Description of Transmutation Process

Briefly and roughly stated, the transmutation process is understood to proceed at follows: Uranium normally and continually emits atomic particles called neutrons, traveling at high velocity. In the proposed transmutation process, uranium is surrounded by graphite, so that the neutrons issuing from the uranium must pass through graphite. By their passage through the graphite, the velocity of the neutrons is so reduced that, upon encountering more uranium, the neutrons are absorbed by it with the production of (a) the new chemical element, Plutonium, (b) highly radioactive by-product chemical elements, (c) neutrons slightly greater in number than the neutrons absorbed, and (d) large quantities of heat (about 11,000,000 kilowatt-hours per pound of Plutonium formed). There is thus a multiplication of neutrons and, provided the escape of neutrons from the system is sufficiently small, there will be a continually repeated chain of reactions: neutron emission - neutron slowing - neutron absorption - Plutonium formation and neutron emission - and so on, accompanied by the evolution of heat and radioactivity. The rate at which heat can be removed determines the rate at which the reaction may be permitted to proceed. The multiplication of the neutrons, and hence the chain reaction, can be controlled or stopped by inserting into the mass of uranium and graphite, materials such as boron or cadmium, which absorb neutrons without emitting more neutrons. These materials are used in the form of rods which, upon withdrawal from the uranium-graphite mass, will permit the neutron chain reaction to increase in rate, or to start again if it has been stopped. For protection of personnel against radioactivity, the uranium-graphite mass must be surrounded by heavy shielding (about 5 feet of iron, steel and pressed wood); must be controlled from outside the shielding; and the treated uranium must be discharged from it within an additional shield (about 5 feet of concrete) by indirect operation.

The treated uranium discharged from the transmutation unit after three months’ operation at rated capacity is expected to contain, on the average, about twenty-five hundred-thousandths (0.00025) parts by weight of Plutonium, together with about an equal quantity of highly radioactive chemical by-products. The uranium containing Plutonium must be handled mechanically by indirect means and surrounded by shielding (about one foot of lead) to protect against its radioactivity, and will spontaneously evolve sufficient heat to require constant cooling during handling and during much of the necessary two months’ storage (under 16 feet of water for shielding) before it can be treated further. The material will then be dissolved in acid and the Plutonium will be separated from the uranium and radioactive by-products by a series of chemical precipitations and re-solutions carried out, because of the radioactivity, by remote control behind shielding (about 7 feet of concrete) and without the opportunity for direct observation of the operation.

Returning to the process carried out in the transmutation unit, usually referred to as a "pile": Neutrons continually escape from the outer portions of the mass of uranium and graphite. Too great a loss of neutrons in this way would prevent the neutron chain reaction. Consequently, the pile must be made large enough so that its surface is small in relation to its volume; i.e., the transmutation process will "go" only in large piles. Each large pile requires large quantities of very pure graphite (about 1500 tons) and of very pure metallic uranium (about 200 tons). Since only limited quantities of uranium are available, the piles to be designed, constructed and operated must be limited in number. Therefore, in order that the desired quantity of Plutonium may be made promptly, each pile must be designed for, and operated at, a high capacity; i.e., at a high rate of heat removal. It is proposed to build and operate three piles, each rated at 250,000 kilowatts. The 250,000 kilowatt capacity is so great that the design of the proposed piles is thereby made slow and difficult; and the operability of the piles is doubtful, particularly their operability at full rated capacity. Nevertheless, it is believed that the proposed pile capacity represents the best choice when consideration is given to the factors of availability of construction and other materials, time for construction, cost, safety, probability of successful operation, estimated output, and urgency of need of Plutonium for military purposes.

The very large amount of heat produced along with the Plutonium must be dissipated in the face of severe limitation of the means to be used for heat transfer and heat removal. The sizes, shapes and space-relationships of the graphite and uranium components in each pile have been shown by laboratory tests to be critical to the success of the transmutation process, so that they must be adhered to closely, regardless of their effect on heat removal and other operational requirements. Only minimum amounts of other materials than uranium and graphite may be built into the pile to serve for uranium-handling and heat-removal purposes; and of the very few having a sufficiently small effect on the neutron reaction, only aluminum and magnesium are available for use in sufficient quantity and suited for use as structural materials. Choice of cooling agent to be used is limited by the fact that many otherwise suitable cooling agents would be destroyed by the transmutation operation. The cooling agent must be available in sufficient quantity and efficient when applied under conditions which do not too greatly interfere with the neutron reaction. The chosen cooling agent is water, filtered, further purified and refrigerated if necessary, forced through the pile, and then discarded. Recirculation of cooling water is prevented by the fact that the water would become increasingly corrosive upon recirculation, because of the effect of radioactivity upon it during its passage through the pile.

