PM Files
The Chairman of the United States Atomic Energy Commission (Dean) to the Secretary of State
Dear Mr. Secretary: In reply to your letter of August 22, 1950, the effects of technical developments on the international control of atomic energy, which were referred to in our letter of June 26, 1950, are discussed in more detail below. These matters are considered under four main headings which we believe cover the points raised in your letter.
I. Accumulation of Fissionable Material
Now that the USSR has produced fissionable material, the technical question arises as to how the United States could be sure that the USSR had turned over all fissionable material to an international agency, if such an agency were established in accordance with the United Nations plan for control of atomic energy. The plan states that the international agency would own all nuclear fuel and that nations and persons would be prohibited from producing, possessing, transferring, storing, transporting, handling, or using nuclear fuel, except as authorized [Page 95] by the international agency. The plan, to the extent that it has been worked out in the various reports of the United Nations Atomic Energy Commission, does not specifically consider the questions of when or how stocks of fissionable material would be turned over to the international agency or what safeguards would apply to prevent or detect the withholding of fissionable material, although it does deal with safeguards against diversion of fissionable material after the plan has gone into effect. Presumably, the transfer of fissionable material to the international agency was left to be considered under the subject of “stages” of transition in putting the plan into operation.
Whether there exist safeguards against withholding of fissionable material is a crucial matter for any control plan, as indicated in your letter. One approach to this problem is to inquire whether methods are available for the determination of the total past output of facilities for the production of fissionable material.
With regard to reactors, waste material in the form of radioactive fission products must necessarily be produced in the fuel elements and must be disposed of after extraction of the fissionable material. These fission products would be subject to control by the international agency because of their possible use in radiological warfare. In this country, the fission products are stored in large underground tanks and it would be possible by sampling and analyzing their contents and estimating their volume, to get some idea of the amount of plutonium produced to date and the time when production began. Representative samples and accurate results might be difficult to obtain in this way. In the case of the USSR, if the purpose were to deceive the international agency, the liquid wastes containing fission products might be allowed to run off into the ground without regard to health hazards or, with considerable effort, might be reduced in bulk and scattered, hid, or altered in such a way that significant measurements could not be made. Some of the fission products are gases and escape into the atmosphere during chemical processing of the reactor fuel elements, .... Methods of sampling and analyzing the atmosphere are being investigated. …
In addition to the fission products, radioactive materials are formed in the moderator, shielding, and structural elements of reactors. Some of those are so short-lived that they could not be used to obtain information on the past history of the reactor, but there are others which may be suitable for that purpose. Where graphite is used as a moderator as at Hanford, carbon-14 with a half-life of 5000 years is produced by absorption of neutrons in ordinary carbon. Measurement of the specific activity of samples of graphite taken from various positions in the reactor would give an indication of the total number [Page 96] of neutrons which had been released in the reactor and therefore the total fissionable material which could have been produced from the time operation started. Where heavy water is used as a moderator as at Chalk River, tritium with a half-life of 12 years is produced by absorption of neutrons in deuterium. However, it would be possible to replace the heavy water after a certain period of operation, since it is not a structural part of the reactor. Investigation of long-lived activities produced in such materials as steel, concrete, and aluminum may be required.
With regard to isotope separation plants for the production of U–235, “technical methods for determining total past production do not appear to be nearly as promising as for reactors. The Waste material from isotope separation plants, which contains uranium depleted in U–235, has very little radioactivity and could readily be hidden or otherwise disposed of. The feed material used in the Oak Ridge plant is uranium hexafluoride, which corrodes a number of structural materials. However, the special materials used in the barriers, piping, and pumps and the operational procedures which have been developed make corrosion an inadequate means of determining the age of the plant or the time during which it has been operated. No other specific technical means of obtaining such information from an examination of the plant has been suggested to date, but further study is being given to the matter.
