Science and Global Security, 14:151­162, 2006 Copyright C Taylor &…

Science and Global Security, 14:151­162, 2006
Copyright C Taylor & Francis Group, LLC
ISSN: 0892-9882 print / 1547-7800 online
DOI: 10.1080/08929880600993071




Feasibility of Eliminating the
Use of Highly Enriched Uranium
in the Production of Medical
Radioisotopes
Frank N. von Hippel and Laura H. Kahn
Program on Science and Global Security, Princeton University, Princeton, NJ, USA
Significant quantities of highly enriched uranium (HEU)--more than enough to make a
Hiroshima bomb--are used annually as neutron target material in Canadian, European,
and South African reactors to produce the short-lived fission products used in nuclear
medicine. The most important of these fission products is 99 Mo, which decays into 99m Tc,
which is the most widely used medical radioisotope.
     The U.S. supplies weapon-grade uranium to the Canadian radioisotope producer
and might in the future provide it to the European producers as well. As a condition
for receiving U.S. HEU, the 1992 Schumer Amendment to the U.S. Atomic Energy Act
requires that a foreign producer cooperate with the United States in converting to low-
enriched uranium (LEU) targets. Some smaller producers have already done so. The
Canadian producer has asserted, however, that the cost of conversion would be too
high. The 2005 Burr amendment therefore exempted radioisotope producers in Canada
and Europe from the Schumer amendment's requirements but requested a National
Academy of Sciences study of the feasibility of conversion, setting as a feasibility test
that the production cost be increased by no more than 10 percent.
     We show that paying for the conversion for the largest European production facility
would increase the cost of 99 Mo production there by only a few percent. For the Canadian
facility the production cost could be more than 10 percent but the increase in the cost of
the final 99m Tc-containing radiopharmaceutical would be only about 1 percent. It is also
pointed out that savings in security could well dwarf the costs of converting to LEU if
HEU were no longer present at the production and radioactive waste sites.




Address correspondence to Frank N. von Hippel, Program on Science and Global Secu-
rity, Princeton University, 221 Nassau St., 2nd Floor, Princeton, NJ 08542, USA. E-mail:
fvhippel@princeton.edu



                                                151
152    von Hippel and Kahn

      INTRODUCTION
      The most important nuclear material to keep away from potential nuclear ter-
      rorists is highly enriched uranium (HEU). If a terrorist group acquired about
      50 kilograms of weapon-grade uranium (90-percent 235 U), it could cause a
      nuclear explosion using a simple gun-type device to assemble two sub-critical
      masses into a supercritical mass. This was the design used for the Hiroshima
      bomb. It cannot be used with plutonium because the high level of spontaneous
      neutron emission in plutonium would result in the chain reaction beginning
      before the supercritical mass was fully assembled. As a result the explosive
      power of a gun-type plutonium weapon would be reduced a thousand fold.
           In the 1970s, recognizing the risks of nuclear proliferation and terrorism
      associated with civilian use of HEU, both the U.S. and Soviet governments
      launched programs to facilitate the substitution of non-weapon-usable low-
      enriched uranium (LEU, containing less than 20 percent 235 U) for HEU in
      civilian research-reactor fuel and in radioisotope production targets. This pro-
      gram is now international.1 Its progress and limitations with regard to the
      conversion of research-reactor fuel have been discussed elsewhere.2 This arti-
      cle discusses the issues associated with the use of HEU in the production of
      medical radioisotopes.