Each pile will be an approximate cube (about 28 feet long, 36 feet high, and 36 feet wide) built of graphite blocks about four inches square and of varying lengths. The pile will stand on a deep concrete foundation and be otherwise totally and closely surrounded by a radiation-absorbing shield about five feet thick, made of iron, steel and pressed wood. Horizontally through the pile and the end walls of the shield will pass about fifteen hundred (1500) aluminum tubes, spaced uniformly on about eight-inch centers. Each tube will be of about one and sixty-one one-hundredths (1.61) inches inside diameter and about six hundredths (0.06) of an inch wall thickness. Each tube will carry throughout most of its length, cylinders of uranium about one and thirty-eight one-hundredths (1.38) inches in diameter and eight inches long. Ribs projecting from the internal surface of the tube will support the cylinders centrally in the cross-section of the tube. Each cylinder will be sheathed in a tight jacket of aluminum about three hundredths (0.03) of an inch thick. In the narrow annular space about eighty-five thousandths (0.085) of an inch wide, between each aluminum tube and the aluminum-sheathed uranium cylinder within it, will flow the cooling water. The uranium must be sheathed in aluminum as described in order to protect it from chemical action by the cooling water, and in order to prevent the passage of radioactive materials from the uranium to the water.

To avoid undue interference with the neutron reaction, only a minimum amount of cooling water may be present in the pile at any one time; yet very large amounts of water (about 30,000 gallons per minute per pile at the proposed capacity) must flow through the pile for safe and sufficient heat removal. The flow of these large amounts of water through the pile without more than a very small fraction (about 400 gallons) of the total water flow being in the pile at any time, requires the use of the very small water channels in the pile and consequently, of very high water velocities through these channels. Even if the water is purified by the best feasible means, it may corrode away the aluminum sheaths or tubes, or impurities or corrosion products deposited from the water may partially or wholly block the narrow water channels, necessitating operation at a lower output or even shutdown and possible abandonment of a pile.

The "radioactivity" which accompanies the formation of the Plutonium is so great that the energy of radiation within each pile will be equivalent to the energy of radiation from about 1700 tons of radium, or over 1,000,000 times as much radium as is known to have been produced since the discovery of radium. The existence of this enormous amount of radiation during normal pile operation requires the beforementioned complete, heavy shielding (about 5 feet of iron, steel and pressed wood) to absorb the radiation to protect operating personnel. Without such shielding, exposure to the radiation from the operating pile for even a few seconds, would be fatal. Insufficient information is available to provide protection with complete assurance of adequacy, and provision of protection calculated to be adequate complicates and limits the design, construction and operating practicability of the pile greatly. Since the radiation cannot be detected by the senses, injury to operating personnel can be minimized only by elaborate precaution and careful surveillance. The presence of large amounts of highly radioactive material represents an ever-present hazard to personnel in surrounding areas should control of the process fail or be interrupted.

While many experiments have been made concerning the effects of extreme radiation on the properties of the materials of construction and of the cooling agent to be used, there remain uncertainties and hazards. Even the stability of the graphite under intense radiation is questionable. Assuming that too great "normal" corrosion of aluminum sheaths and tubes can be prevented by never allowing the cooling water to attain too high a temperature, under the influence of the radiation the rate of corrosion may nevertheless become so great as to make operation of the piles, especially at rated output, impracticable for even short periods of time. The large amounts of effluent water are calculated to be safe for discharge to the large river adjacent to the location of the piles, but experiments being made to attempt to prove this point in advance of actual plant operation, may be inconclusive. Large retention reservoirs to permit decay of radioactivity before return of water to the river, are planned.

Not only does the overall intensity of radiation which will be experienced exceed manyfold that which has been experienced previously, but the nature and distribution of the radiation present new problems. In addition to radiation of the types (beta and gamma rays) previously experienced in relatively small amounts, neutrons will be encountered for the first time in enormous concentrations. Information concerning the physiological effects of neutrons and their effects on materials of construction is very incomplete. A considerable amount of radiation may be present in the form of neutrinos, hypothetical particles of which little is known.

Once operated, a pile will not become substantially completely inactive until months have passed, even if all the uranium has been discharged from it. Radioactivity will have been induced in the graphite, aluminum, iron and steel in the pile and shield. This may make it difficult, if not impossible, to modify or repair a pile once it has been operated, except after a long waiting period for decay of the radioactivity. Consequently, even if the design of the pile proves to be adequate and even if no serious accident occurs in operation, a series of minor leakages or breakages may require complete abandonment or shutdown for a period so long as to make the shutdown equivalent to abandonment for present war purposes. The possibility of abandonment can be measured by the fact that the failure of as few as 15 of the most critically-placed of the 1500 tubes in each pile, might make the pile inoperative by preventing the neutron reaction.

