Chapter 3: First Steps for Weapon Development
In those incredibly busy two months following the planning meeting of 25 June 1942, the military leaders – working closely with project scientists and technicians – energetically set about not only to organize the operational requirements for the Army’s administration of the project but also to carry out the specific steps for development of an atomic weapon. On the twenty-sixth Maj. Gen. Eugene Reybold, chief of the Corps of Engineers, held a briefing with Brig. Gen. Thomas M. Robins, the assistant chief; Col. Leslie R. Groves, the deputy assistant; Col. James C. Marshall, the new district engineer; and Lt. Col. Kenneth D. Nichols, the deputy district engineer. During the session Reybold reviewed some of the immediate problems of the atomic project, placing special emphasis on two that required prompt action: selection and acquisition of a site for atomic production facilities in the Tennessee Valley, and securing a contract with the Stone and Webster Engineering Corporation to serve as architect-engineer manager (AEM).
Securing an Architect-Engineer-Manager
Consistent with Army policy that the industrial operator of a proposed installation should have a strong voice in selection of the specific site, the district engineer gave his first attention to securing a working agreement with Stone and Webster, which was slated to have the chief responsibility for the Tennessee plants.1 Following Colonel Marshall’s orders, Colonel Nichols went to New York on Saturday, 27 June, to visit Stone and Webster President John R. Lotz. Nichols outlined the role projected for the firm and Lotz responded enthusiastically. The following Monday, Lotz and other company officials met with Robins, Groves, Marshall, and Nichols in Washington, D.C. Lotz assured them the AEM job would not seriously interfere with the firm’s work on other important Corps of Engineers contracts and that the firm could meet the strict security requirements of the atomic project. The group then drew up a letter of intent, which Lotz
and Marshall signed, authorizing Stone and Webster to begin work immediately on preliminary investigations and surveys, procurement of supplies, and initiation of design. The following afternoon, Vannevar Bush, director of the Office of Scientific Research and Development (OSRD), gave the Stone and Webster representatives a thorough explanation of the technical processes involved in the atomic project. This completed the preliminary discussions.
Stone and Webster now became the Army’s agent for managing the atomic energy project, charged with overseeing and subcontracting all research and development, procurement, engineering, and construction that fell within the Army’s sphere of responsibility. Company officials established a separate engineering group to operate with the utmost secrecy under the direct control of the firm’s senior engineers. Project leaders had hoped that a single company could perform all AEM tasks; however, by the time Stone and Webster signed the formal contract (backdated to 29 June) several months later, the Army had to seek the assistance of other major firms to share with Stone and Webster the vast and complex job.2
The Stone and Webster agreement required immediate funds. The approved program had allotted $85 million – $54 million for the Army Corps of Engineers and $31 million for the OSRD – but had not indicated the source of this money.3 An effort in early June to obtain this sum from the President’s Emergency Fund was unsuccessful. Marshall’s pressing financial obligations totaled $38 million: $10 million to cover the letter of intent issued to Stone and Webster, $15 million to repay the sum advanced to the OSRD, $6 million for site acquisition, $2 million for the projected Argonne pilot plant, and $5 million for the purchase of materials. He also required “practically unlimited authority,” as he put it, to spend it.4
By 16 July, Marshall was able to arrange for an allotment from the Office of the Chief of Engineers, specifically from the Engineer Service-Army category of available funds. The $15 million for the OSRD had already been provided, $5 million was furnished immediately, and the remainder became available a few weeks later at the time of the formal allocation of the total sum. Marshall also received assurances from the War Department’s budget officer that all restrictions on the use of these funds that could legally be removed had been set aside. These included regulations on establishing title to property, the placing of government contracts, employment in the United States and abroad, rentals and improvements
of property, and several other controls. Colonel Marshall appeared to be well on the way to attaining the fiscal means and independence that the atomic project required.
Securing a Priority Rating
Even as fiscal problems eased, the atomic project encountered serious difficulties on the matter of priorities. In the summer of 1942, competition for critical materials was strong and unremitting as America prepared to halt the worldwide Axis offensive. With these conditions prevailing, Colonel Marshall soon realized that access to the scarce supplies and equipment needed for atomic research, construction, and production might be blocked unless he could secure a high-priority rating for the project.