It is assumed that reliance would not have to be placed on the examination of reactors and isotope separation plants alone, but that the whole range of controls and inspections given in the United Nations plan would be used. This would be necessary in any case to assure that all the facilities for the production of fissionable material had in fact been located. In addition, it would provide cross-checks on what the total production had been and, in particular, might give useful clues as to the time when a particular reactor or isotope separation plant was constructed or began operation. Investigations would have to be made of the mines, mills, and dumps from which source material had been obtained in the past, the facilities for processing and purification of source material, the supply of special equipment and materials such as mass spectrometers, diffusion barriers, pure graphite, and heavy water, and the consumption of large quantities of electricity. Records would have to be examined, personnel interviewed, and ground and aerial surveys made. Considerable effort on the part of the USSR would be necessary to falsify records, remove or alter other evidence, and suborn witnesses in an attempt to withhold large quantities of fissionable material from international control.
An attempt might be made by the USSR to cover up the withholding of fissionable material by claiming that the missing material had been [Page 97] consumed in atomic explosions for peaceful purposes or for tests. Methods for detecting atomic explosions would provide a check on such claims, although the quantities of fissionable material used might be difficult to determine in some cases.
With all the controls and inspections contemplated under the United Nations plan, it would appear that quantitative information could be obtained about the total amount of fissionable material produced in the past. There is still the question of whether such information could be made sufficiently accurate. The Second Report of the United Nations Atomic Energy Commission defines dangerous activities or facilities as those “which are of military significance in the production of atomic weapons,” but does not give a quantitative meaning to “military significance,” leaving this to the international agency to decide. Volume VI of “Scientific Information Transmitted to the United Nations Atomic Energy Commission by the United States Representative” includes the following statement:
“It is difficult to define the amount of activity in the illicit production of atomic weapons which is significant. The illicit construction of a single atomic bomb by means of a decade of successful evasion would not provide an overwhelming advantage, if it can be assumed that it would take another decade to produce a second bomb. But the secret production of one bomb per year would create a definite danger, and the secret production of five or more per year would be disastrous. This report assumes arbitrarily that the minimum unit of noncompliance is the secret production of one atomic bomb per year or of a total of five bombs over any period of time.”
This statement was made at a time when atomic bombs were much scarcer than they are now. As the stocks of fissionable material accumulate in various countries, the difficulties which the international agency would encounter in detecting the withholding of fissionable material for a few bombs become greater. At the same time, weapon development is increasing the energy release which can be obtained from a given quantity of fissionable material, a trend which may reach its culmination in thermonuclear weapons.
These considerations emphasize the importance of two aspects of the United Nations plan which were never worked out in detail, the stages of transition to full control and the strategic distribution of fissionable material. Before the crucial stage of turning over fissionable material to the international agency was reached, each country would have to assure itself that no other country was withholding significant quantities of fissionable material. Since there will always be the possibility of error or inaccuracy in this determination, the strategic distribution of the bulk of the fissionable material might have to be made in such a way as to override the effects of possible withholding by one or more countries.
[Page 98]II. Production of Fissionable Material by Particle Accelerators
The possibility of producing fissionable material by means of high-energy, high-current particle accelerators was of course not known at the time the United Nations plan was formulated, but seems to be covered by the general language and definitions of the plan. However, there are a number of technical aspects of the plan which may need to be made more explicit.
Thorium may increase in importance. The First Report of the United Nations Atomic Energy Commission stated that the danger inherent in the diversion of thorium is less immediate than in the case of uranium, because thorium cannot be used by itself for the production of atomic energy. This may no longer be true, since the particle accelerator may provide a means of utilizing thorium alone in the production of U–233. The particle-accelerator development may have the same general effect as a reactor breeding program, in that it may transform U–238 and thorium into fissionable material without being limited by the U–235 content of natural uranium, thus greatly increasing the potential supply of fissionable material. The Specific Proposals of the Second Report made no distinction between the control measures to be applied to thorium and uranium, although it was realized that the control of thorium would be more difficult because of the extent and mode of its occurrence and its use outside of the atomic energy field.
