LAUR-05-376G
Ancroved ox put rolease.
datetime anketed
1
Twe-) Enrichment Separative Gapacity for SILEX
Authors}
John L. Lyman
‘Submited te"! sp.1 POTAS Project, "Enrichment Separative Capacity"
L
» Los Alamos
NATICNAL LABORATORY
2 Slamas Naver Laboraery anathmatws acto-vecqual opperinty employer. cporated byt Ureareny ot Cabomnia otha).
Bpanenent Energy under convac #-7405 ENG- 36 By acceptance of hs anil. the puberer rosogeizns that the U.S, Government
reams anonexctsive.reysty-ree teense te puoieh Or eprcauce Me published form of ms coranbuvon. orto low Oro" 8050.1 LS.
Gavemnmentautposes ies Alamns Natonalt abora-oy comets thatthe nub neney hs article as work parnemad under ths
swapces d tne US Departentet Energy Lor Alamias Nationa Laboratory sienghy supports academic oadomandarsecarcho? ttt
‘BiDISh: aga mBtlon. However the Laboratory 4CBs not endorse he vewpoet oa puOMCEUON oF Quaranio8 fs lechncal comeciness.
Form as61s00}Enrichment Separative Capacity for SILEX
John L. Lyman
Los Alamos National Laboratory
Intraduction
The objective of this study is to provide guidetines for estimating the isotope enrichment
separative capacity for the technology known as SILEX.
Separation of Isotopes by Laser EXcitation (SILEX) is the name given to the technology
developed by Australian scientists, Michael Goldsworthy and Horst Struve. They formed
4 public company, Silex Systems Ltd. (SSL), to further develop and market the isotope
enrichment capability for uranium and several stable isotopes. Their current facilities are
in Australia at Lucas Heights. New South Wales, The facilities are in a secure area ina
building sented from the Australian Nuclear Science and Technology Organization
(ANSTO). ANSTO provides security and other services for the SSL. Michaet
Goldsworthy is the CEO of 8SL and Horst Struve is the Chief Sciemtist.
From 1996 to 2002 the United States Ensichment Corporation (USEC) provided funding
and technical assistance to SSL for their uranium enrichment process. [ was part of that
technical assistance from 2000 to 2002. My understanding of the process allows me to
make the guidelines presented here, USEC no longer has an interest in the SILEX
sechnology. All contractual arrangements between SSL and USEC have been terminated.
(understand, however, that SSL is pursuing agreements with other US companies. The
SSL objective is to build a uranium enrichment pilot plant in the United States. [alsa
understand thar SSL has no interest in marketing the technology outside of the United
States. | know al'n0 organization outside of the Linited States that has an interest in
developing the SILEX technology. Even Australia cannot build a pilot plant or
enrichment facility
During the time of the USEC-SSL cooperative research the United States and Australian,
governmems concluded an “Agreement for Cooperation '" that allowed joint research on.
uranium isotope separation to take place. It also provided a basis for the classification of
the work by both dhe Australian govemment and the United States Department of Energy.
Many aspects of the uraniurr. enrichment technology are also company proprietary. As
many aspects uf the SILEX process are eovered by Australian and American
classification end the proprietary restrictions. many details of the SILEX process cannot
by revealed in this report
in addition tw the time J worked for USEC with SSL in Australia, 1 also wrote the review
ol the SILEX process for the International Atomic Energy Agency (IAEA) in February
20S", The report discusses a visit af three [AEA representatives and myself to SSL to
assess the pucential for the company (SSL) to produce significant quantities of highly
enriched uranium (HEU 1. That report is classified “Safeguards-in-Confidence” by the
{AFA and is not available for distribution beyond the Agency. The report did conclude
thet SSL. is not currently capable of producing significant amounts HEU. | use some the
same techniques as used for the report in this document to establish the guidelinesDescription of the SILEX methodology
SSL has several corporate objectives. The one that is of interest to us is to develop a
method for uranium enrichment for nuclear fuel for electric power generation.