As pile operation proceeds there will be an increasing accumulation of Plutonium and of radioactive by-products in the uranium in the pile. It is known that the by-products may affect the neutron chain reaction adversely, counteracting the known accelerating influence of the increasing concentration of Plutonium. It is not known whether the overall effect will be to speed the reaction or to slow it down. Conceivably, the pile may be "poisoned" by the by-products so rapidly as to delay or prevent the building up of the planned concentration of Plutonium in the uranium, or perhaps even to preclude the attainment of a Plutonium concentration sufficient for practicable recovery. To obviate this, it is necessary to provide excess reactivity in the pile, and that can be done only by keeping the amounts of neutron-reaction-inhibiting water and aluminum in the pile very small; so small that the narrow pressure-water channels may block with water impurities or corrosion products, and the thin aluminum tubes and sheaths are fragile and will be permeable after only slight corrosion. Radiation-disintegration of the graphite, blockage of water channels, breakage or leakage of tubes or sheaths, and the radioactive by-product effect on the pile reaction, represent known possible causes for failure to attain Plutonium production as planned.

The heat produced by the operation at full capacity of the three piles to be constructed under the Contract is calculated to be equivalent to about three-fourths of the total authorized power output from the Grand Coulee Dam (1,000,000 kilowatts); presuming that it proves to be possible to operate the Plant at a rate approaching this capacity. For future peacetime plants for Plutonium manufacture, it may be practicable to develop and provide means to convert the heat evolved in the transmutation process into useful power. This cannot be attempted in the Plant to be built under the Contract because neither time nor the necessary technical information is available for the purpose. In fact, the mere removal and dissipation of pile heat, no matter how wastefully, represents one of the major problems of design and operation. Should the supply of cooling water to an operating pile be interrupted for even a few seconds, disaster of catastrophic proportions may immediately result.

In order that the formation of Plutonium may proceed successfully and that catastrophe may be avoided, it is necessary not only that there be an uninterrupted and sufficient overall flow of cooling water to the pile, but also that the flow of water at each point within the pile be adequate to remove the quantity of heat being generated at that point. Distribution of water at high pressure (200 pounds per square inch) to each of the 1500 tubes in the pile, in the amount required by each tube, is a difficult engineering problem requiring complicated piping arrangements. Despite such difficulties, the water flow must be continuous, sufficient, and adequately distributed under all conditions. Two sets of water pumps driven by independent power sources will be installed, and water reservoirs for emergency use (including overhead reservoirs) will be provided; but there remains the danger of stoppage or breakage of water piping.

Water flow must continue even when the pile is shut down in the course of normal operation. As has been described, the rate of the neutron chain reaction will be controlled by moving rods containing cadmium or boron, into and out of the pile. The reaction will stop when enough rods are inserted into the pile to a sufficient depth; and "safety" rods designed for automatic, "instantaneous" (1.5 seconds) insertion will be provided, as well as other safety measures, all to be actuated by stoppage of water flow or other operating accidents. But heat generation will not cease immediately upon application of the controls. The neutron reaction will stop only as the rods absorb neutrons, and a few seconds must pass before neutron absorption is substantially complete. Moreover, after the neutron reaction has ceased, the radioactive materials in the pile will continue to evolve heat at a rate starting at about one-tenth of the heat generation when the pile is operating at full capacity, gradually diminishing, and stopping only when all the uranium is discharged from the pile.

To appreciate the rapidity with which disaster might occur with a pile delivering high power, and the extent of the catastrophe which might follow upon the failure of all of the several safety devices which will be provided: Suppose that through some mischance, the water supply to the pile is interrupted for a few seconds, the control rods cannot be inserted immediately or completely, and all "safety shutdown" devices fail to act or act too slowly. Heat generation in the uranium cylinders will continue. The water surrounding the uranium in the control tubes of the pile will begin to boil in three seconds, and the boiling will travel swiftly in a wave from the central to the outermost tubes. A steam explosion of the pile may result. If this explosion is not too violent yet violent enough, it may save the situation by immediately disrupting the pile just sufficiently to disturb the critical space-relationship of the graphite and uranium and so causing the neutron reaction and the heat generation to cease. If the steam explosion is more violent, it will break open the pile and throw its radioactive contents to a distance dependent upon the exact force of the explosion. The pile will have to be abandoned, the area surrounding it will have to be evacuated, and casualties will have been caused by flying fragments or by exposure to radioactivity from those fragments or from the radioactive dust cloud which will be the factor that will most greatly extend the area affected by the explosion.

