V095PCWP 040496PCWP as delivered 04/23/96
The Post-Cold War World and Nuclear Weapons Proliferation
***As distributed at Nagoya, but with Table 1 added for presentation.***
That this problem is real is demonstrated by the experience in Japan with the Aum Shinrikyo-- one of the largest and most technically aggressive terrorist groups that has come to light.
The good news is that there is a strengthening
consensus against proliferation, as evidenced by the fact that 175
nations in May 1995, without exception,
achieved the indefinite extension
of the Non-Proliferation Treaty.
by which they all voluntarily deny themselves
the development or possession of nuclear weaponry except for the
five nuclear weapon states-- Britain, China, France, Russia, and the
United States.
India, Israel, and Pakistan are widely
believed to have nuclear weapons (even if unassembled),
and they are not signatories to the NPT.
In exchange for this self denial on the part of the non-nuclear
weapons states (NNWS), the NWS have undertaken to agree a Comprehensive
Test Ban Treaty (CTBT) in 1996, and in principle to take other measures
to reduce the grossly excessive destructive power of nuclear weaponry
that exists in the world today.
I will briefly review what would be needed to acquire nuclear weapons, the mechanisms for preventing or countering such acquisition by sub-national groups, the new problems of the post-Cold War world, and specific actions by various nations or groups of nations or international organizations that might help to prevent or to counter proliferation of nuclear weapons.
There is much that can be done not only by the G-7, but by other nations, and even by nongovernmental groups such as the Pugwash movement, which shared the 1995 Nobel Peace Prize with one of its founders, Professor Joseph Rotblat, largely for its efforts and considerable success in helping to limit proliferation and use of nuclear weapons.
But the world has changed enormously since 1945 when the United States exploded over Hiroshima a fission bomb yielding some 15,000 tons of high-explosive equivalent (15 Kt) of energy, destroying the city. Four days later, the city of Nagasaki was destroyed by a 21 Kt fission bomb.
The Nagasaki bomb contained approximately 6 kg of plutonium (almost pure Pu239), while the Hiroshima bomb had as its fissile core some 60 kg of high-enriched uranium (HEU) about 95% U235.
The refinement of nuclear weaponry proceeded rapidly in the 1950s, after the Soviet Union exploded its first atomic bomb in 1949, with the result that nuclear weapons have been built weighing tens of kilograms instead of tons, and with yields from 20 tons of high explosive to 60 megatons or more.
The United States has conducted 1149 nuclear tests, the Soviet Union some 1100, Britain 45, France 210, and China 43 (to the end of 1995). In addition, India conducted one nuclear test in 1974.
Many of the nuclear weapons in the five NWS consist of two-stage systems in which the explosion of a "primary" nuclear explosive leads to the assembly of a "secondary" charge usually containing both fusion fuel and fissionable material-- U235, Pu, and/or U238.
Over the last half century, computer technology has evolved so that even the smallest PC has more computing capability than the largest computers that were used to design the nuclear weaponry of the 1940s, and the general evolution of technology has made simple many of the operations required to produce a nuclear weapon. The control over the proliferation of nuclear weapons, in my judgment, must be exerted in two ways-- over the will to build nuclear weaponry, and over the necessary fissile material.
In addition to the general evolution of technology, there exists in 1996 a nuclear industry that is widespread and that handles enormous amounts of fissile materials. Furthermore, as a result of the 1955 Atoms for Peace Initiative, education and training in nominally peaceful aspects of nuclear power has been widely available, including the techniques required to obtain plutonium from spent fuel. Of course, such Pu has a perfectly legitimate function in "recycle" in order to reduce the amount of uranium and enrichment services required for the operation of a nuclear reactor system; unfortunately, with the current availability of low-cost uranium, such "reprocessing and recycle" appears to cost more in every country than would direct disposal of spent fuel into a mined geologic repository. (1)
With the end of the Cold War and the elimination of the enmity between the United States and the states of the former Soviet Union, it became abundantly clear that the nuclear weapon stocks of the US and the CIS were grossly excessive. At its peak, the US stock was some 33,000 warheads, while the Russian holdings were some 45,000. Accordingly, both formal arms control agreements (START I and START II) were accelerated as well as less formal agreements between US President Bush and Presidents Gorbachev and Yeltsin, with major consequences.