Because the process involves proprietary information and is classified by the two
governments, this description will be incomplete. The description in the Physical Mode)?
is also incomplete. It contains only one short paragraph that says that little is known
about the process. Another document’ is aiso available that describes the process, but is
classified and has restricted distribution,
The process uses a mixiure of UF, with a carrier gas, The gas mixture is cooled to
Separate the resonance peaks for the two isotopes CU and 4), A 16-jem laser then
selectively excites the “UF. One or more infrared laser frequencies may be used. The
laser excitation results in a separation into a product stream and a tails stream. The
product stream is enriched in 7“U while the tails strcam has an increased fraction of '4U.
Ina plant situation the «wo streams would be further processed to achieve the desired
enrichment and physical form. th the research laboratory in the current SSL facility the
‘wo streams are combined, remixed with the carrier gas, and retumed to the feed tank.
The main SSL experimental facility is the Direct Measurement (DM) laboratory. It
consists of:
1, A large process vessel where the separation occurs.
2. A feed vessel that contains the UF, and carrier gas mixture,
3. Vessels for containing the product and tails,
4. A quadrupole mass spectrometer for measuring the isotope ratios in the feed,
product and tails streams.
‘The necessary plumbing and processing facilities for carrying out the
experimental tests. and
6. All necessary optics for bringing the laser radiation to the process vessel.
Pulsed laser radiation at a specific! 6-cam wavelength is necessary 10 achieve the isotope
enrichment. The lasers are in an adjacent room and the laser radiation is brought through
2 port in the wall to the process vessel. The laser radiation for the process begins with one
of several pulsed COs lasers. These are capable of producing iaser pulses at a 300 Hz
repetition rate, The lasers are capable of producing one joule pulses, but not at high
repetition rate. The fact that the lasers must be at a specific frequency further reduces the
energy per pulse. Tae faser pulses in the 10.8-ym region are taken to a Raman conversion
cell where they are converted to the 16-y4m region. The Raman cell is a large vessel filled
with high-pressure para-hydrogen. The 10.8-ym laser radiation is passed through the
Raman celt many times (~25), and on each pass it goes through a focus near the center of
the cell. The result 1s a nontinear optical process that reduces the laser photon energy by 4
quantum of para-hydragen rotational energy to produce the desired 16-am laser
radiation. To obtain the exact laser frequency it is necessary ta use some additionalnonlinear optical tricks on the initial CO» laser pulse and to operate the laser at high
pressure.
Guidelines for assessing separative capacity
Exed and product materials
‘The feed for a hypothetical isotope enrichment plant based on the SILEX technology is
uranium hexatluoride (UF,). [0 is really the onty possible uranium feed material. All other
uranium-bearing species have insufficient volatility 10 be used in the process, and the
process must occur in the gas phase.
The amaunt of LF, required to produce a significant amount of HEU is substantial. As an
example, aysume the plant produces 10 ky of 85% U-235 and that process is sufficiently
selective to take natural UF. from 0.721% 10 0.221%" “UF. in the tails stream. This
requires 2500 kg of natural UF, feed,
A camer gas to dilute the UF, is necessary for the process 10 enrich uranium, The amount
of the carrier gas is substantially greater than the UF,, during the isotope separation
process. It. however. is not consumed and can be separated and reused. The identity of
the carrier gas and the dilution fraction are protected by the classification and proprietary
restrictions
“The nature of the product stream is again covered by the proprietary restrictions, but we
do know that the processing takes considerable time, at least under the current
experimental conditions. SSL has stated that an experimental run in their current facitity
Tequires ten hours for preparation, about an hour tor the run to occur. and another ten
hours to get ev erything ready for the next run, This preparation and processing time is 2
significant restriction on the amount of entiched material produced.