      USE OF HEU FOR THE PRODUCTION OF MEDICAL RADIOISOTOPES
      "Targets" of weapon-grade uranium placed in high neutron fluxes near the cores
      of high-powered research reactors, are the principal sources for the production
      of a number of short-lived fission products that have become important to mod-
      ern medicine. In this article, we focus primarily on technicium-99m, which is
      currently used in about 80 percent of all nuclear-medicine diagnostic procedures
      worldwide.3
           Technicium-99m (99m Tc) has a 6 hour half-life and emits a gamma ray when
      it de-excites. Attached to various chemicals, it can be followed by its gamma
      emissions through the body and thereby can be used to examine the functioning
      of various organs. Its short half-life and lack of beta radiation minimizes un-
      necessary radiation doses. It is derived from molybdenum-99 (99 Mo), which has
      a half-life of 2.7 days and decays into 99m Tc. 99 Mo is adsorbed onto the surface
      of a bed of small alumina particles in "generators" from which the 99m Tc decay
      product is drawn off in solution.
           99
              Mo is produced in about 6 percent of all fissions of 235 U.4 Ninety-five
      percent of the global supply is produced by placing a "target" of HEU (usually
      weapon-grade) in or near a reactor core.5 Very roughly 85 kg of HEU are being
      used for this purpose per year in Canada, Europe, and South Africa.6 Less
      than five percent of the 235 U in the target is consumed and, in most cases,
                      Highly Enriched Uranium in Medical Radioisotope Production           153
                               99
Table 1: Reactors producing         Mo for major international distributors in 2005.10

                               Initial  Percent                              Av/Peak
                      Power Operation   of year                            production
Reactor/Country       (MWt) (shutdown) operating          Distributor   (% world demand)
NRU/Canada             135    1957               86      MDS-Nordion          40/80
HFR/Netherlands         45    1961               79      Mallinckrodt         20/30
                                                         IRE                  10/20
BR2/Belgium            100    1961               31      Mallinckrodt          5/15
                                                         IRE                   4/20
Osiris/France            70   1964               60      IRE                   3/20
FRJ-2/Germany            23   1962 (2006)        57      IRE                   3/10
SAFARI South Africa      20   1965               86      NTP                  10/45
Other                                                    Other                 5/10
                                                         Total               100/250


it is not recycled. The HEU in the waste is therefore still weapon-usable and
has accumulated in the 99 Mo-producing countries in amounts that would be
sufficient to make many Hiroshima weapons.7 The gamma radiation dose rate
from this HEU waste is not sufficient to make it self-protecting by international
standards.8
     The world's major 99 Mo production reactors are currently in Canada (99 Mo
distribution by MDS-Nordion), Europe (Tyco-Healthcare/Mallickrodt in the
Netherlands and the Institute for Radioelements [IRE] in Belgium), and South
Africa (NTP) (see Table 1). Although the U.S. accounts for about half the global
99
   Mo demand,9 it currently does not produce 99 Mo.


U.S. EFFORTS TO ELIMINATE HEU TARGETS AND
INDUSTRY OPPOSITION
The U.S. and Russia are the major international suppliers of HEU for use
in research-reactor fuel and isotope-production targets. In 1992, the Schumer
amendment was added to the U.S. Atomic Energy Act to help motivate foreign
consumers of U.S. HEU to switch to LEU.
     One of the requirements in the Schumer amendment is that, as a condition
for the supply of U.S. HEU to foreign reactors, the operators of those reactors
must make the commitment "that, whenever an alternative [LEU] nuclear re-
actor fuel or target can be used in that reactor, it will use that alternative."11
     Small 99 Mo producers in Argentina and Australia are now using LEU tar-
gets and Indonesia's producer is converting to such targets.12 The major pro-
ducers, however, have been resisting conversion.13
     Only one of the four major companies that distribute 99 Mo is currently
importing U.S. HEU for targets, MDS-Nordion of Canada, which accounts for
about 40 percent of global production of 99 Mo.14 It imports about 20 kilograms
154    von Hippel and Kahn

      of weapon-grade uranium from the U.S. per year.15 The European producers
      currently are using weapon-grade uranium that has either been acquired from
      another nuclear-weapon state (France, Russia, or the U.K.) or was exported
      by the U.S. prior to the Schumer amendment.16 South Africa is using highly
      enriched HEU that it produced prior to 1991.
          In 2005, a lobbying campaign sponsored by MDS-Nordion and Mallinck-
      rodt resulted in the Burr Amendment in the National Energy Policy Act of
      2005. This amendment exempts target HEU used by medical radioisotope pro-
      ducers in Canada, Belgium, France, Germany, and the Netherlands from the
      Schumer Amendment's requirements.17 Some U.S. physician groups supported
      this exemption because they were persuaded that enforcement of the Schumer
      requirement would endanger U.S. radiopharmaceutical supplies.18 As will be
      seen later, there is a question as to the future adequacy of world 99 Mo produc-
      tion capacity, but that is because of the aging of the production reactors--not
      the potential impact of converting the targets from HEU to LEU.
          Supporters of the Schumer Amendment were unable to stop the Burr
      Amendment but were able to insert into it a requirement for a National
      Academy of Sciences study on "the feasibility of procuring supplies of medical
      isotopes from commercial sources that do not use highly enriched uranium."17
      The definition of "feasibility" includes an "average anticipated total cost in-
      crease from production of medical isotopes [of] less than 10 percent."
          In 2004, the average price of the 99 Mo used per dose of 99m Tc was about
      $7.50.19 The average cost to hospitals of radiopharmaceuticals containing 99m Tc
      in 2002 was $87 per dose.20 Therefore, if the 10-percent criterion is applied to
      the production cost of the radioisotope, it corresponds to a requirement that the
      cost of radiopharmaceuticals be increased by less than 1 percent.


