the end of
February 2010, Hutcheon said. By that time the IAEA will be
sent one-time password tokens from LLNL to be able to access
the database system from their IAEA computers. LLNL experts
explained that the server had not yet been physically
transferred to its new location at LLNL or connected to
LLNL's "Blue" network for secure, unclassified communication.
LLNL also needs to complete the installation of new
web-based protocols to allow VPN access to the Blue network.
These two actions are expected to take place before the end
of January 2010.

--------------
TECHNICAL DISCUSSION OF iDAVE
--------------


3. (S) During the meeting with DDG Heinonen, LLNL gave a
live presentation of iDAVE, using an internet connection to
the iDAVE server, to demonstrate how one would use the search
engine to derive sourcing information about uranium ore
concentrate, or other uranium bearing, samples. During the
presentation, there was discussion of some of the technical
aspects of the database. For example, one issue concerned
the potential difference between the point of origin and
point of collection of samples that the IAEA sends to LLNL
for analysis. Inspectors cannot always know that samples
found in a particular facility originated in that location.
Nonetheless, the IAEA reaffirmed that it tries to send as
much information as possible about the origin of the samples
to LLNL. In (rare) cases where the IAEA simply does not have
enough information to determine the point of origin or other
useful information about a sample, it will not send that
sample to LLNL for analysis.


4. (S) The discussion also covered the basic operating
principles of iDAVE, which currently uses 16 parameters
common to all samples in the database. While this is a
relatively small number (the IAEA expects there to be at
least 60 parameters for samples measured since 2007),it was
noted that the number of parameters will grow as additional
analyses of samples already in the database are completed by

the participating DOE laboratories and the number of samples
in the database gets larger. (Note: There are approximately
2,000 samples from other sources in addition to those
provided by the IAEA currently in the database. End Note.)
The technical discussion also addressed the various types of
information that iDAVE will be capable of providing, i.e.,
not only country or facility of origin but also descriptive
statistics relating to the certainty with which the
assignment of a source location is made.


5. (S) During the presentation, the IAEA raised a
question concerning the ability of iDAVE to return
information concerning the number of samples in the database
with a specific set of characteristics in addition to the
normal query based on specific measurements of a sample to
find point of origin matches. LLNL replied that while iDAVE
does not currently provide this information, it would be
relatively simple to add the option to generate this
information in the next version of iDAVE. It is likely that
the IAEA will be using the latter type of query once access
is granted.

-------------- --
POTENTIAL COOPERATION WITH AUSTRALIA AND CANADA
-------------- --


6. (S) During the discussion, Senior Safeguards Analyst
Maxim Penkin noted that there is potential for cooperation
with Canada and Australia to study how trace elements behave
during the process of converting uranium ore to UF6. He
explained that the amount of data the IAEA expects to gain
from Australia and Canada could be quite large and should
therefore be entered into the database. LLNL and Msnoffs
agreed that this would be a very good idea. (Comment:
LLNL/DOE has a current project with the Australian National
Safeguards Office to provide similar information for U-ore of
Australian origin.)

--------------
FOLLOW-UP TECHNICAL EXPERTS DISCUSSION
--------------


7. (S) Following the initial briefing to DDG Heinonen, a
technical experts discussion took place with Penkin, IAEA
Safeguards Analyst Ke Zhao, Hutcheon, Robel, MsnOff and
Washington rep to review the details of iDAVE. LLNL walked
through the various steps to access the iDAVE search engine
and answered the IAEA's questions. One issue that
continually arose concerned the IAEA's access to the raw data
in the database. LLNL reiterated that at this juncture, the
strong U.S. preference is to maintain one copy of the actual
data at LLNL to ensure additions to the data are uniform.
Further technical questions concerned how samples are grouped
together when an analysis is performed by iDAVE, and whether
or not multiple samples can be analyzed at one time. LLNL
attempted to log in to the database using an IAEA computer
(the previous demonstration was performed using an LLNL
laptop computer). However, there were technical
difficulties, possibly related to a firewall at the IAEA,
which will have to be resolved at LLNL. Ian Hutcheon and/or
Martin Robel agreed to return to Vienna before the end of
February to provide an additional tutorial once the web-based
version of iDAVE is fully operational and access is provided
to the IAEA. The meeting concluded with a discussion of next
steps.

--------------
NEXT STEPS
--------------


8. (S) At the conclusion of the meeting with Heinonen,
the issue of maximizing IAEA sample-taking was discussed in
light of Heinonen's strong support for continued
collaboration on the database. DepCounselor reiterated that
samples of uranium from states of critical safeguards
interest are of particularly high value, given that the U.S.
does not generally have other means for collecting such data
from those states. DepCounselor urged the IAEA to take as
many such samples as possible and indicated U.S. willingness
to undertake and fund the requisite analysis. Heinonen noted
that the IAEA is beginning to receive an increasing number of
samples from Kazakhstan and Uzbekistan, which could help
populate the database and provide useful information. Maxim
Penkin further noted that the majority of the IAEA's current
samples are from Iran. (Comment: All parties recognized
that samples from critical countries (i.e. Iran, DPRK, Syria)

are of major benefit to the value of the database and further
collections will continue to be a vital part of this project.
End comment.)
DAVIES