(This document has subscripts and superscripts which may be seen with an appropriate browser.)

I. Nuclear Warhead Dismantlement, Storage and Disposal

Table of Contents

I. Nuclear Warhead Dismantlement, Storage and Disposal

  • Introduction
  • Context
  • Comment on the accelerator-driven reactor
  • Future NATO Science Council activities in Nuclear Warheads Dismantlement, Storage and Disposal



    Richard L. Garwin, Rapporteur

    IBM Research Division
    Thomas J. Watson Research Center
    P.O. Box 218
    Yorktown Heights, NY 10598

    (914) 945-2555

    Adjunct Professor of Physics,
    Columbia University;

    Adjunct Research Fellow,
    Kennedy School of Government
    Harvard University)

    August 19-23, 1993

    A subtopic of
    The NATO workshop on
    Global Stability Through Disarmament
    within the framework of the
    "18th Session of the Planetary Emergencies Seminars"
    of the
    Ettore Majorana Centre for Scientific Culture
    Erice, Sicily

    S245NATO                      090293NATO DRAFT 9                      09/30/93



    IBM Fellow Emeritus, Thomas J. Watson Research Center, P.O. Box 218, Yorktown Heights, NY 10598

    ABSTRACT. The discussions on this subtopic in a small NATO workshop on Global Stability Through Disarmament addressed primarily disposal of excess weapons plutonium (W-Pu), discussing both options for processing W-Pu in current and future reactors, as well as non-fission disposal options such as launching W-Pu packages into solar orbit away from the Earth, and vitrification of W-Pu with high-level nuclear waste for disposal in an engineered repository. The protection of W-Pu from acquisition by terrorist groups or rogue nations that might attempt to acquire it for use in nuclear weapons demands strict measures for safe and secure storage until disposal, and also implies limitations on its wide use in civil reactors (or substantial costs associated with such use). Since a nominal 100 tonnes of W-Pu has a small fraction of the energy value of the 1000 tonnes of Pu produced thus far by civil reactors (largely contained in the spent fuel), the W-Pu must be treated as a security problem rather than primarily as an energy problem. Recommendations are advanced for near-term further study, to narrow the range of practical options.

    The NATO workshop on Global Stability Through Disarmament was held in Erice, Sicily, within the framework of the "18th Session of the Planetary Emergencies Seminars" of the Ettore Majorana Centre for Scientific Culture 08/19-23/93. What follows is a report (1) on the subtopic "I. Nuclear Warhead Dismantlement, Storage and Disposal" as requested by the organizer, Dr. Thomas Ypsilantis.

    There were four presentations relating to this topic, as follows:

    "Safe, Secure Dismantlement and Disposal of 50,000 Excess Nuclear Warheads," by Richard L. Garwin,

    "International Plutonium Management," by Frans Berkhout,

    "A Project of High Safety Level-- Soon, and at Appropriate Cost," by E. Hicken, and

    "High-value Use of Weapons Plutonium by Burning in Molten Salt Accerator-driven Subcritical Systems or Reactors," by Charles D. Bowman.

    An informal paper was presented by Edward Teller, and significant discussion involved also Gregory Canavan, Enno Hicken, Yuri Izrael, Evgenii Velikhov, and Thomas Ypsilantis. A paper had been expected also from Prof. Nikolay Laverov, Vice President, Russian Academy of Sciences, but Prof. Laverov cancelled his appearance in Erice. In addition, three other invited participants, Viktor Mikhailov, Robert Dautray, and Johan Swahn, could not participate. (2)

    Appended are the long prepared paper by GARWIN, as modified by him in the light of discussions at Erice. Also attached is a copy of the foils he used in his oral presentation, and also foils used in his requested summary presentation of 08/23/93.

    In view of his having been invited only a few days before the conference, BERKHOUT did not have a formal paper, but spoke from extensive recent publications (3) and a current Scientific American article. BOWMAN provided a paper, attached.

    The workshop took place in part as follow-up to the NATO Science Committee Meeting of 13 March 1993 with Cooperation Partners, and in the environment of a new NATO emphasis on global security through disarmament. Furthermore, the U.S. government, which had in the Bush Administration been a somewhat reluctant follower to the congressional initiative (to which the names of Senators Nunn and Lugar are attached) now has some of the leading experts in this field in important posts-- e.g. Profs. Ashton B. Carter and Graham Allison in the U.S. Department of Defense. Two successive appropriations of $400 million from the Department of Defense budget have been made by Congress ("Nunn-Lugar funds"), although only a few tens of millions of dollars of this have been spent, and those almost entirely in the United States. The U.S. Administration is seeking ways to ease some of the peripheral restrictions on the expenditure of the funds in the Newly Independent States (NIS).