      COMPARISON BETWEEN HEU AND LEU 99 Mo PRODUCTION
      PROCESSES
      There appears to be no significant technical or safety reason not to replace HEU
      with LEU targets. G. F. Vandergrift from Argonne National Laboratory who
      provides technical support for replacing HEU with LEU targets, has examined
      the impact of conversion on production of: 99 Mo per target, 99 Mo extraction
      time, solution volume, solid-waste and plutonium production, and 99 Mo purity.
      His most important findings are as follows.21


      Production Per Target
          The dilution of the 235 U by four times as much 238 U in LEU in the target
      increases the total volume of uranium in the target. A typical target contains
      only about 15 grams of 235 U with a volume of about 1 cubic centimeter in a
                   Highly Enriched Uranium in Medical Radioisotope Production      155

target volume of hundreds of cubic centimeters, however. The quantity of 235 U
is limited not by volume but by the rate at which the water flowing through
and around the target can carry off the fission heat. Therefore, the addition of
the 238 U can be easily accommodated.


Byproduct Plutonium Production
    The added 238 U increases the amount of 239 Pu produced by neutron capture
in the 238 U. Plutonium is a proliferation concern. The quantity of produced
plutonium is still relatively small, however. For a case in which 0.5 percent of
the 235 U in the target is fissioned, about 1 kg of plutonium would be produced
for every 1,600 kg of weapon-grade uranium that otherwise would be in the
waste.22 For 5-percent 235 U fission, the ratio would still be less than 0.01.


Purity of the 99 Mo Product
     For the same amount of purification, there will be more plutonium left
in Mo made with LEU. The product contains less than 1.6 × 10-14 grams
   99

of 239 Pu per Curie of 99 Mo, however.23 The associated radiation dose to pa-
tients therefore would be less than one ten millionth of the dose from the
99m
    Tc.24

COST OF CONVERSION FROM HEU TO LEU
The economic arguments made by the big producers against conversion to LEU
targets have focused primarily on the costs of the conversion rather than the
cost of operating with LEU targets thereafter. Because there appears to be
no economic advantage to conversion, however, as long as conversion is not
required, the big producers cannot be expected to volunteer to incur the costs
and whatever risks there might be in going first.
    Most of the public debate over conversion has involved the Canadian pro-
ducer, MDS Nordion because it uses U.S.-supplied HEU for its 99 Mo production
targets. The U.S. Nuclear Regulatory Commission licenses these exports. In
1999 and 2000, the NRC held public hearings on these exports because ques-
tions had been raised as to whether MDS-Nordion had been cooperating in
good faith with the Argonne National Laboratory to convert to the use of LEU
targets--i.e., complying with the Schumer Amendment.
    MDS-Nordion currently uses the Atomic Energy of Canada Limited (AECL)
NRU reactor at Chalk River, Ontario to irradiate its 99 Mo-production targets.
The NRU is a multi-purpose reactor that began operations in 1957. An older
reactor, the NRX, provided backup irradiation services until 1993, when it was
permanently shut down. With the age of the NRU becoming an increasing con-
cern, MDS-Nordion decided, for redundancy, to build two replacement reactors,
156    von Hippel and Kahn