Material pro q
‘The current SSL research facility uses infrared lasers that operate at 50 Hz. The nature of
che enrichment process is that a considerably higher repletion rate would be necessary to
process a large fraction of the feed material. The $0 Hz laser system only allows a 1%
duty factor, That is, 99° of the feed material is unprocessed. This resuhs in a high
fraction of the feed material getting into the product stream. Consequently, the observed
enrichments are low. A working enrichment plant would require large improvements in
the laser repetition rate
Equipment signatures
The lasers themselves are a significant indication of the maturity of the process and the
capability to produce significant enrichment, We noted that a 50 Hz repetition rate is not
sulficiem, A mature facility would have the capability 1o increase that rate by more than
an order of magnitude, The pulse energy of the COs laser also needs ta be in the 1 J
Tange. It 1s not sust that the rate needs to be high for high production. At low rates the
Process does not work because of large amounts of unprocessed material in the product
ream,
The Raman converters are essential and at the projected high repletion rates considerable
actention needs to be paid (o the problem of energy deposition in the para-hydrogen
convening gas. This heating can produce oprical interference that obscures or deflectslater Jaser pulses. The Raman converters must be well engineered with facilities w flow
she hydrogen gas.
Restrictive povential of irradiated volume
‘One feature of the physical arrangement of the Laser processing provides a limit on the
amount of material that can be processed. tis the volume irradiated per pulse. It provides
a better restriction on. the enrichment capacity than the electrical power consumption or
the laser output power. Although we can give no indication of the process of enrichment,
we can conclude that the process is similar to other MLIS enrichment technologies. That
is. it requires Vapor phase irradiation at low temperature with the infrared lasers. We
know fram those processes that rapid, catastrophic condensation occurs if the UF.
molecular density gets to high. [know from experience with many molecular laser
enrichment programs that the motecuiai density cannot exceed 1x10"* molecutes/em’,
Given that density. irradiated volume. the laser repletion rate, and the natural abundance
of "4. we can calculate the amount of time to process some quantity, say 1.0 kg of *°U.
For a 1.0 / irradiated volume and a natural abundance of 0.721% "Li we find that it takes
12 eight-hour days to process the 1.0 4g of ™U. This calculation ignores that fact that the
laser proves is not perfect and multiple stages witl probably be required. It does,
however. give a generous upper limit on the enrichment capacity. One can estimate the
inodiated volume from the physical length of the process vessel and the diameter of the
input beam,
Liilities requirements for production
For a plant that is using the SILEX technology’ to enrich uranium the main utility
requirement 1s electrical power for operating the lasers. Significant electrical power
would also be necessary for maintaining vacuum conditions in the process vessel, for
pumping the process gas, and separating the camponents.
We can make an assessment af the electrical power required far the operation of the
lasers. The (Oz lasers themselves are about 1% efficient. The conversion to 16-zan laser
radiation is about 25% efficient. The laser power required for tests in the current SSL
experimental facility is 12 W, This would need 1o be enhanced by at least a factor of
txenty to obtain high enrichment. By considering all of these factors we obtain an
electrical requirement of about [60 kW for the lasers for a single process vessel. This
would be the minimum power requirement for processing the } 0 kg of “Lin eight days
mentioned in the previous paragraph.
Potential separative capacity from number and type of units
‘The previous paragraphs irradiation volume and electrical power requirements for a
single process vessel indicates what a single unit is capable of achieving in the
hypothetical SILEX enrichment facility. Because of difficulties with laser transpory, it
\ould be difficult to have larger separation units the test facility curently being used by
SSL in Australia. The value given above of 12 days to process 1.0 kg of *""U is an upper
limit. As we can say nothing about process efficiency or enrichment factor. we just make
the note that any inefficiency will increase the amount of time required to process 1.0 kg
ofEvaiuation of workforce
We address here the question of what the characteristics of the workforce tells us about
whether a particular facility is devoied to R & D or isotope production. The necessary
expertise for running a production facility has some common features with the expertise
required for R & D. Both will require someone with the capability to operate complex
laser systems. Since a production facility will not require the development of methods of
doing such things as changing the laser frequency, a complete knowledge of laser physics
and quantum aptics would be unnecessary. Good technicians with the capability 10
operate (and repair) existing laser facilities would suffice. Because the production facility
would run for long, periods of time and because the processes need to be continuously
monitored, the production facility would require several people with this capability.