All US tactical weapons were removed from ships, and with other tactical weapons were brought back to the United States-- many of them to be destroyed. All Soviet tactical weapons were returned to Russia from the other former Soviet states, and all strategic warheads are committed to return to Russia-- the only ones remaining outside Russia being those on missiles in Ukraine-- these warheads to be returned to Russia in 1996. This was not a minor force, with some 1900 strategic warheads and another 600 warheads on cruise missiles in Ukraine at the time of dissolution of the Soviet Union in 1991!
So Russia is the sole nuclear heir to the Soviet Union, and the other former Soviet states have joined the NPT as NNWS.
In conjunction with these weapon reductions and dismantlements, more than 50 tons of weapon plutonium is expected to be made surplus from the US inventory by the year 2003, and at least as much in Russia. Arms reductions thus pose a new aspect of the proliferation danger, since W-Pu has been transformed in just a few years from a valuable commodity in short supply to a material in surplus. Similarly, those nuclear weapons slated for destruction have lost their status as a scarce and valuable resource; instead, they are a costly (2) and dangerous problem.
For planning purposes, the US DOE authorized the statement (3) that US nuclear weapons to be dismantled contain on the order of 4 kg of W-Pu each. Instead of multi-ton weight, such weapons may be in the tens or 100 kg class. The major improvement over the 1945 weapons arose from the adoption of a "hollow pit" instead of a solid mass of Pu, enabling the Pu to be not only assembled by the explosive but also more readily compressed, thereby not only reducing the amount of Pu needed but also improving the safety of the unexploded weapon.
The first Chinese fission bomb was tested in 1964 with a yield of some 22 Kt and weighed some 1550 kg. It was an implosion device using U235, and could therefore use less fissile material than the gun assembly method.
South Africa recently revealed that prior to joining the NPT it had constructed six gun-assembled uranium weapons, which have all been dismantled and the U235 blended down so that it is no longer HEU (to below 20% U235).
Thus, if the fuel is removed at relatively low burnup (as is necessary in any case in a natural-uranium reactor) the requisite 6 kg of Pu would be obtained from something like 6000 MW(t)d or from about 17 MW(t)-year of reactor operation. So a 17 MW(t) reactor will produce about 6 kg of Pu239 per year.
This is a very large amount of thermal energy (5) for an ordinary householder, but not enormous for a substantial industrial operation. If such a reactor were used to produce heat at temperatures high enough to produce electrical power at about 30% efficiency, if would provide 5 MW(e) or about 40 million kWh in a year, which at $0.05 per kWh would be worth some $2 million per year. Alternatively, if the reactor were used simply to heat river water, with a 40 C temperature rise, one would need about five tons per minute of coolant.
Taking the U235 route to an implosion weapon and assuming that some 20 kg of 90% U235 would be needed, one can calculate that some 176 kg of natural uranium would be required per kg of product, and some 230 SWU (6) per kg of product. At $20 per kg for natural uranium, the feed material would cost some $80,000, but the 4600 SWU at about $100 per SWU would cost some $460,000. By taking a commercial price for separative work from either centrifuge or gaseous diffusion plants, we are considering plants that are amortized over years of operation. The investment for a proliferator to make the first 20 kg in a year or so would be much higher than this allocated cost.
Furthermore, the enrichment process uses UF6, which is not a common article of commerce except for nuclear energy.
The acquisition of kilograms of the artificial element Pu requires well known techniques for chemical separation of Pu from the irradiated fuel and, especially, from the highly radioactive fission products, together with the disposal of the waste material. It would be simpler for a sub-national group interested in proliferation to attempt to acquire (by purchase or theft-- or by purchased theft) the requisite material than to process spent fuel to obtain the plutonium. But the plutonium in "fresh" (not yet irradiated) Pu-bearing mixed-oxide ceramic ("MOX") fuel does not have significant penetrating radiation and is thus a much easier chemical task than is irradiated fuel.
On the HEU side, the problem is slightly different. There is not a lot of surplus HEU outside the weapon (and naval fuel) industry. However there is a great deal inside: The United States has contracted to purchase from Russia 500 tons of surplus HEU, to be blended down to LEU and delivered as 4.4% U235. Six tons of Russian HEU were delivered in this way in 1995, and 12 tons will be delivered in 1996. However, the Russian Minister of Atomic Energy has indicated that Russia has a good deal more HEU-- probably 750 tons in addition to this 500 tons. Even 100 tons of this HEU would suffice for 5000 implosion weapons.