      Maple 1 and Maple 2, which are to be fully dedicated to the production of 99 Mo
      and other fission products for radiopharmaceuticals.
           Despite the requirements of the Schumer Amendment, however, the de-
      sign of the new 99 Mo recovery facility associated with the Maple reactors was
      optimized for HEU targets.
           In 2000, MDS-Nordion officials stated to the U.S. Nuclear Regulatory Com-
      mission that only one design change would be required to adapt its new pro-
      cessing facility for LEU targets: increasing the capacity of its waste-calcining
      (drying and oxidizing) system. The MDS-Nordion officials also asserted, how-
      ever, that the space that had been allocated in the new processing facility was
      too small to hold a larger capacity calciner. MDS-Nordion committed to try to
      adapt the recovery facility to LEU targets after it went into operation or, if that
      proved impossible, to build a new molybdenum-99 recovery line designed for
      LEU targets.25
           In 2003, however, MDS-Nordion informed the Nuclear Regulatory Com-
      mission that conversion would not be feasible and that a new LEU processing
      facility would be too costly: (Cdn)$90 million ($77 million US).26 There has been
      no independent confirmation of these claims because MDS-Nordion broke off
      its cooperation on conversion studies with Argonne National Laboratory.
           The new Maple reactors were supposed to come on line in 2000 but were
      found to have safety defects related to both their design and construction. In
      November 2005, the Canadian Nuclear Safety Commission gave AECL an ad-
      ditional two years to bring the reactors into operation.27 It also granted an
      interim extension of the NRU operating license to the end of July 2006 to allow
      preparation of an application to extend NRU operations until 2012.28
           Subsequently, AECL took over project completion and operating costs for
      the Maple reactors and processing facility, relieving MDS-Nordion from a debil-
      itating drain on its corporate finances. Because AECL is a "Crown Corporation,"
      that is, wholly owned and subsidized by the Canadian government, this means,
      in effect, that the Canadian government has taken over the ownership and op-
      eration of the facilities, leaving MDS-Nordion with the role of distributing the
      radioisotopes. The reason given in the AECL press release was to "maintain
      Canada's position as market leader in a high-tech medical enterprise.29
           A.A. Sameh has provided us with his estimate of the cost of converting
      the 99 Mo recovery facilities at Mallinckrodt's Radiochemical Center in Petten,
      Netherlands. Sameh developed the patented KfK 99 Mo recovery process used
      there and directed the Radiochemical Center from 1995 to 2004. He estimates
      the total conversion cost at about $10 million. Most of this expenditure would
      be required for construction of a hot-cell facility to optimize ("polish") the LEU
      process at production scale and obtain test data on the product for the European
      and U.S. pharmaceutical licensing agencies. Use of such a hot-cell facility would
      be necessary to avoid shutting down and using one of the production lines for
      the development and certification tests.30
                   Highly Enriched Uranium in Medical Radioisotope Production        157
IMPACTS ON RADIOISOTOPE AND RADIOPHARMACEUTICAL COSTS
In 2005, roughly 25 million diagnostic procedures using 99m Tc were conducted
worldwide.31 Roughly 40 percent of global sales were delivered by MDS-
Nordion--about 10 million doses (see Table 1). Charges of $0.5­1.6 per dose
would pay off a $77 million investment in the new recovery facility in 30 years,
assuming 6­21 percent fixed charge rates.32
     This estimate is consistent with that which can be derived from information
about the "extraordinary price increases" MDS-Nordion reported in 2000 that
its customers had agreed to accept to help it defray the cost of building the
new Maple-reactor complex--originally estimated at $140 million.33 This price
increase has been reported as being "an initial increase of about 40%" to pay
for the cost for the Maple reactors and the associated 99 Mo recovery facility.
At the time, 99m Tc was being used in about 10 million procedures per year
worldwide, MDS-Nordion controlled about 85 percent of the market and had
estimated $50 million gross earnings per year from its 99 Mo sales--i.e., about
$5 per dose.34 A 40 percent price increase therefore would have been in the
range of $2 per dose. This price increase is roughly in the same ratio to the
$140 million estimated capital cost as our estimated $0.5­1.6 per dose price
increase from a $77 million processing facility.
     A $1 price increase per dose of 99m Tc would be somewhat more than 10
percent of the current production cost for the associated 99 Mo but it would
be less than 2 percent of the cost of the associated diagnostic procedure. The
estimated impact of the $10 million conversion cost for Mallinckrodt Radio-
chemical Center would be lower. This facility supplies roughly 25 percent of the
global market or about 6 million doses per year (see Table 1). A price increase
of $0.12­0.35 would pay off the investment in 30 years with a 6­21 percent
rate of interest. This price increase would be about 2­5 percent of the produc-
tion cost of the 99 Mo and a few tenths of a percent of the cost of the associated
radiopharmaceutical.