    Apparently all Soviet non-strategic nuclear weapons were transported to Russia by June, 1992, although thousands of strategic warheads on ICBMs and cruise missiles remain deployed in Belarus, Kazakhstan, and Ukraine.

    According to the U.S. government, nuclear weapons slated for destruction are being dismantled at nearly the target rate of 2000 per year, and Russian officials state that their dismantlement rate is similar or slightly greater. U.S. dismantlements occur entirely at a Department of Energy facility in Pantex, Texas, while Russia has four dismantlement facilities of unstated location.

    There is significant concern that existing U.S-Russian agreements do not commit either side to the dismantling of most of the weapons that will be excess under the arms control and disarmament agreements already committed. Russian officials also state that the rate of dismantlement is limited in Russia by the lack of adequate storage facilities for the plutonium (Pu) and high-enriched uranium (HEU) arising from the dismantlement process, which parts must be stored in special individual containers to ensure that these fissile materials do not approach other such materials sufficiently closely to induce a self-sustaining chain reaction.

    In the United States, adequate low-cost and secure storage is available at Pantex for the storage containers housing individual weapon "pits" arising from dismantlement, and the HEU is shipped to Oak Ridge for storage and eventual blending to make fuel for nuclear reactors.

    Although the U.S. has committed to purchase 500 MT of HEU from Russia over the next 20 years, in the form of LEU for power reactor fuel, those two decades and at least a decade for the dismantlement process means that there must be safe, secure storage for these fissile materials during that interval.

    The requirements on the storage are to prevent the theft of fissile material, a matter of world concern because of the potential use in the fabrication of nuclear weapons; as well, there is concern about the diversion of fissile material for rearmament in Russia or the U.S. The proliferation hazard is dominant at the present time.

    The primary topic of the workshop (after setting the stage in greater detail than above) was to focus on the determination of value of weapons plutonium ("W-Pu"), and means for its disposal.

    In brief, Garwin's analysis shows that given the availability of uranium ore at low cost, a surplus of enrichment capacity, and the experience in burning LEU in conventional power reactors, even free W-Pu would not be accepted by a reactor operator or fuel fabricator, because the fabrication cost alone of W-Pu into so-called mixed-oxide fuel (MOX) (4) is greater than the free market price of fabricated LEU. Nevertheless, the amount of subsidy required in world security interests to fabricate the unit of account of 100 MT of W-Pu is only some $11,000/kg of W-Pu, or about $1.1 B. Unfortunately, normal light water reactors (LWR) can without modification burn only 1/3-core loads of MOX, and the cost of guarding against theft and proliferation the MOX fuel in transport and storage at the reactor (before loading into the reactor) would be substantial if many reactors were involved. As a benchmark, it is often stated that commercial enterprises in Europe charge some $2000-3000 annually per kg for the storage of separated Pu, but the investment cost of a storage facility of 50 MT capacity is said to be some $100 M ($2000 per kg) or some $235/kg-yr if the facility is used for 20 years, at a capital charge rate of 10% (in real dollars-- i.e., assuming zero inflation).

    GARWIN notes also that a few reactors are designed for full-core MOX, and that the cost of secure shipping and storage would be reduced if such reactors could be used. In particular, it might be possible to modify the approximately ten modern 1000 MWe reactors under construction in Russia to take full-core MOX. The options considered for reactor processing of W-Pu include SPIKING, in which no significant amount of Pu is destroyed, but enough fission-product radioactivity produced so that the fuel is to some extent self protecting for some years.

    SPIKING incurs the full cost of fuel fabrication, for normal LWR reactors SPIKING can be performed only at the cost of "capacity factor" so that the reactor may run only 40% of the time instead of a typical 75% of the time, and the radioactivity barrier to access to considerably lower than that of normal spent fuel (in the proportion of exposures of 1500 MWD/MT vs. some 40,000 MWD/MT for normal fuel).