The particle accelerator may require a smaller inventory of source material for a given rate of production of fissionable material than a reactor, but may be comparable as regards size of installation, cooling and shielding requirements, and chemical processing plant. The requirements for electric power and specialized electrical equipment will be much greater than for a reactor. It would therefore appear that the problems of detection and control would be no more difficult than for a reactor. It should be emphasized that the feasibility of a particle accelerator for production of fissionable material in quantity has not yet been demonstrated.
III. Prospects for Atomic Power
At the time the discussions of atomic energy began in the United Nations, there were some persons in this country and elsewhere who were sanguine of the early demonstration both of the practical generation of atomic power and of its economic feasibility. During the course of the discussions, it became apparent that numerous technical and economic problems were involved and that the time scale was a matter of years before atomic power could be demonstrated in a practical way and perhaps decades before atomic power might be utilized on a large [Page 99] scale, if it should prove to be economically feasible. The production of large stocks of fissionable material by the international agency in anticipation of possible future uses in atomic power plants therefore seemed to be unnecessarily dangerous. There resulted the provision which you quoted from Specific Proposal 12, Chapter 4, of the Second Report, on keeping the production of weapon-level material to a minimum. This provision as written and other provisions in the same chapter would apparently permit work to continue on the development of atomic power, including the construction of pilot plants, and would allow power reactors eventually to be constructed by arrangement between the international agency and the nation concerned. Meanwhile, in the absence of international control, there has been a large accumulation of stocks of weapon-level fissionable material, so that the limitation of Specific Proposal 12 has to some extent been nullified.
Various proposals have been made for further limitations on dangerous facilities for the production or utilization of fissionable material, such as a moratorium on atomic power. An important question raised by such proposals is the disposition of the large stocks of fissionable material now in existence, which are more dangerous in the sense of the United Nations plan than development or production facilities. This is also an important problem in connection with the United Nations plan itself, and one that has not been worked out in detail. Its seriousness continues to increase as more and more fissionable material is produced.
The possibility that particle accelerators may be developed for the production of fissionable material from U–238 and thorium without being limited by the U–235 content of natural uranium would have an effect on the prospects for atomic power by making fissionable material more readily available for this purpose. The cost of fissionable material produced in this way would be a factor in determining the economics of atomic power.
IV. Thermonuclear Weapons
Since thermonuclear weapons would be dependent on the availability of fissionable material and reactors or particle accelerators for the production of tritium, the United Nations plan in its control of fissionable material and facilities for its production would also be controlling thermonuclear weapons. This is also in accordance with the general language and the definitions of the United Nations plan. If an opportunity presented itself at some future time, there are several places where explicit reference could be made to thermonuclear weapons and to the light elements. The provisions for the control of heavy water might also be strengthened. At present, the plan only [Page 100] calls for periodic reports to the international agency from nations regarding the production, shipment, location, and use of heavy water. Because of the increasing importance of heavy water in production reactors and because of the possible use of deuterium in thermonuclear weapons, it might be desirable for the international agency to own stocks of these materials and to own, operate, and manage facilities for their production.
One important feature of the increase in energy release which thermonuclear weapons may make possible is that the seriousness of the diversion or withholding of fissionable material together with tritium and deuterium might be intensified and the accuracy required in accounting for such materials might be increased.
Conclusions. After consideration of the above matters, we arrive at the following conclusions:
- a.
- Effective international control of atomic energy on a current and continuing basis after the United Nations plan went into full operation would be technologically feasible.
- b.
- The detection of the withholding of fissionable material produced prior to the implementation of the United Nations plan would be more difficult, and the amount of fissionable material which might be withheld without detection would increase as the accumulation of fissionable material increases. This difficulty stresses the need for (1) stages of transition to full control of such a nature as to provide opportunities for obtaining assurances against the withholding of significant quantities of fissionable material and (2) strategic distribution of fissionable material in such a way as to minimize the effects of possible withholding.
Sincerely yours,