Security Cost Savings
There could be a very large cost saving associated with using LEU targets--the
elimination of the very high security costs associated with HEU transport and
storage. It is puzzling that this factor has not been introduced into the debate
at a time when the U.S. National Nuclear Security Administration (NNSA) is
de-inventorying HEU-using facilities because of the associated huge post 9/11
increases in its security budget. The number of attackers (19) involved in the
September 11, 2001 aircraft hijackings has required the NNSA to increase the
size of the "design-basis threat" (DBT) that its guard forces are required to be
prepared to defend against.
158    von Hippel and Kahn

           The estimated total cost per guard is $125,000 per year. For every at-
      tacker added to the design-basis threat against a facility where nuclear-weapon-
      useable materials are used, it would be necessary to add a guard to each of at
      least three posts for five shifts or a total of fifteen full-time guards. On this
      basis, the guard-force cost associated with a design-basis threat of 19 would be
      $36 million per year.35 This dwarfs all the annual conversion charges discussed
      earlier.
           We do not have sufficient information to make an analysis of the security
      cost savings that would result from conversion from HEU to LEU targets but
      it should be taken into account in future cost-benefit analyses such as the Con-
      gressionally mandated study by the National Academy of Sciences.



      RELIABILITY OF 99 Mo SUPPLY
      The redundancy of the 99 Mo supply has improved since the molybdenum-99
      distribution networks have become global. If all the reactors were operating
      at full capacity, they could have supplied 250 percent of 2005 world demand.
      Taking into account the fraction of the year that each operates, they could
      produce on average 175 percent of 2005 world demand (see Table 1).
          This excess capacity is fragile, however. In 2006, the ages of the production
      reactors ranged from 41 to 49 years. The FRJ-2 shut down in 2006. If the NRU
      shut down, the combined production capacity of the remaining 4 reactors, if
      scheduled optimally, would drop to just 100 percent of world demand, which,
      has been increasing by 5­10 percent per year (see Table 1). It may be that some
      of the other reactors could increase their peak production capacities. The dis-
      ciplined schedule of 99 Mo production can conflict, however, with other missions
      at multipurpose reactors. The high level of operation of the NRU, HFR, and
      SAFARI reactors reflect the fact that they are committed to be available to pro-
      duce 99 Mo with only short interruptions. The other reactors currently operate
      as backup producers.
          If the two dedicated 10 MWt Maple reactors come on line, they will alleviate
      the situation considerably. It has been proposed that Europe also build at least
      one new reactor dedicated to molybdenum-99 production in addition to the
      new multipurpose reactors that are being built.36 In the U.S., there have been
      discussions of the possibility of using various Department of Energy or U.S.
      university reactors to provide a U.S. source of molybdenum-99 and proposals
      also have been made to build dedicated reactors.
          Concerns about reliability of 99 Mo supply should not, however, be used as an
      argument for delaying conversion of 99 Mo production targets from HEU to LEU.
      Based on our analysis, conversion appears both technically and economically
      feasible.
                      Highly Enriched Uranium in Medical Radioisotope Production                159
CONCLUSIONS AND RECOMMENDATIONS
The major molybdenum-99 producers are currently using more than enough
weapon-grade uranium each year to make a Hiroshima bomb. Very little of this
HEU is consumed, and large stocks of weapon-grade uranium are accumulating
at the associated waste-storage sites. All national governments should be con-
cerned about this issue. The theft of HEU in any country represents a potential
threat to all the cities of the world.
     To date, only the U.S. government has been working seriously to persuade
medical radiopharmaceutical companies around the world to convert to LEU
targets. Canada's government, for example, which supplied a $100 million
interest-free loan for the construction of the new Maple reactors and associ-
ated target processing facility37 could have required MDS-Nordion to design
the processing facility to be able to handle LEU as well as HEU targets, but
did not, despite a 1997 exchange of diplomatic notes with the U.S. in which
it committed to do so.38 Now that AECL, a Canadian Crown Corporation, has
bought the facilities, the Canadian government should be able to require that
the facility be modified to accommodate LEU targets before it goes into produc-
tion. Once the facility is in use, transitioning to LEU may become much more
difficult if it is impossible to interrupt HEU-target processing for development
and certification testing on LEU targets.
     Europe should not repeat Canada's mistake. Euratom, the European
Union's nuclear regulatory agency, should require that any new molybdenum-
99-production facility in Europe be designed to use LEU targets and require
peer-reviewed feasibility studies on the conversion of existing facilities. South
Africa should do so as well. The costs of these initiatives would be trivial in
comparison to the potential consequences of a theft of some of the HEU.