In the wartime economy, the establishment of priorities for military and civilian demands was the responsibility of the War Production Board (WPB), succinctly characterized by one World War II historian as “the supreme industrial mobilization control agency.”5 The Army and Navy Munitions Board (ANMB) administered the priority system for military and related agencies, theoretically subject to WPB approval, but in 1942 the War Department’s Services of Supply (SOS) gradually began to take over the ANMB’s responsibilities relating to the Army. The SOS, particularly through its staff divisions for requirements and resources that formed the ANMB’s Army Section, controlled and coordinated all War Department procurement activities. SOS officers, moreover, served on WPB committees. For example, Brig. Gen. Lucius D. Clay, the SOS deputy chief of staff for requirements and resources, had an important voice in the establishment of policy and would play a key role in the matter of assigning priorities to the atomic project.
Major programs received ratings of AA-I through AA-4, in decreasing order of precedence, whereas lesser projects received ratings in a more extensive category, the highest designation of which was A-1-a. A special top rating of AAA, reserved for emergencies, could not be assigned to an entire program but was limited to expediting delivery of small quantities of critical items. Although the program approved by President Roosevelt did not mention a specific priority designation for the new project, it did imply that the program should be given a relatively high rating, which was to be balanced against the needs of other critical projects.6 When Colonels Marshall and Nichols met with General Clay on 30 June, they requested only an AA rating – without, apparently, asking for a specific classification within that category. Marshall assured Clay that the project “would issue such lower ratings as were possible whenever we did not need the A.”7 General Clay,
who had known Marshall since their days as West Point classmates, told Marshall and Nichols that all DSM requests would be given prompt attention and the highest preference in processing, and that he personally would take immediate steps to obtain an AA rating for the project and would be available at any time for any specific request.
Despite Clay’s assurances, nearly two weeks passed with no priority rating forthcoming. Finally, on 13 July, following some persistent prodding by Colonel Nichols, the ANMB approved a rating of AA-3 for the atomic project.8 This rating, with which Clay concurred, came as a grave disappointment. It was based, however, on an ANMB directive that limited AA-1 and AA-2 ratings to the most essential and urgently needed weapons and equipment – airplanes, ships, guns, and tanks scheduled for production in 1942. Even AA-3 ratings were reserved for those items of military equipment and construction that constituted an essential part of the 1942 program or were required in 1942 for the 1943 program. Under the circumstances, a rating of AA-3 was the highest the atomic project could have received. Indeed, given the as yet unproved nature of the project, the cautious estimates of how long it might take to produce atomic weapons, and the absence of a specific presidential directive assigning it a high priority, the wonder is that the atomic program fared as well as it did. Because the ANMB was limiting AA-1 and AA-2 ratings, Clay told the protesting Nichols that an AA-3 should be adequate for the atomic project. If difficulties did arise, he promised the project could obtain an AAA priority to pry loose certain critical items. With this assurance, the atomic project leaders had to be satisfied.
The anticipated problems were not long in appearing. Badger and Sons soon reported that the heavy water reconversion work on the Trail plant was coming into competition with its commitments in the synthetic rubber program. Both projects had an AA-3 rating, with the rubber program having first choice on materials and skilled workmen because of its earlier start. By mid-August 1942, Badger officials estimated the Trail plant would probably not go into operation until August 1943, although an AA-1 rating might better this date by at least two or three months. This, however, would cause a delay in the rubber program and, as S-1 Committee Chairman James B. Conant pointed out to Colonel Nichols, it would be bad politics to push for a higher priority at Trail at the expense of such a critical project as synthetic rubber. As a matter of fact, General Clay had already indicated his opposition to such a move. Thus, for the moment, the best policy seemed to be to go ahead at Trail under the AA-3 rating.