    The SPENT FUEL option would fabricate W-Pu into MOX or other appropriate fuel and burn it to exposures similar to that of LEU. If higher concentration of Pu could be achieved by the use of burnable poison, fuel fabrication costs would be reduced for the 100 MT nominal amount of Pu.

    The DESTRUCTION option of burning almost all the Pu in nuclear reactors would in all cases require chemical processing. It is important to distinguish between burning almost all the Pu239 and the more difficult task of burning all the isotopes of Pu. Pu240, present to the amount of about 6% in W-Pu, amounts typically to 23% of the Pu discharged from a normal LWR. Although Pu240 is not subject to fission by the thermal neutrons in a LWR, nevertheless, Pu240 and essentially any mixture of isotopes of Pu can be used to fabricate a potent nuclear weapon. (5)

    The DESTRUCTION option must be considered in the context of the combination of economics and security aspects of the much larger quantities of civil Pu present largely as spent fuel, but increasingly constituting a stockpile of separated Pu, as a result of reprocessing of power reactor fuel carried out by a number of nations. There is almost 1000 MT of reactor Pu (R-Pu) in existence now, with the amount growing by about 100 MT per year.

    GARWIN also discussed non-reactor options for disposal of W-Pu, considering them on the basis of immunity from theft, resistance to rearmament, and their environmental suitability and cost. In view of the subsidy required for use of Pu in LWR, a hypothetical non-energy-using disposal option at zero cost would be very attractive. Unfortunately, the options that score highest against proliferation and rearmament pose severe legal and environmental problems (e.g., dilution in the ocean). Shooting Pu into solar orbit is vastly expensive, and storing Pu underground after destruction of many pits in a single nuclear explosion, storing pit fragments in a spent-fuel repository, or in deep boreholes all are susceptible to access for rearmament.

    Nevertheless, in view of the fact that W-Pu will need to be stored for many years before it is used in reactors (if that is the choice), both reactor and non-reactor disposal need to be studied seriously in a programmatic way.

    GARWIN notes that reactors of new design offer no significant benefit in the rate of destruction of W-Pu per unit thermal power (6) (about 1000 kg per year of fission per 1000 MWe plant), and that reactors of existing design should therefore be considered most seriously for this mission. It is clear that the avoidance of modest subsidy required for burning W-Pu in LWR could not in any way support the development of a new-design reactor.

    BERKHOUT's presentation drew heavily upon his book and also his current article in Scientific American, (7) from which the following phrasing of Berkhout's position is largely taken.

    BERKHOUT divided his presentation into 3 parts:

    1. Military and civil plutonium stocks and arisings.

    2. Existing controls on plutonium production and use.

    3. Extending controls--
    He sketched the magnitude of the nuclear materials problem by noting the World inventory of HEU (end 1990) as some 1330 metric tons (MT), of which 810 MT are in nuclear warheads and 20 MT in research reactors. (8) As for Pu, BERKHOUT assesses the end-1990 world total as 910 MT. Of the 258 MT of military Pu, 178 MT are in warheads and 80 MT outside; of the civil Pu, 531 MT is in spent fuel, 72 MT stored as PuO2, and 49 MT have been recycled as reactor fuel. (9)

    BERKHOUT notes that although there exists in Japan and some Western European nations a partial infrastructure for recycling plutonium recovered from spent power-reactor fuel, the electric utilities in these countries have no interest in incorporating W-Pu in their reactors, because they already anticipate a significant surplus of civilian plutonium, and the cost of manufacturing MOX is currently considerably greater than the cost to purchase LEU.

    Furthermore, neither the U.S. nor Russia has a MOX plant capable of operation.

    Considering the reluctance to use W-Pu in LWR plants in countries which are already using MOX, and the inability immediately to use MOX in the U.S. or Russia, BERKHOUT and his colleagues propose vigorously that the W-Pu be vitrified with high-level radioactive waste, from which the W-Pu was initially separated. Taking specifics from the vitrification plant soon expected to begin operation in South Carolina, he notes that this facility is expected to produce at least 8000 tons of radioactive glass in the form of massive steel-sheathed "logs" three meters long and 0.6 m in diameter, each containing about half a ton of high-level wastes slurry mixed with 1.2 tons of borosilicate glass. 70 tons of W-Pu could be dissolved in these logs without raising the concentration to levels above those in spent power-reactor fuel; the nominal 100 tons of Pu would raise the Pu concentration to about 40% above that of spent fuel.