NOTES AND REFERENCES
1. Information about the Reduced Enrichment Research and Test Reactor program
may be found at http://www.rertr.anl.gov.
2. "A comprehensive approach to elimination of highly enriched uranium from all
nuclear-reactor fuel cycles" by Frank von Hippel, Science & Global Security, 12 (2004),
137; and "Global Cleanout: Reducing the Threat of HEU-Fueled Nuclear Terrorism,"
Alexander Glaser and Frank von Hippel, Arms Control Today (January/February
2006) 1­18.
3. "Production of Mo99 in Europe: Status and perspectives," H. Bonet, B. David,
and B. Ponsard, 9th International Topical Meeting on Research Reactor Fuel
Management, Budapest, Hungary, April 10­13, 2005, available at http://www.
euronuclear.org/pdf/RRFM2005-Session1.pdf. Other fission products used in medicine
are 131 I (8.0-day halflife), 133 Xe (5.2 days), and 89 Sr (50.5 days), "Nuclear Medicine Fa-
cility Survey, SNM 2003: Survey Reporting on 2002 Cost and Utilization" by Denise
Merlino, Journal of Nuclear Medicine Technology, 32(4), (December 2004): 215.
160    von Hippel and Kahn

      4. Most of this 99 Mo comes from the decay of the short-lived fission products, 99 Y(1.47
      seconds) and 99 Zr(2.2 seconds), Evaluation and Compilation of Fission Product Yields
      1993, T. R. England and B. F. Rider, Los Alamos National Laboratory, LA-UR-94­3106,
      ENDF-349 (1994), http://ie.lbl.gov/fission.html
      5.    "Production of Mo99 in Europe: Status and perspectives," op. cit.
      6. RERTR program project execution plan, U.S. Department of Energy, National Nu-
      clear Security Administration, February 16, 2004, Table B5.
      7. Recycling the HEU (as well as conversion to LEU) was seriously investigated at
      Mallinckrodt's Radiochemical Center at Petten, Netherlands around 2000, "Produc-
      tion of fission Mo-99 from LEU uranium silicide target materials" by A. A. Sameh,
      Radiochemical Center Mallinckrodt Medical, presented at 2000 Symposium on Isotope
      and Radiation Applications, May 18­20, 2000, Institute of Nuclear Energy Research,
      Taiwan.
      8. The IAEA standard for self-protection is a radiation dose of one Sievert (100 rems)
      per hour at a distance of one meter, The physical protection of nuclear material and
      nuclear facilities, International Atomic Energy Agency, INFCIRC/225/Rev.4. Five Siev-
      erts is a median lethal dose. The canisters of HEU-containing waste that are shipped
      from Mallinckrodt's Petten facility to the Netherlands interim radioactive-waste stor-
      age facility two years after target irradiation contain 0.4 kg of HEU each and have an
      unshielded dose rate at one meter of 0.1 Sievert/hour, personal communication, Fred
      Wijtsma, Director, High-flux Reactor, Petten, Netherlands, June 1, 2006.
      9.    Marvin Burns, Bio-Tech Systems Inc. personal communications, December 2005.
      10. "Production of Mo99 in Europe: Status and perspectives," op. cit. Years of initial
      operation from Nuclear Research Reactors in the World (IAEA, 2000).
      11. Atomic Energy Act (42 U.S.C. 21 et seq.) Chapter 11, Section 134. Other require-
      ments set by the Schumer Amendment are that a reactor operator can only request HEU
      if no LEU fuel or target suitable for use in the reactor is available and if suitable LEU
      fuel or targets are under development.
      12. "Facts and myths concerning 99 Mo production with HEU and LEU targets," G. F.
      Vandergrift, Argonne National Laboratory, Proceedings of the International Conference
      on Reduced Enrichment for Research and Test Reactors, Boston, MA, Nov. 7­10, 2005.
      13. See, e.g., DOE Needs to Take Action to Further Reduce the Use of Weapons-Usable
      Uranium in Civilian Research Reactors, U.S. Government Accountability Office report,
      GAO-04-807 (2004), 2.
      14.    "Production of Mo99 in Europe: Status and perspectives," op. cit.
      15. Nuclear Regulatory Commission (NRC), "Briefing on proposed export of
      high enriched uranium to Canada," June 16, 1999, public hearing transcript,
      http://www.nrc.gov/reading-rm/doc-collections/commission/tr/1999/19990616a.html, 15.
      After September 11, 2001, the NRC stopped making such information public but recently,
      in response to a request from Alan Kuperman of the Nuclear Control Institute, made
      public the fact that the U.S National Security Administration had requested a license