Procurement was generally an S-1 Executive Committee responsibility, and only when the OSRD was unable to secure the necessary priorities did it turn to the Army for help. During
July, difficulties in obtaining small but essential quantities of scarce materials held back progress on important experimental work. Two much-needed nickel shipments totaling less than 85 pounds, for example, were threatened with a delay of several months and were only cleared for delivery after two weeks of effort by OSRD members, General Clay, and Maj. Gen. Wilhelm D. Styer, the SOS chief of staff.9
On 30 July, the S-1 Committee raised this problem with Colonels Marshall and Nichols, and the group decided to urge OSRD Director Vannevar Bush to ask WPB Chairman Donald Nelson for a blanket AA-1 priority for all atomic project orders below a value of $1,500 or $2,000, to eliminate bottlenecks without interfering unduly with other wartime programs. The next day Marshall, accompanied by Nichols, went again to see General Clay, making one last attempt to secure the desired rating before going over his head. Clay repeated that the atomic project was entitled to no higher rating than AA-3, except in very few specific instances, and said he would oppose any effort to secure a blanket AA-1 rating. That afternoon, Marshall, Nichols, Conant, and others met with Bush, emerging with an agreement that the OSRD director would confer with the WPB chairman. Receptive to Bush’s proposal, Nelson promised to discuss the matter further with Army Chief of Staff General George C. Marshall, but whether or not he actually did is unclear. In the end, the matter was referred back to the ANMB, which still refused to grant a higher rating but worked out a procedure that eliminated the bottleneck on small orders.
Meanwhile, the priorities situation worsened. Securing materials became progressively more difficult. Steel, for example, would soon be virtually unobtainable with less than an AA-2 rating. Without access to this basic material, the atomic project would come to a standstill. Marshall was already receiving reports of delays in plant construction and, in mid-August, the ANMB questioned continued assignment of even an AA-3 priority to the Trail project. Prompt action by General Clay ended that threat, however.
On 26 August, Marshall, Nichols, and Stone and Webster representatives met with the S-1 Executive Committee, and again priorities were a major topic. Most small orders were now being handled without undue delay, but there was serious general concern about the large-scale procurement soon to be required for the production plants. A limited number of firms had the organization and experience needed to build and operate the major facilities, and they were all heavily engaged on other AA-3 programs for which orders had been placed before atomic project orders. The only way to push ahead of other programs was to get a higher priority. With an AA-1 priority, the electromagnetic separation pilot plant would probably be ready by April instead of August 1943 and earlier completion dates for other plants would also be assured. The effect of achieving this end, however, would be that of delaying the progress of other vital
projects. Clearly a decision was needed, perhaps from the President himself, on the relative importance of the atomic project and other war programs. Either atomic energy should be pushed with a higher priority, or it should remain an experimental project for postwar application, with a lower priority.
As a result of these conclusions, on 29 August Colonel Nichols again called on General Clay. With concurrence from General Styer, Nichols now outlined the status of the atomic energy program and presented the unanimous opinion of its Army and OSRD leaders that a higher priority was necessary. If Clay would indicate exactly what procedure must be taken to secure an AA-1 priority from the ANMB and WPB, Bush would obtain a letter signed by the President and addressed to whomever Clay thought necessary. Clay suggested that a letter go from the Joint Chiefs of Staff to the ANMB and that it simply state that the atomic project should be granted a higher priority. But he himself opposed this course. He did not believe that the presidential approval of 17 June ever implied the granting of an overall AA-1 rating and he was convinced that the project was less important than “tanks and other munitions of war.” Clay would support the AA-3 priority, but nothing higher. In Nichols’s presence, he telephoned Brig. Gen. Theron D. Weaver – director of the SOS Resources Division and, thus, the ANMB’s senior Army representative – and directed that the AA-3 rating assigned to the atomic project should not be questioned.10
On the same day, Vannevar Bush wrote to Harvey Bundy, Stimson’s special assistant who served as the Secretary’s personal agent in scientific affairs. Bush knew that his memorandum would come to Stimson’s attention. He summarized the current status of the atomic energy project and its plans and hopes for the future in relation to the problem of priorities. He emphasized that if the ANMB persisted in its view that Manhattan did not need a higher priority rating, the entire atomic bomb program would be delayed. The time had come, he continued, for weighing the relative importance of the atomic program against other wartime programs with which it might interfere and, on that basis, deciding the best way to expedite its development. “From my own point of view,” he concluded, “faced as I am with the unanimous opinion of a group of men that I consider to be among the greatest scientists in the world, joined by highly competent engineers, I am prepared to recommend that nothing should stand in the way of putting this whole affair through to conclusion, on a reasonable scale, but at the maximum speed possible, even if it does cause moderate interference with other war efforts.”11
Bundy showed Bush’s memorandum to the Secretary a few days later,
but there is no indication that Stimson took any immediate action.12 Meanwhile, Stone and Webster representatives reported that steel companies had reacted negatively to their attempts to place orders and the ANMB warned Manhattan officers that the rating was scarcely sufficient to secure the steel needed for the projected electromagnetic pilot plant.