    BERKHOUT discusses the five-year delay probably required to complete safety and other preparations and concludes that W-Pu vitrified with HLW in this way and incorporated in a geological repository would be no more accessible to a proliferator or even to rebuilding nuclear weapons than is Pu from spent fuel.

    BERKHOUT notes that Russia vitrifies HLW at a site near Chelyabinsk. However, Russia uses phosphate glass instead of the borosilicate glass used in Western Europe, Japan, and the U.S. Phosphate glass is more soluble and does not contain the boron that prevents criticality with substantial amounts of Pu.

    BERKHOUT notes that the Russian nuclear establishment has shown "little enthusiasm for glassification or, more generally, for processing plutonium into more diversion resistant forms. This material was produced at enormous human and environmental costs; Russian nuclear officials consider it a national heritage."

    But in view of the cost and hazard of safeguarding the W-Pu for decades, Russia should take an informed decision on what to do about the W-Pu. BERKHOUT advocates comprehensive monitoring of disposal of surplus warheads, and a cutoff of production of weapons-usable materials worldwide (not just in the U.S. and Russia).

    Focusing on ultimate disposition of W-Pu, TELLER proposed that the fissile materials from nuclear weapons should be used in nuclear reactors, and in specially constructed reactors that would make a major contribution to energy supply, that would isolate dangerous substances, and that would eliminate the waste disposal problem. He said this should appeal to everyone so that there will be enthusiasm from all sides.

    TELLER maintains that reactors and people do not belong together, so that reactors should be placed 100 m underground and only communicated with via regulation of the coolant flow. The reactors should be buried without possibility of access, and it is to be hoped that they would have a core of 30-year life after which they would end their operation and entomb themselves.

    There should be no moving parts in the reactor, and it should be foolproof, because there should not even be fools involved. TELLER says "Don't use dangerous fast breeders, which use a metal coolant that is flammable. Use a thermal breeder, using Th232. Then end of life will not be due to lack of fuel but probably to components giving out." The resultant core will be highly radioactive and self protecting because of the U232 produced by (n-2n) on U233, which after a number of alpha decays ends in Tl208, which is a potent gamma emitter.

    TELLER notes that there are some technical problems, since one needs to ensure a negative temperature coefficient of reactivity for 30 years, with the reactor changing completely from its initial to its equilibrium and final configuration. One needs an appropriate amount of fissile fuel, thorium, actual tests, etc.

    Then the concept is to command a meltdown at shutdown, yielding a radioactive mass at high temperature, which will then burrow its way into the Earth. Of course, this should be "put as far as possible from a water table."

    TELLER would like a strong international effort to design such things, although not necessarily exactly what he has sketched. According to TELLER, this is a scientific jump well worth considering.

    VELIKHOV commented that he did not like the idea of a meltdown at at the end.

    GARWIN agreed that it would be useful to have reactors 100 m underground, and HICKEN indicated that studies in Germany showed that this was feasible but would increase cost by about 40%. GARWIN opined that TELLER was too negative about other reactor concepts and too positive about one that had had essentially no analysis. In particular, liquid-metal cooled reactors using lead instead of sodium were being designed in Moscow, with support by Westinghouse, with the advantage that the lead does not react violently with water or air, and that it is much farther from its boiling point than is sodium in normal operation. As for the thermal breeder, a reactor with a 30-year core would need to convert most of the fertile material into fission products during the core life, causing a great problem for fuel integrity, unless it had an enormous core. As an example of other options, proposals have been made in Japan for reactors with a candle-like core, with a reflector that makes the core critical as the reflector is moved up at a rate of about 1 mm per day.

    GARWIN also indicated that the core-melt scenario for ending the reactor's life was extremely problematical, that such a design would be difficult to certify, and that it was unlikely that one would really want to put a reactor out of contact with possible repair for 30 years.

    HICKEN spoke on "A Project of High Safety Level-- Soon and at Appropriate Cost," a presentation of new, publicly acceptable, safe reactor concepts. These are French (Framatom) and German (Siemens) projects for thermal spectrum reactors, especially high temperature, gas cooled (graphite) reactors. They seem to be about double the cost per kWh of existing French LWR plants, and putting them underground would add 20% additional. HICKEN emphasized that it is necessary to be very conservative in these analyses, and that not only the concept, but the detailed design, fabrication, and licensing were important to such a project.