Personnel with the skills of optical tecknicians would be necessary in a production
facility. These are people who not only could aperate and repair lasers, but could
facilitate the transport and control of the laser radiation,
Likewise. it would be unnecessary for the production facility to have on the staff people
who understand the physical processes that occur during and after the interaction of the
laser radiation with the feed gas. These processes should have been investigated and
understood during the R & D phase of the project.
One production facility requirement would be workers that would handle the product and
tails streams. Someone would need to be able to make whatever chemical and physical
changes are necessary to the product stream and to prepare the matenal for additional
stages of processing. Again. this does not require a high level of scientific education and
expertise. It does require sufficient technical training to effectively carry out the
necessary aperations.
Juideline
yn personnel for processes that ond R&D
Information in open sources should be avaitable for determining the education and skill
level of personnel in different countries that could be used in a SILEX type isotope
production facility. Some the people mast likely to be involved would include optical
techmiczans who could operate. repair, and contro] infrared lasers and the faser radiation
they produce. The optical elements such as salt windows, infrared reflecting mirrors, and
focusing opnes require specially trained personnel. The handling of uranium hexafluoride
requires some special skills that few people have. {1 is a volatile solid that has its own
chemical and radiation hazards. People trained in radiation and chemical hazard safety
need to be part of any production facility.
Recommendatsons for inspectors
The first recommendation is that inspectors take environmental swipes to be analyzed
later for the presence of uranium. particularly for uranium fluorides such as UO:Fo. This
material is produced by reaction of UE, with ambient water. The uranium isotope ratios
should be measured. If the isotope ratia is not natural it indicates one of two things.
Bither the facility couid be producing enriched uranium or the researchers had obtained
ennched uranium tor R & D purposes. As the natural fraction of “U is so low, it is
eremely difficult to develop an isotope separation process with only natural uranium.
Enriched uranium is almost a necessity in R & D.Radiation detectors that could detect the presence of uranium are essential unless the
facility is known to have uranium present.
I the SILEX process is being used, lasers will be present. They are noisy. A high
repetition rats lascr issues a loud hum. They are casy to hear when they arc operating.
Another indication of the presence of fascrs is laser beam transport facilities. These cari
be in tubes or through open ports in room walls.
Raman converters arc fairly large. They require in excess of ameter length with a
diameter near a balf-meter. They will also have optical ports and controls. The Raman
converters also require access to high-pressure hydrogen with the capability to cool them
with something like liquid nitrogen.
Laser goggles for eye protection will be present in the laser room and where the laser
radiation is present.
The process vessel itself requires a great deal of plumbing with pipes and tubes for
transport of the feed gas mixtures. They will also require large tanks for storage and
mixing of the foed gas. The product and tails stream must also be handled similaty.References
L
William J. Clinton, “Agreement for Cooperstion between the Linited States of
America and Australia Concerning Technology for the Separation af Isotopes of
Uranium by Laser Excitation—Message from the President.” Conpressional
Record--Senats $13793, November 3. 1999
John L. Lyman, “Report of IAEA Review Visit to Silex Systems Lid, 17 February
2005," Report to the International Atomic Energy Agency, Safeguards-in-
Confidence, February 25, 2005,
Physical Model, Vol. 3 “Enrichment of Uranium,” STR-314 Restricied
Distribution, International Atomic Energy Agency, Departmenn of Safeguards,
September 2004
Edward (0. Arthur, Jefttey J. Bedell. Hary J. Dewey, John L. Lyman, Anne
Menefee. and Emest J. Moniz, Uranium Enrichment Technolagiet: Materials
from a lus Alamos Workshop, LA-CP-03-0233, March 30, 2003