Stone and Webster experienced a similar response to its efforts to obtain copper required for the Trail project. Capt. Allan C. Johnson, assigned in August to head the project’s liaison office in Washington, D.C., found that WPB and ANMB officials viewed the AA-3 rating as indicating that the atomic bomb program was, as he phrased it, “an unimportant miscellaneous type.”13 On 12 September, Marshall asked the ANMB for an AAA rating for the Trail copper. Three days later, backed by Colonel Groves, the district engineer went to see General Weaver of the ANMB and the following day the board assigned the rating, but only with the understanding that the metal would be drawn from the normal quota of the Corps of Engineers. Unfortunately this delayed other engineer projects, but Marshall had no alternative. His action opened the way for the work at Trail to proceed on schedule.
Despite the victory on copper procurement for Trail, there was universal agreement among those concerned with the atomic energy program that improvement in the whole priorities picture was an absolute necessity if the entire project was not to founder. Groves felt that DSM leaders would be able to justify a higher priority rating only after sites were definitely selected, plans were firmly adopted, and actual construction was under way. He urged Marshall to move ahead on these matters with all possible speed. As chairman of the S-1 Executive Committee, Conant had concluded that nuclear developments had become more important than the highly rated synthetic rubber program and now believed that they should be given preference. Bush, too, saw the immediate need and called for assignment of a higher priority. The problem in mid-September 1942, as Groves later recalled it, was “quite simple.” If atomic energy “was really the most urgent project, it should have the top priority.”14 The solution to this problem was not far off, but it would not come before the atomic project itself had undergone major organizational changes.
Procuring Essential Materials
Certain materials essential to the program had never been in sufficient demand for industrial or commercial use to have been produced in quantities. At the time the Army entered the atomic project, three such materials were urgently required: processed uranium feed material (chemical compounds and metal), highly purified graphite, and heavy water. The Manhattan District had to develop its own sources of supply for these essential materials.
In early 1942, the OSRD S-1 Section’s planning board had located sufficient raw uranium ore in North America to satisfy the anticipated requirements of the project for many months to come. But the means for converting this uranium into the various kinds of feed materials needed for the different methods of producing fissionable materials were almost wholly lacking. While the OSRD had taken some steps to secure these materials, the major task of procurement remained to be carried out by the Army.
The most immediate demand was for processed uranium in the form of metal for the Metallurgical Laboratory. Raw uranium ore is customarily refined either as uranium oxide, commonly termed black oxide, or as uranium salts. The oxide or the salts can be converted into metal by additional processing; however, at the beginning of 1942, this was still complicated and expensive and only a limited quantity was available in the United States – several grams of good quality produced experimentally by the Westinghouse Electric and Manufacturing Company and a few pounds in the form of pyrophoric powder manufactured by Metal Hydrides, Inc., of Beverly, Massachusetts. Both Westinghouse and Metal Hydrides had obtained the black oxide from the Canadian Radium and Uranium Corporation of New York.
Canadian Radium’s source was the mine owned by Eldorado Gold Mines, Ltd., at Great Bear Lake in Northwest Canada (Map 2). Eldorado processed the ore in its refinery at Port Hope, Ontario, and then marketed it in the United States through Canadian Radium. The mine itself had been closed and allowed to fill with water in the summer of 1940, because sufficient ore had been stockpiled to satisfy anticipated demand for five years. The uranium for early atomic research in the United States had come from these stockpiles. When the OSRD placed a sizable order in 1941, it obtained additional equipment and supplies for getting the mine back into operation and, meanwhile, Canadian Radium continued to supply amounts of black oxide refined from the stockpiled ores.15
As deliveries increased during the spring of 1942, project scientists intensified their efforts to develop better methods of purifying the material and transforming it into metal. Experiments at the National Bureau of Standards demonstrated that an ether process, long known, could remove all impurities by a single extraction method, greatly simplifying the conversion of black oxide into uranium dioxide, or brown oxide, the starting point for uranium metal production. Arthur Compton arranged with Edward Mallinckrodt, an old friend who owned the Mallinckrodt Chemical Works in St. Louis, to develop large-scale production of brown oxide, using the ether process. To ensure an adequate supply of uranium oxide, Colonel Nichols directed Stone and Webster to buy 350 tons from Canadian Radium to cover the project’s needs for the year ahead.16
Thanks to these measures, by the fall of 1942 Mallinckrodt’s production of brown oxide from Eldorado’s ore had increased sufficiently to supply the project’s requirements. Mallinckrodt and other chemical firms converted the brown oxide into uranium tetra-fluoride, or green salt, the feed material employed in most uranium metal-making processes. Westinghouse had abandoned a photochemical method in favor of a faster process using green salt and soon was producing at a satisfactory rate. At first Metal Hydrides was less successful, failing to provide a metal of sufficient purity with pyrophoric powder.