    BOWMAN made a presentation on "Accelerator-Driven Subcritical Systems for Energy Production," with considerable bearing on the disposition of W-Pu. BOWMAN advocates this approach, primarily as a way of handling both fission products and waste actinides from existing and future LWR. BOWMAN quoted a statement that "no repository will be open anywhere in the world for at least 20 years" to indicate that there is ample time to develop any promising approach competitive to geologic storage.

    The concept is to use a sub-critical chain reacting system with a K = 0.95, with a high-energy proton or deuteron accelerator being used to produce neutrons that are then multiplied by a factor 1/(1-K) in the sub-critical reactor. For K = 0.95, the multiplication factor is just 20.

    The combination of actinides and fission products is neutron-poor, and does not readily make a critical thermal reactor, although a fast-reactor can burn the actinides. However, the fast reactor will do little to transmute the fission products. BOWMAN proposes systems sized at 3000 MW(t) and burning about 1200 kg of actinide per year. BOWMAN proposes to use a molten fluoride salt LiF-BeFsub2 used in the Oak Ridge molten salt reactor experiment (MSRE) which operates at about 1100 K and allows a thermal-to-electric conversion efficiency of about 40%. BOWMAN's calculations indicate that it takes 55 mA of 800-MeV protons to drive a K = 0.95 system at 3000 MWt (about 44 MW of beam power, requiring on the order of 100 MWe of electrical power to the accelerator. This assumes that each of the protons incident directly on the molten salt produces on the average two high-energy neutrons that then cause fission or spallation in a U238 or Pb cylinder surrounding the target itself and within the reactor. All together, it is assumed that each 800 MeV proton produces 25 neutrons.

    BOWMAN proposes that fertile Th232 be added to the molten salt, which would allow breeding with a Keff = 0.97 if no LWR waste is burned, and about K = 0.94 if the annual waste from one LRW is burned per year. This assumes isotopic separation of Cs in order not to burden the neutron economy with stable Cs.

    According to BOWMAN Keff falls below the K = 0.95 needed for the accelerator-driven sub-critical system if two LWR wastes per year are burned, but may be restored to 0.95 if the fraction of Pu in the (Th+Pu) feed is raised to 40% (240 kg of Pu per year).

    Among advantages claimed for the accelerator-driven sub-critical system are

    With continuous reprocessing to remove fission products and with isotope separation of the Cs, there is to be "no long-term, high-level waste stream"; all short-lived fission products are to be retained in engineered storage on site.

    Fission products are to be separated from the heavy fluorides in a novel centrifuge system operating at the temperature of the molten salt (450 C in the centrifuge, compared with 1100 C in the reactor).

    According to the paper provided by BOWMAN, the partial substitution of Pu239 for accelerator beam saves an investment that amounts to $250,000/kg of Pu, which BOWMAN ascribes to the "value for the plutonium for this system."

    In response to questions, BOWMAN noted that the baseline proposals from his laboratory (Los Alamos National Laboratory) were for an aqueous approach using heavy water slurry of oxide fuel and fission products, which has much lower thermal efficiency for the production of electricity and so poorer economics. Furthermore, the aqueous approach has a much higher "inventory" of actinides, amounting to about 12,000 kg for a 6000 MWt system that would provide about the same net electrical output as the molten salt system. Similarly, the capital cost of the aqueous system would be much larger, in view of the necessity for a 450 MWe accelerator, in contrast with the 100 MWe accelerator for the molten salt system.

    BOWMAN stated that in future briefings by LANL, the aqueous system will be replaced by the molten salt concept. Apparently, the aqueous approach has been the baseline at LANL a large part because much more is known about the separation technology than in the case of the molten salt approach, although there are many unknowns and gaps in our knowledge in both approaches.