Fortunately, the HEU problem is much more tractable than is the Pu problem, both because enriched uranium is a valued article of commerce and because it can be blended quickly and at low cost with natural uranium or depleted uranium to form an intermediate stock of 20% U235, no longer usable in weapons, but retaining its value for use either as 2-5% U235 or as 10-20% U235 in fast-spectrum reactors.
"It follows that both separated reactor-grade plutonium and weapons-grade plutonium are credible targets for sub-national theft or seizure, and that protective measures must be designed to take this into account."
The weapons usable materials-- HEU, W-Pu, and R-Pu-- in separated form share the characteristics that an individual can carry enough such fissile material to supply a nuclear weapon, without hazard from radiation in the process. >From this point of view, and from that of time between acquisition of the material and the availability of a bomb core, these separated fissile materials are in a class by themselves. The next stage in hazard is posed by "fresh" MOX fuel-- somewhat mitigated by the 5% Pu content, the 500-cm length of the fuel elements, and the need for chemical separation.
CISAC-1 introduced a rationale for handling the proliferation threat of excess weapons plutonium (and uranium):
This was not the result of a conclusion that there is no proliferation hazard associated with spent fuel, but simply the recognition that to increase the amount of highly radioactive spent fuel from 1000 ton Pu content by 10% to a total of 1100 tons of Pu would not significantly increase the chances of proliferation. However, to have 100 tons of weapon Pu in metal or oxide form would pose a substantial additional hazard; in addition, it would indicate to NNWS that the Pu might be rebuilt into weapons at some early time-- no matter how unlikely that seems.
The US is currently dismantling some 1500-1800 nuclear warheads each year, with the metal "pits" containing the nominal 4 kg of Pu being put individually in a steel storage container and the "secondary" uranium parts returned to Oak Ridge, Tennessee. The pits are stored in concrete igloos at Pantex (Amarillo), Texas.
The management and disposition of excess Russian W-Pu is making slow progress, with construction begun on a large storage site for such weapon pits and other materials. Russian technical analyses of the burning of W-Pu as MOX in LWRs show feasibility, but enthusiasm is directed more toward the building of several BN-800 sodium-cooled fast reactors (which would in any case probably be fed civil plutonium first). Some Russians claim that the fast reactors are no more expensive than the VVER-1000 LWRs when the latter are upgraded to modern safety standards, but primary data are not available to consider this claim.
While considering the "energy value" of Pu, it is easy to ignore the cost of releasing this energy-- particularly the cost of MOX fabrication, which at current costs with zero-cost Pu metal exceeds (12) the total cost to purchase equivalent LEU fuel elements. Although vast sums and human capital were invested in the acquisition of the military Pu, these additional costs over LEU reduce the economic value of Pu to zero; nevertheless, the energy from fission of a gram of Pu is just that available from U235, so it should be possible to arrange an exchange of excess W-Pu for an equivalent amount of HEU blended to the 19.9% level. Indeed, such a swap would be highly advantageous to the initial holder of W-Pu, since the blended HEU would be much cheaper to use and hence more salable.
The option for earliest disposition of excess W-Pu still seems to be the conversion of VVER-1000 reactors to take full core MOX. Russia lacks an adequate MOX fabrication plant, but the Siemens-built plant at Hanau, which has not been allowed to begin operation, could be used as the source of equipment for a high quality plant within Russia for the fabriction of LWR MOX fuel elements-- the so-called Hanau East option.
One principal impediment to the completion of a CTBT is the ongoing Chinese nuclear test program. France conducted nuclear tests while negotiating, but stated that its series would terminate by May 1996 with 8 tests or fewer, (14) and China should give a date certain by which time their tests will be finished, preferably in 1996, and commit to obeying a CTBT after that time.