Intensive research during the summer at Massachusetts Institute of Technology, Iowa State College, and the Bureau of Standards had developed new and improved metalmaking techniques. Most important was a steel-bomb process for reducing green salt to metal, employing highly purified calcium – and later magnesium – as a reduction agent. By early 1943, using this method, Iowa State had developed a manufacturing program and Metal Hydrides had significantly increased its output. New Army contracts with the Electro Metallurgical Company of Niagara Falls, New York, a subsidiary of Union Carbide and Carbon Corporation, and with E. I. du Pont de Nemours and Company further increased production, and the acute metal shortage was largely relieved by 1944.17
In September 1942, Colonel Marshall placed Capt. John R. Ruhoff in charge of all uranium metal production. Ruhoff had been a chemical engineer at Mallinckrodt, assisting in the uranium oxide program, and when he was inducted into the Army that summer, Manhattan officials had arranged his assignment to the District as its area engineer in St. Louis, with headquarters at the Mallinckrodt firm. Then in October, Marshall formed a Materials Section in the District office to administer the whole feed materials program. He selected Lt. Col. Thomas T. Crenshaw of the California Area Engineers Office to head the section and had Ruhoff transferred to New York to serve as Crenshaw’s assistant.18
Meanwhile, project leaders knew the reopened Eldorado mine would probably not be able to produce and ship ore for at least another year and that stockpiles at the Port Hope refinery were insufficient for the 350 tons of oxide ordered for the project in July. They urgently needed a source that could provide high-grade uranium on short notice. Such a source, in fact, had long been close at hand. Late in 1940, when German seizure of much of Africa appeared likely, Edgar Sengier – head of Union Miniere with whom Alexander Sachs of Wall Street and Harold Urey of Columbia had earlier conferred – had ordered shipment of approximately 1,200 tons of high-grade ore from the Shinkolobwe stockpile in the Congo via Portuguese West Africa to New York. Storing the
ore in a warehouse in Port Richmond on Staten Island, Sengier apparently made no effort to call it to the attention of American government officials until after the United States entered the war. Attending a meeting in Washington, D.C., in March 1942, he mentioned his Staten Island cache to Thomas K. Finletter and Herbert Feis, State Department officials concerned with international economic affairs, but neither state nor defense officials indicated any immediate interest in the ore – why is not entirely clear. Nevertheless, it was soon common knowledge in trade circles that Sengier was interested in selling the ore.19
It was early September 1942, however, before word of the Congo ore reached Manhattan District officials. The Standard Oil Development Company, working on the centrifuge process, had opened negotiations with Sengier for procurement of the uranium oxide it needed. Through Standard Oil, Metallurgical Laboratory staff members learned of the Staten Island ore and sought to purchase additional quantities. Through his Union Miniere outlet in New York, the African Metals Corporation, Sengier had submitted a request to the State Department for a license to ship ore from Port Richmond to Eldorado’s refinery in Ontario, for processing into black oxide. On 7 September, Colonel Nichols received a query from Finletter concerning the request from African Metals – his first inkling of the existence of the Congo ore. Nichols acted promptly; he met with Finletter and Feis at the State Department on 12 September and then dispatched Captain Ruhoff to consult with Stone and Webster in Boston and Sengier in New York, while he himself hurried to California for the meeting of the OSRD S-1 Executive Committee on the thirteenth and fourteenth. The committee recommended that all Sengier’s ore be acquired.20 Thus, at just the time when an acute shortage of uranium threatened to seriously delay the atomic project, the store of rich Congo ore became available to provide most of its wartime requirements.