    GARWIN observed that according to HICKEN "runaway is not the most critical problem in reactor design" and so a principal advantage claimed for the accelerator-driven systems is the supposed absence of a problem that has long been conquered in almost every one of the operating reactors worldwide. Chernobyl was an example of runaway, due to a combination of poor design choices and flagrant mishandling by operators and managers. As for the "instant shutdown" provided by the accelerator, the motion of control rods will shut down a critical reactor in 0.01 s. (10) In addition, fluid-fuel reactors have substantial variation in reactivity as a function of temperature, especially the aqueous approaches. (11) Furthermore, the fission product decay heat from the sub-critical reactor operating at a given thermal power is precisely that from a normal reactor. If there is a loss of flow accident, fuel will vaporize. If there is a failure (12) of the reactor vessel, fluid fuel will escape, depositing on surfaces where it will deliver initially some 400 MWt of heat energy. (13)

    BOWMAN considers several concepts for these molten salt systems and recognizes that Pu239 can provide plenty of neutrons so that an accelerator is not needed. For instance, If no thorium is fed to the reactor, the system will still produce electrical power, without an accelerator, and could burn the waste from somewhat more than two LWR per year.

    Comment on the accelerator-driven reactor
    But one must be extremely cautious about ascribing a "value" to an element of a process (14) in this way. The derived "value" of $250/g for W-Pu (or for HEU) has merit only if such systems are otherwise economical with accelerators. This high "value" simply shows how much one would need to pay to replace neutrons packaged in HEU or Pu by those created by an accelerator beam.

    Indeed, the operators of such a system who would wish to minimize their costs would be strongly motivated to move to as much HEU or Pu239 as could be accommodated while maintaining the safety margins of the system. The "value" of Pu or HEU would still be that derived from other applications, which are buffered by the cost of supply of enriched uranium-- much more in the neighborhood of the $20/g that is being paid by the U.S. to Russia for the 500 MT of HEU to be delivered over the next 20 years. And the value of Pu would continue to be reduced below that of medium- enriched uranium by the cost of safeguards and security required by the proliferation and health hazard of plutonium. (15)

    For the disposal of actinide and fission product waste, the accelerator-driven approach ultimately must be compared with the cost of mined repositories, if those continue to be judged acceptable from an environmental point of view; and from the point of view of long-term nuclear power production, beyond that able to be supplied by enriched uranium, the accelerator approach is in competition with the fast breeder reactor or with the Th-U233 breeder reactor itself.

    As for the economics of operation of such systems, any of the accelerator-driven or reactor-based systems shown in Table 1 of BOWMAN's paper is indicated to produce "power value" just about four times the invested capital. The "power value" represents the undiscounted sales price of the power produced over the assumed 30-year Pu-destruction campaign. Society might then want to know which of these options to select in order to achieve the goal of destruction of the nominal 100 MT of W-Pu, together with the destruction of LWR waste and the production of electrical energy.

    If the process were indeed profitable-- if the net present value of the revenue stream according to competitive discount rates were positive (net benefit), then the largest investment-- that with the least capability to destroy W-Pu-- would be preferable. If the approach is not profitable from the point of view of sale of electrical energy, then one needs to ask how much one is paying and what are the alternative approaches for the destruction of the W-Pu or the LWR waste. Keep in mind that a subsidy on the order of $1-2 B is all that is required for the destruction of 100 MT of W-Pu in reactors of conventional type. Specifically, if we inquire about the worth of the investment from the point of view of power generation, and we assume that the capital is invested at a steady rate over ten years, while the power is produced at a steady rate for the following 30 years, then the discounted net present value (DNPV) of the $10 B investment (at 4% real discount rate) is $8.6 B, while the discounted NPV of the 30-year revenue stream is $16.4 B, for a total DNPV of $7.8 B benefit.

    This does not include any operating cost for the system, any contingency, or the like, and a 4% annual discount rate is very low for such analyses. If on the other hand one took an annual discount rate of 10%, then the same calculation would give a $3.7 B negative net present value (loss), so that the approach would not be profitable, although it might be desirable if it achieved other "goods" not captured in the price of the power produced and the cost of the capital investment.

    The cost-averse approach to destroying the W-Pu would clearly involve a simpler reactor, with no accelerator, no thorium, and with some burnable poison like B10, or one that provides a more negative temperature coefficient, like erbium.

    Future NATO Science Council activities in Nuclear Warheads Dismantlement, Storage and Disposal
    The transport, dismantlement, and destruction of nuclear warheads agreed between the US and Russia is largely a matter for those two nations, and one in which NATO probably has little role. However, all NATO nations have competence in nuclear energy and an understanding of the importance of safe and secure storage of the fissionable materials being removed from the excess nuclear weapons.