The most serious problem, perhaps is that represented by the statement of the Chinese disarmament ambassador Sha Zukang (15) formally to the 38-member UN negotiating Conference on Disarmament that a CTBT should permit peaceful nuclear explosions (PNEs). Even though peaceful nuclear explosives may differ considerably from military explosives (in needing to have a low fission fraction, or a low tritium content, or to be able to withstand high temperatures and pressures in a borehole), the point is that such charges would still be useful military explosives. A nuclear-weapon state (and perhaps many non-nuclear states) will not tolerate one nation going ahead unimpeded with its design efforts; and if all proceed with imaginative PNE designs such a program will prevent the achievement of a meaningful CTBT.
On the other hand, if the aim is simply to avoid impediments to inertial confinement fusion, induced by laser beams or particle beams, this can both be understood and accommodated. The heat required by a normal power plant corresponds to the energy of about one ton of HE per second, so an ICF pellet explosions every ten seconds would need to give ten tons of HE equivalent. The US would certainly not count this as a forbidden "nuclear explosion" even though it is 10,000 times as powerful as fission explosions that will clearly be banned under the CTBT. So an exemption is clearly in order in a CTBT for pure fusion energy release (without any fission component) taking place in a permanent, reusable containment above ground, and properly monitored.
As for the other uses to which peaceful nuclear explosions (as traditionally understood) were put in the 123-test Soviet program and the 27-test US program-- excavation, stimulation of oil and gas, creation of underground reservoirs, deep seismic sounding, and quelling runaway gas or oil wells, none of these has benefits that would repay its proliferation damage. Indeed, in a published interview a year ago, Victor Mihailov, head of the Russian MINATOM concluded about the Soviet and American PNE program, "So far, they have not proven to be economical." Still, Mikhailov is reluctant to ban any tool of scientific progress.
Humanity may need PNEs in the long term, and a CTBT now
will not stand in the way.
For instance, enormous nuclear explosions emplaced by
high-performance rockets are one approach to diverting asteroids or
comets from striking the Earth;
such an impact was responsible for the
extinction not only of the dinosaurs some 65 million years ago
but of 70% of all species on Earth at the time.
However, we should not start now to prepare for a disaster that
will happen with good probability in tens of millions of years and
with smaller magnitude of damage perhaps in 10,000 years.
The reasonable solution would seem to be to sign a
CTBT that bans any explosive energy release in tests from fission,
and any unconfined pure fusion energy release, but includes a provision to
call a well-prepared international conference on PNEs ten years
after the CTBT enters into force, with a view to understanding not only the
costs and problems of PNEs,
but also whether and how they might be carried out without
compromising the security goals of a CTBT.
The further reduction of 6000 warheads need not depend on an expansion in the capacity for dismantlement; the fissile material in the warheads themselves could be designated irrevocably for civil use, declared and monitored under the agreement; the weapons could be demilitarized, even before being disassembled.
One example of the scope of the problem is given by project SAPPHIRE, in which the US agreed with Kazakhstan to buy and transport to the United States some 600 kg of HEU. At the $12 B agreed with Russia for the transfer of 500 tons of HEU, this would be worth about $14 M. The purpose of the transfer, of course, was the inadequate protection offered this material, which could have provided the fissile core for some 20-30 nuclear weapons.
The US Department of Energy has requested $95 M from the US Congress to fund an MPC&A plan on which agreement has been reached with Russia to upgrade security and accounting systems that cover more than 70% of all the locations in the former Soviet Union where weapons-usable plutonium and HEU are located, but these security and accounting upgrades would not be completed until the year 2002. Funding at a level of $150 M per year would permit more rapid reduction in these highest priority proliferation risks.
But it would be good to have other nations looking at themselves and their neighbors (and helping with the states of the former Soviet Union) to increase the international standards for security and accounting of weapons-usable materials worldwide, in view of the serious proliferation risks that these pose. To the extent that these standards are not equivalent to those for restricting access to nuclear weapons, they should be brought to that level.
IAEA must adopt a consistent analysis to the health effects of Chernobyl. Dr. Rosen estimated 2500 initial premature deaths from late cancer among the 3.7 million people in the contaminated zone, (16) emphasizing that the additional radiation from Chernobyl was less than the natural background in France and Spaie. Exactly for this reason, the concept of committed population dose that leads to the estimate of 2500 deaths among these 3.7 million people leads to an estimate of 58,000 late cancer deaths among the much larger population outside the contaminated zone (see Table 1).