      to export 15.5 kg of 93 percent enriched HEU to Canada for use in 99 Mo-production tar-
      gets. However, the NRC refused to make public Canada's annual requirements for this
      purpose because the applicant considered that "proprietary information," letter to Alan
      Kuperman from NRC chairman Nils J. Diaz, April 26, 2006, www.nci.org/06nci/06/NRC-
      HEU-export-licenses-2006-Response-May-AK.PDF.
      16. As of the beginning of 1993, Euratom had 13.7 tons of HEU originally exported
      from the U.S., Plutonium and highly enriched uranium 1996 by David Albright, Frans
      Berkhout, and William Walker, (Oxford University Press, 1997), Table 8.1. Euratom
                      Highly Enriched Uranium in Medical Radioisotope Production               161
does not inform the U.S. of transfers of this material within the EU when it is no longer
needed for its original purpose, such to fuel a critical facility.
17.   National Energy Policy Act of 2005, Sec. 630.
18. "Bomb-grade bazaar," Alan J. Kuperman, Bulletin of the Atomic Scientists, (March­
April 2006).
19. Total U.S. 99 Mo sales in 2004 were $150 million for 20 million doses of 99m Tc,
Marvin Burns, Bio-Tech Systems Inc. personal communications, December 2005.
20. Calculated from Table 4 of "Nuclear Medicine Facility Survey, SNM 2003: Survey
Reporting on 2002 Cost and Utilization," op. cit. The average charge for four 99m Tc-
containing drugs to Medicare between July 1, 2003 and June 30, 2004 was $78 per dose,
Medicare: Radiopharmaceutical Purchase Prices for CMS Consideration in Hospital
Outpatient Rate Setting, Government Accountability Office, letter report to the Secretary
of Health and Human Services, July 14, 2005, Table 1.
21. When no other reference is provided, our source is "Facts and myths concerning
99
   Mo production with HEU and LEU targets," op. cit.
22. "Preliminary investigations for technology assessment of 99 Mo production from
LEU targets," G. F. Vandegrift et al., Proceedings of the 1986 International Conference on
Reduced Enrichment for Research and Test Reactors, Gatlinberg, Tennessee, November
3­6, 1986, Table 1.
23.   One Curie (Ci) of 239 Pu has a mass of 16 grams.
24. Because of the difference in half-lives, 1 Ci of 99 Mo would produce 11 Ci of 99m Tc.
We assume that only 2.4 of these 11 Ci are used, however. The standard dose of 99m Tc is
24 mCi ["Nuclear Medicine Facility Survey, SNM 2003: Survey Reporting on 2002 Cost
and Utilization," op. cit.]. One dose of 99m Tc would be associated with less than 1.6 ×
10-16 grams of 239 Pu. The effective dose from inhaling this much 239 Pu is 2.4 × 10-11
Sieverts (Sv), "The hazard from plutonium dispersal by nuclear-warhead accidents,"
Steve Fetter and Frank von Hippel, Science & Global Security 2 (1990): 21. The average
effective doses from 99m Tc procedures are in the range of 1­10 mSv, Sources and Effects
of Ionizing Radiation, U.N. Scientific Committee on the Effects of Atomic Radiation (UN,
2000), Annex D, Table 42.
25. "Briefing on proposed export of high enriched uranium to Canada," July 10, 2000,
Nuclear Regulatory Commission, public hearing transcript, http://www.nrc.gov/reading-
rm/doc-collections/commission/tr/2000/20000710b.html
26. "Nordion headed for a showdown with the U.S.?" Daniel Horner, Nuclear Fuel,
March 15, 2004.
27. The principal safety design problem is a positive power coefficient of reactivity, i.e.,
if the power increases, the reactivity increases--which further increases the power. As of
November 2005, the cause of this problem was still not understood, "Application for the
renewal of the operating license for the MAPLE reactors at the Chalk River Laboratories:
Record of proceedings, including reasons for decision," Canadian Nuclear Safety Com-
mission, November 24, 2005, http://www.nuclearsafety.gc.ca/eng/commission/pdf/2005-
10-18-Decision-AECL-MAPLE-e.pdf.
28. "Application to continue operation of the National Research Universal (NRU) re-
actor beyond December 31, 2005: Record of proceedings, including reasons for decision,"
Canadian Nuclear Safety Commission, November 24, 2005, http://www.nuclearsafety.
gc.ca/eng/commission/pdf/2005-10-18-Decision-AECL-NRU-e.pdf.
29. "AECL and MDS enter into long-term supply agreement for medical isotopes,"
AECL news release, February 22, 2006, http://www.aecl.ca/NewsRoom/News/Press-
2006/060222.htm
162    von Hippel and Kahn