If a steam explosion does not occur or is so mild that it does not disrupt the graphite-uranium space-relationship in the pile, nevertheless the water will have started to boil in the control tubes of the pile as explained above. The large volume of gaseous steam formed by the boiling will rapidly push the neutron-absorbing liquid water out of the central tubes, whereupon the rate at which neutrons are produced throughout the pile, and consequently the rate at which heat is liberated throughout the pile, will increase swiftly and enormously, becoming progressively greater as boiling proceeds in a wave from the central tubes to the surrounding tubes so that the liquid water is expelled from more and more tubes. If this process is uninterrupted, i.e., if no steam explosion occurs, the water will be expelled from the outermost tubes of the pile within five seconds after the water has been expelled from the central tubes. Meanwhile heat will be liberated at the rate of many millions of kilowatts. (If all liquid water were instantaneously removed from the pile, heat generation would increase one millionfold in one second.) Meanwhile also, the temperature in the central tubes will have risen rapidly, passing successively through the melting point of the aluminum sheaths and tubes (1200EF.), the melting point of uranium (2600EF.), the boiling point of aluminum (3200EF.), and the boiling point of uranium (7800EF.). The melting of the aluminum and the uranium will not sufficiently disturb the graphite-uranium space-relationship in the pile to stop the neutron reaction and heat generation, if for no other reason than because the molten uranium will not have time to flow away before it reaches its boiling point. Beginning at the center of the pile, the aluminum and uranium will vaporize with explosive violence, literally bursting the pile apart, perhaps even before the water has been forced from the outermost tubes of the pile. The issuing uranium vapor carrying radioactive by-products, will ignite as it comes in contact with the air. The resulting cloud of extremely radioactive uranium oxide smoke will kill or injure persons exposed to it until it has spread over an area so great that sufficient dilution of it by air has been obtained. The affected area will have to be completely evacuated.

One phenomenon can prevent this catastrophe: Sufficient decrease in rate of the neutron reaction as the temperature rises, so that the reaction and heat generation level off before the temperature in the pile has reached a point such that catastrophe will result. It is known that the rate of the neutron reaction will be effected as has been described, by displacement of water from the pile. It is not known with finality and accuracy whether the neutron reaction will be slowed or accelerated by temperature rise in uranium in a pile in which the graphite-uranium space-relationships, the amount and position of aluminum, and the relative temperatures of uranium, aluminum and graphite are as they will be in the large-scale pile. Attempts have been made and are continuing, to determine accurately whether temperature rise will slow or accelerate the neutron reaction, and to what extent. Such experiments and measurements are difficult, especially when made outside a relatively small temperature range around the normal operating temperature of the pile (150EF.). As the result of the temperature tests carried out to date, it is believed that rise in temperature slows the neutron reaction to such an extent that the heat generation, and consequently the temperature, will level off at a point below the melting point of uranium (2600EF.). It is conceivable, though improbable, that rise in temperature might be shown to slow the neutron reaction within a temperature range, yet that slowing would shift to acceleration at higher temperatures which would be reached in the pile in spite of the slowed rate of reaching them.

If it be presumed that heat generation will level off at some temperature far below the boiling point of uranium, nevertheless disaster is still possible; inasmuch as the only condition necessary for the formation of the radioactive smoke cloud which is the hazard most to be feared, is that the uranium be exposed to the air at a temperature high enough so that it will burn to uranium oxide at a reasonably rapid rate. Such a temperature is the melting point of aluminum (1200EF.). At that temperature, the thin aluminum sheaths (0.03 of an inch thick) on the uranium cylinders will melt away, exposing the uranium. The uranium will then burn as soon as it comes in contact with air. Air will enter the pile relatively slowly through breaks in the water connections to the pile or through breaks in connections to the air-tight casing on the pile. Such breaks might have been caused by the original accident which interrupted the water flow to the pile; they would be caused by a relatively mild steam explosion after the accident. Air will enter the pile rapidly if the graphite structure itself is opened up. Such disruption of the graphite structure may be caused by steam explosion mild enough not to disturb the graphite-uranium space-relationship to a degree which will stop the neutron reaction, yet strong enough to shake the structure. Or expansion-disruption and air access to the pile may accompany or follow attainment of moderately high temperatures in the uranium, merely because of rapidly-formed large temperature gradients in the pile structure due to the greater heat generation and smaller heat loss in the center of the pile. The burning of even a small fraction of the 200 tons of uranium in the pile is sure to start combustion of the 1500 tons of graphite in the pile.