Table 1: Estimated deaths due to committed population dose: M. Rosen 3,700,000 x 7 mSv --> 26,000 person-Sv 2500(1) pop. dose pop. dose add'l deaths UNSCEAR: EC + USSR 600,000(2) p-Sv 58,000(3) (30,000) Each on top of 200 times as many natural deaths due to cancer. ---------------- 1 M. Rosen, op. cit. Evidently he took 0.096 late cancer deaths per person-Sievert. 2 02/17/95 "The Perception of Radiation Effects on Humans: The Case of Chernobyl", by A.J. Gonzalez, NATO/AAAS, Atlanta, GA. (February, 1995). 3 (Using same 0.096 per person-Sv as selected by M. Rosen.)
(1) See Special Panel Report of the American Nuclear Society "Protection and Management of Plutonium" (August 1995) page 53. Citing data from the OECD fuel cycle study, the ANS report notes that reprocessing and recycle "R/R" would cost less than the once-through "OT" cycle if the cost of U3O8 rose to $80 per kg; however, uranium price has remained in the range below $25/kg for the last 5 years.
(2) There is legitimate profit to be made in the safe disposal of waste, but only if governmental subsidy, open or concealed, is introduced at some point in the process. Rather than resort to open subsidy, governments often resort to regulation-- thus shifting the cost to the industry and hence to the consumer of the product.
(3) "Management and Disposition of Excess Weapons Plutonium," Report of the National Academy of Sciences, Committee on International Security and Arms Control (January 1994), p. 68. ("CISAC-1")
(4) I.e., one MW(t)d. See "Management and Disposition of Excess Weapons Plutonium: Reactor-Related Options," Report of the National Academy of Sciences, Committee on International Security and Arms Control, Panel on Reactor-Related Options for Disposition of Excess Weapons Plutonium, (July 1995), p. 33. ("CISAC-2")
(5) Specifically, the industrialized nations use some 8 kW(t) of primary energy per person, so 17 MW(t) is the share of about 2000 persons.
(6) One "SWU" is a "separative work unit" suitable for measuring the capacity of isotope separation facilities such as centrifuge or gaseous diffusion plants. For general formula, see standard references or CISAC-2, p. 288.
(7) CISAC-1, pp. 32-33.
(8) Op. cit.
(9) For instance, October 24, 1995 White Paper on Atomic Energy published by the Atomic Energy Commission of Japan lists 126 kg of plutonium oxide at the PNC Reprocessing Plant, 2032 kg at the PNC Facility for Plutonium Fuel Fabrication, 1412 kg in Britain, and 7309 kg in France. Another 2000 kg (approximately) is listed in various other forms and facilities. (Table in Plutonium, No. 11, Autumn 1995, p. 15).
(10) "Storage and Disposition of Weapons-Usable Fissile Materials Draft Programmatic Environmental Impact Statement" (February 1996).
(11) "The current-reactor/spent-fuel and vitrification-with-wastes options are the two leading contenders for plutonium disposition to the spent fuel standard. Because it is crucial that at least one of these options succeed, because time is of the essence, and because the costs of pursuing both in parallel are modest in relation to the security stakes, the panel recommends that project-oriented activities be initiated on both options, in parallel, at once." (In CISAC-2, p. 14).
(12) CISAC-2, pp. 292-298.
(13) Nobel Prize Acceptance Spech, "Remember your humanity" Oslo (December 10, 1995)
(14) In the event, it terminated February 1996, with 6 tests.
(15) "The Chinese delegation maintains that as an important principle, any disarmament or arms control treaty should not hinder the development and application of science and technology for peaceful purposes. Therefore it would be incorrect if CTBT should ban PNEs." (Reuters, March 28, 1996).
(16) This symposium page i-1-3 "For the 3.7 million residents of other contaminated areas the predicted lifetime excess is 2500 over the normal 430,000."
(17) "Nuclear Electricity Generation Using Seawater Uranium," by Toru Hiraoka in Atoms in Japan, pp. 14-16, Dec. 1994.
(18) But note the caution, "Studies that offer estimates of costs without providing sufficient detail about the derivation of these to permit such a procedure (that is, providing the estimates of direct construction costs, or of labor or materials requirements) are not useful for the purpose of making systematic comparisons." (In CISAC-2, p. 75). CISAC-2 itself goes to great efforts to present primary data and assumptions, as well as the procedure for deriving cost totals from time-spaced expenditures.