      30. Sameh assumed that the highly enriched UAl3 , targets used at Petten would be
      replaced by 20-percent enriched U3 Si2 targets, A. A. Sameh, private communications,
      January 2006. If the hot cells at the Petten HFR could be use to polish the LEU process,
      the cost of conversion would be only perhaps $1 million.
      31.   "Production of Mo99 in Europe: Status and perspectives," op. cit.
      32. We have used the following approximation to the mortgage formula: Annual pay-
      ment = iC/[1 -exp(-iT)], where C is the cost of the facility, i is the interest rate, and T
      is the payback period in years. The range of annual fixed charge rates considered come
      from an analysis of spent-fuel reprocessing economics in which a 5.8 percent charge
      rate was obtained for a government-owned plant and 20.8 percent for a private venture,
      Nuclear Wastes: Technologies for Separations and Transmutation (National Academy
      Press, 1996), Table J-5.
      33. NRC, "Briefing on proposed export of high enriched uranium to Canada," June
      16, 1999, public hearing transcript, 15, op. cit. By 2003, MDS-Nordion had spent $304
      million on its new 99 Mo production complex, MDS Inc, Veritas Investment Research,
      July 2004, 2.
      34. Evaluation of medical radionuclide production with the accelerator production of
      tritium (APT) facility, Kenneth M. Spicer et al., Medical University of South Carolina,
      University of South Carolina, and Westinghouse Savannah River Co, 1997, 12, 46.
      35. U.S. Nuclear Weapons Complex: Homeland Security Opportunities, Project
      on Government Oversight, May 2005, http://www.pogo.org/p/homeland/ho-050301-
      consolidation.html, 15. This report cites (on p. 9) a NNSA estimate that the new DBT
      will require adding about 100 guards around the clock at each of seven facilities.
      36. "Production of Mo99 in Europe: Status and perspectives," op. cit. Germany brought
      the 20-MWt FRM-II research reactor on line in 2004 and France is constructing a new
      100-MWt Jules Horowitz materials test reactor, which is expected to go into operation
      in 2014. But radioisotope production will be at most a backup mission for these reactors.
      37. MDS-Nordion, Annual Information Form for the period ending October 31, 2005,
      http://www.mdsinc.com/reports/2005 engaif.pdf, 11.
      38.   Cited in "Bomb-grade bazaar," op. cit., endnote 13.