Graphite, Heavy Water, and Silver
Either highly purified graphite or heavy water to use as a moderator in the atomic pile was essential for the plutonium program and the other work under way at the Metallurgical Laboratory. Ample graphite was already being produced commercially in the United States; the question was one of “purity and priority.” The main quality required in the graphite was low-neutron absorption, which was directly dependent on its purity. Unfortunately, the standard product had too many impurities, particularly boron. Scientists at the National Bureau of Standards traced the boron in commercial graphite to the coke used for its production. By substituting petroleum for coke and altering certain manufacturing techniques, both National Carbon Company and Speer Carbon Company were soon
producing highly purified graphite that absorbed 20 percent fewer neutrons and satisfied the stringent requirements of the Metallurgical Laboratory. With the WPB’s cooperation in arranging the necessary priorities, the OSRD was able to place large orders with these firms, essentially solving the atomic energy program’s graphite problem.21
Heavy water was another matter. Scientific leaders knew that heavy water could not be available in large quantities for many months or even years. Researchers at the Metallurgical Laboratory had directed their primary interest toward developing a uranium-graphite pile, viewing heavy water as an alternate solution should the problems with graphite prove insuperable. Meanwhile, the OSRD moved ahead with its plans for a heavy water plant at Trail (see Map 2), but priority difficulties delayed construction and the plant did not begin operating until June 1943.22
A store of approximately 400 pounds – almost all the heavy water in the world outside of that being produced by the German-controlled Norsk Hydro plant in southern Norway – was in the hands of British scientists. This heavy water had an interesting history. Nuclear research in France by Frederic Joliot-Curie and his collaborators, Hans von Halban and Lew Kowarski, had concentrated on using heavy water as a moderator to achieve a slow-neutron reaction. In March 1940, just before the German attack on Norway, Joliot-Curie had secured about 160 to 165 liters (169 to 174 quarts) of heavy water from Norsk Hydro. Shortly before the fall of Paris in mid-June, he sent von Halban and Kowarski with most of this precious store to England, where, after a hazardous trip, the two men joined the growing team of British and refugee scientists doing atomic research. Work with this stock of heavy water had contributed to the optimistic British reports on a uranium-heavy water system. When the group relocated to Canada at the end of 1942, the heavy water went along.23
The need for large quantities of silver had not been anticipated. At the Army-OSRD meeting on 9 July, Ernest Lawrence of the University of California, Berkeley, pointed out that he needed several thousand tons of copper for magnet coils. Because copper was high on the list of critical materials and might be impossible to obtain, he thought that silver, a good electrical conductor and not on the critical materials list, would do as well. Accordingly, Colonel Nichols
visited Under Secretary of the Treasury Daniel W. Bell to find out if silver would be available. Although Nichols did not provide specific details of the DSM project, Bell appeared receptive. “How much silver do you want?” he asked. “About fifteen thousand tons,” answered Nichols. Visibly startled, Bell exclaimed: “Young man, ... I would have you know that when we talk of silver we speak in terms of ounces.”24
Ounces or tons, that the DSM project would get what it wanted was soon clear. With relatively good speed, considering the need for secrecy and the number of clearances required, the Department of the Treasury, the ANMB, and the WPB approved the necessary arrangements. On 29 August, in a letter to Secretary of the Treasury Henry Morgenthau, drafted jointly by Manhattan and Treasury representatives, Secretary of War Stimson requested the transfer of 175 million fine troy ounces (about 6,000 tons of silver) “to the War Department to be used as a substitute for copper” for an “important project” that was “highly secret.” “At this time,” read Stimson’s letter, “the interests of the Government do not permit my disclosing the nature of the use.”25
The endorsement of the Treasury Department on a second letter that day constituted an agreement between the two agencies for the transfer. It provided that the silver would remain in the United States; would be returned to the Treasury in five years, or sooner if required; would be utilized in government-owned plants essential to the war effort; and would be protected against loss. Subsequent agreements in 1943 and 1944 would raise the quantity involved to roughly 14,700 tons, worth about $304 million.26 Under constant heavy guard, the bars of silver were transferred – after being melted, cast in cylindrical billets, rolled into strips, and finally fabricated into magnet coils. Because the electromagnetic process seemed the most promising in the summer of 1942, this turn of events was indeed encouraging.