    Every nation and indeed, every individual, has an interest in preventing the proliferation and use of nuclear weapons. An important question is for how long the weapons materials (high-enriched uranium and plutonium) will be accessible, and what is to be done with them. The NATO Science Committee has already requested E. Hicken as a Panel 1 member to elicit proposals to organize a workshop in Moscow on use of MOX in LWRs and has asked R.L. Garwin to organize a workshop in Paris in December or early in 1994 on the more general topic of plutonium prospects and problems. The Erice workshop was far less specialized than either of these, and was organized much more rapidly and with fewer participants experienced in the field; nevertheless, it does point to the need to encourage and to support

    Nuclear power beyond the year 2100 will need to compete with supplies right now feasible, e.g., solar power sources, perhaps with fusion, and with approaches perhaps yet to be discovered. It is important to have a number of technical options for fission power available 60 to 80 years from now-- not to select one now for early deployment if it is less economical than LWR with current low-cost uranium.

    This it would seem that detailed analysis and exploration of the adequacy of our knowledge should proceed in interactive and cooperative fashion on at least the following concepts:

    In addition to these options for supplying energy from abundant U238 or thorium, substantial analytical effort and innovative technological exploration should be directed to waste disposal and waste destruction approaches: While the nature of a technological approach may be compelling to an individual, another approach may be just as attractive to another. The common ground is analysis with economic, environmental, and security components, (16) and it is both important and urgent that scientists and technologists interested in nuclear energy both learn such techniques and standards and also involve the professionals from those fields.

    Richard L. Garwin


    (1) Remarks ascribed to a person are intended to be an accurate abbreviation of that person's presentation or comments. Statements not so ascribed are the views of the Rapporteur. An earlier but not very different version of this report was sent 09/15/93 to the speakers and to some of the more active commentators, and the Rapporteur has taken into account responses received by 09/30/93; this document, however, does not purport to be a consensus, but rather a factual report of positions presented, including those of the rapporteur, with additional commentary by him. Details of the presentations and interchange are appended.

    (2) All individuals named are identified further in the list following this report.

    (3) F. Berkhout, A. Diakov, H. Feiveson, H. Hunt, E. Lyman, M Miller, and F. von Hippel, "Disposition of Separated Plutonium" in Science and Global Security, Vol. 3, Nos. 3-4, pp. 161-214 (March 1993).

    (4) Typically 3-4% Pu in U238 ceramic oxide.

    (5) J.C. Mark, "Explosive Properties of Reactor-grade Plutonium," to be published.

    (6) (if that is the criterion of goodness)

    (7) "Eliminating Nuclear Warheads" by F. von Hippel, M. Miller, H. Feiveson, A. Diakov, and F. Berkhout, Scientific American, August 1993.

    (8) The October NUKEM Market Report contains an interview with MINATOM Minister Viktor Mihkailov, in which Mikhailov states that the 500 MT of Russian HEU to be sold to the U.S. constitutes some 40% of the Russian total. This would make the Russian HEU alone total some 1250 MT.

    (9) Current figures used by BERKHOUT in his talk indicate the world stock of separated civil Pu as 85 MT, of which 40 MT are at Sellafield in Britain, and 30 MT at Chelyabinsk, Russia. Some 650 MT are now present in spent fuel, with about 70 MT of Pu discharged each year from the world's civil reactors.

    (10) Special shut-down rods driven by individual pressurized gas reservoirs can be reliably triggered by local temperature sensors. Whether such a system is categorized as "active" or "passive" is less important than the testable reliability that can be achieved.

    (11) A very large negative temperature coefficient of reactivity can be can be a hazard if there is any prospect of access of cold fuel to the reactor.

    (12) (very unlikely, of course, for an unpressurized molten salt vessel)

    (13) BOWMAN indicates that means have been conceived for mitigating the consequences of such accidents in a fashion simpler and cheaper than for LWRs (personal communication 16 September 1993).

    (14) The very interesting paper by BOWMAN provoked much discussion, some of which is reported above; more comment is to be found in the brief summary paper "Uses of Plutonium for Peaceful Purposes" requested by the organizers to close the Workshop, and presented by GARWIN. Additional commentary is provided by the Rapporteur in this section.

    (15) There is certainly no consensus yet regarding the value of plutonium, usually determined for a commodity by the price at which commercially motivated transfers take place.

    (16) (the inputs to social and political decisions)