Project leaders in the summer of 1942 were well aware that acquisition of suitable sites was as important to the success of the atomic program as obtaining adequate priorities. At the Army-OSRD meeting of 25 June, they had confirmed an earlier decision to build a heavy water plant at the Trail site and approved location of the proposed plutonium pilot plant in the Argonne Forest near Chicago. The Army delayed actual acquisition of a specific area in the Argonne Forest preserve pending receipt of further
information from the Metallurgical Laboratory concerning the size of site needed for the plutonium pilot plant. In early July, Colonel Nichols obtained clarification of the specific Argonne requirements in discussions held in Chicago with Stone and Webster officials and Compton and his staff, opening the way for lease in August of 1,000 acres from Cook County. At the same time, the University of Chicago agreed to provide an additional acre on the campus for future construction of additional laboratory space. To administer the site acquisitions and oversee construction activities, Colonel Marshall established the Chicago Area Engineers Office in August and assigned Capt. James F. Grafton as area engineer.27
For the main production plants, Colonels Marshall and Nichols and representatives of Stone and Webster and the Tennessee Valley Authority (TVA) began a survey of possible sites in the Knoxville area on 1 July.28
(See Map I.) Requisite conditions for the site were a nearby source of a large amount of continuous electric power, enough for a fair-sized city; availability of a very large quantity of water for cooling and processing as well as construction and operating requirements; and proximity to a main line railroad and good access roads, to ensure delivery of heavy construction materials and supplies. Topography, too, was important. An area bounded by natural barriers, such as rivers and hills, would be securer and individual plant sites separated by ridges far safer in case of an explosion, although the slopes of these ridges should be gentle enough for easy construction. The substratum should provide adequate foundation, yet not be so full of rocks as to make excavation unnecessarily difficult and time-consuming. Finally, there should be adequate and suitable space for a town with facilities for housing and serving thousands of workmen and technicians and their families.
The survey and subsequent investigations filled nearly three days, during which Colonel Marshall and his colleagues examined several possible sites. None seemed at first glance exactly right, but one, at least, had possibilities. TVA officials seemed certain that the 150,000-kilowatt power requirement of the plants could be met if Marshall could hasten the delivery of some badly needed heavy-generating equipment. As project priorities were indefinite at this time, Marshall agreed to look into the matter; however, he emphasized that because an entirely suitable site had not been found, he would have to consider an area near Spokane, Washington, where the Bonneville Power Administration might more easily meet his requirements.
Site problems were a key issue at the next Army-OSRD meeting on 9 July. John R. Lotz, head of Stone and Webster, reported that his firm had surveyed the Spokane area and concluded that it lacked sufficient transmission lines to supply the required power. The group reaffirmed in principle its earlier decision for a site in Tennessee. Also, Marshall and the Stone and Webster engineers agreed that half of the 200 square miles previously believed necessary would be adequate, and even a site of this size would not be required were it not for the plutonium plant. The danger of highly radioactive fission products escaping, or even of a nuclear explosion, dictated building this plant 2 to 4 miles from any other installation and an equal distance in from the boundaries of the site.
The 9 July Army-OSRD meeting ended without a decision on a specific Tennessee site or any indication of when one might be made. Nor was there, for that matter, any clear forecast of scientific developments that might help determine the choice. Only a tentative and, as soon became clear, excessively optimistic construction schedule emerged. As Colonel Groves pointed out to Colonel Marshall that afternoon, a general air of vagueness seemed to pervade the whole atomic project, with the starting dates for development of many of its phases still too indefinite. He urged Marshall first to insist upon the prompt and complete programming of all contemplated steps and then to see that this schedule was adhered to as far as possible. An obvious necessity
was a swift decision on the major production site.
Shortly thereafter, Marshall and Stone and Webster officials agreed to try to obtain a site in Tennessee by 10 August so that construction on the project administration building and some housing facilities could begin, even if plant construction could not. Stone and Webster drew up a formal site report on the most promising area, about 12 miles west of Knoxville, and prepared maps indicating the exact tracts of land to be acquired. To avoid having a public highway run through the site, an obvious security hazard, the firm also studied the possibility of relocating Tennessee 61, which then crossed the northern portion of the area. The Ohio River Division of the Engineer Department then prepared an appraisal of the cost of acquiring the approximate 83,000 acres in the area, comprised of land in the Roane, Loudon, Knox, and Anderson Counties of Tennessee. On 30 July, at the next Army-OSRD meeting, Colonel Marshall reviewed the steps taken toward acquisition of the site and the entire group agreed that the Tennessee Valley seemed the best location, although some of the scientists felt that a site farther east in the Great Smoky Mountains, where the climate was not as warm, might prove more desirable in the future for a proposed permanent central laboratory.
Hardly had the way been cleared for immediate acquisition of the Tennessee site when Colonel Marshall, with the approval of General Robins, decided to postpone carrying it out. He knew that the site and making the necessary preconstruction changes and improvements, not including re location of Tennessee 61, would cost an estimated $4.25 million and require resettlement of some 400 families living in the area. Marshall reasoned there could be no harm in delaying acquisition until more definite information on the plutonium process was available. At worst, a postponement would cause only a few weeks delay, for the Engineers’ Real Estate Branch was sure that the land could be acquired to the point of right of entry within ten days of his order to proceed. Meanwhile, he would try to get the TVA the needed priorities and, when scientific developments warranted, order acquisition of the site.
Although Ernest Lawrence indicated he was now willing to have the full-scale electromagnetic separation plant built in Tennessee, locating the plants in the Shasta Dam area of California was seriously studied and the proposal was not completely abandoned until early September. Nevertheless, Colonel Marshall felt he was “about ready to recommend purchase of at least part of the Tennessee site” by 26 August, the next S-1 Executive Committee meeting. When the committee, however, delayed a decision on production facilities, acquisition of the site was postponed, despite the urgings of Robins and Groves to the contrary.29
Reaching Decisions: The Meeting at Bohemian Grove
About 10–12 miles northwest of San Francisco, across the Golden Gate and amidst the giant redwood
trees of the Muir Woods National Monument, there is a beautiful area known as the Bohemian Grove. In this impressive setting, not too far from Lawrence’s laboratory at the University of California, Berkeley, the S-1 Executive Committee met on 13 and 14 September 1942 to consider at length and in detail the major problems of the DSM project.30 Present along with the committee at this fifth Army-OSRD meeting were Colonel Nichols and the California area engineer, Maj. Thomas T. Crenshaw – both in civilian clothes to mask from casual observers the Army’s interest in the work at Berkeley – as well as J. Robert Oppenheimer and two other scientific consultants.
The first major decision was to acquire the Tennessee site immediately. But on which plants could construction begin? The gaseous diffusion and centrifuge separation methods still appeared feasible and promising, but neither had produced any appreciable amounts of U-235 and both would require hundreds or thousands of process stages for large-scale separation. The plutonium process had yet to see a self-sustaining chain reaction, much less production and separation of plutonium. Thus far, only the electromagnetic method had achieved significant production. Because one electromagnetic unit could separate 10 milligrams of U-235 per day, it was not inconceivable that fifty thousand units could separate a pound, and, in the same period, a billion units could separate a ton. To design and build these units would be difficult and expensive, and the full-scale plant would require considerably more research and engineering development as well as the training of large numbers of skilled operators. But the process appeared sufficiently feasible to justify starting work on a production plant. After a visit to Lawrence’s laboratory, where the Executive Committee viewed experimental separation units in actual operation, the group agreed to proceed with the construction of a large-scale electromagnetic plant.
This 100-gram-per-day (the output specified in the 17 June program) electromagnetic installation would be erected in Tennessee at an estimated cost of $30 million. Design and procurement for the plant were to begin immediately, subject to cancellation at any time before New Year’s Day of 1943 if further developments so warranted. On that date, the group hoped, design would be frozen and construction could begin. At the same time, a small electromagnetic pilot plant was projected for Tennessee; however, at a later date, this plan was dropped.
The experimental plutonium plant planned for the Argonne Forest site was now switched to Tennessee. This change was necessitated by growing evidence that operations at this plant – including chemical studies on extracting the plutonium, training of operators, and testing of equipment and processes – would be on a scale too large for the Argonne site. Stone and Webster would arrange a subcontract with a chemical company to develop and operate the chemical engineering equipment needed for plutonium separation.
Now the experimental pile from the Metallurgical Laboratory could be relocated from the heart of south Chicago to the safer Argonne location. In further support of the plutonium project, construction of the heavy water plant at Trail would be pushed as rapidly as necessary to complete this work by 1 May 1943.
The meetings on 13 and 14 September brought an end to much of the indecision that the course of events had imposed on the atomic energy program during the summer of 1942. The decisions reached at the Bohemian Grove, in the words of one participant, “were destined to shape the entire future development of the project.”31 Indeed, even as these decisions were taking form, changes were under way that would have a profound effect on the organization and direction of the atomic bomb program. The early period of Army participation, marked by a slow and deliberate entrance into the project, was coming to an end.