(W)Archives: Industrial Hubris

October 17, 2014

Today the United States often faces questions about when adversaries will develop particular military capabilities. When will Iran get the bomb? When will North Korea develop a true ballistic missile capability? How long until China can project power outside of its current regional limitations?

As the Second World War came to a close, one of the most pressing national security questions was when the Soviet Union would develop their own nuclear weapons. Estimates in response to this question were varied: U.S. intelligence agencies estimated that the most probable date for a Soviet bomb was 1953; Manhattan Project director Leslie Groves predicted 20 years; and President Harry Truman thought they would “never” match the American accomplishment. When told of the Soviet bomb test in 1949, Truman famously proclaimed his confusion at how “those Asiatics” had accomplished such a complicated feat.

Groves’ estimate was based on his belief in the inability of the Soviet Union to provide Soviet science with the industrial support necessary to build an atomic bomb. He understood better than anyone the immense industrial and intellectual effort that went into building the American atomic weapons. Furthermore, he was told by representatives of American industry that the Soviet Union did not have the available infrastructure to provide for such a massive undertaking as an atomic bomb project.

As early as May 1945, Groves had begun to speak to the leadership of American industry in order to gauge Soviet industrial capabilities. A document at the National Archives gives us insight into what these captains of industry told him. For instance, on May 21, Groves called a Mr. G. W. Read, a senior official of the DuPont Company, and asked him how long it would take the Soviets to build a facility like the Hanford, Washington installation that provided the Manhattan Project with fissile material. Read told Groves that the experience of DuPont in building an ammonia plant in Moscow demonstrated that Soviet skilled labor and mechanics “were so poor that they allowed machinery to pound itself to death.” The company also had difficulty finding capable men to run the plant once it was constructed, and he believed that the Soviets would not have enough men to build, or run, an effective uranium separation facility. Even if they were given the exact plans and blueprints for the American plant, it would take the Soviet Union so long to reproduce what the Americans had accomplished that Read argued that the Soviets would not “live long enough to build one of these things.”

In the summer of 1945, the U.S. political, military, and scientific leaders who would shape post-war nuclear weapons policy met periodically as an “Interim Committee” to discuss pressing issues. On June 1, Groves attended one such meeting in which prominent American industrialists were invited to speak on the issue of Soviet industry. According to the minutes, available at the Truman Presidential Library, the industrialists included Walter Carpenter, the president of DuPont, James White, the president of Tennessee Eastman, George Bucher, the president of Westinghouse, and James Rafferty, the vice president of Union Carbide. All were dismissive of the capabilities of the backward Soviets.

Carpenter explained that DuPont had to work with 10,000 to 15,000 other companies in order to complete Hanford on time, and without such help it would have taken significantly longer. He estimated that it would take the Soviet Union “at least four or five years” to construct this type of facility, and this assumed they already had the basic plans for the plant. He believed that the Soviet’s greatest difficulty would be in securing the necessary skilled labor, technicians, and adequate production facilities.

For his part, White stressed the advantage that the United States had over the Soviet Union in standardized mass production capabilities. Special ceramics, vacuum tubes, special stainless steels, and “a great variety of special products” were needed in his Tennessee plant, and he doubted whether the Soviets “would be able to secure sufficient precision in its equipment to make this operation possible.” He also echoed Carpenter’s assertion that the Soviets would have difficulty finding enough skilled and educated personnel to catch up with the Americans.

Bucher estimated that if the Soviets were able to utilize captured German technicians and scientists then they might be able to produce an electromagnetic pilot plant in as little as nine months, but it would take at least three years before the plant would be fully operational. He pointed out that the major problem the Soviets would need to overcome was that this type of plant required large numbers of replacement parts and “extremely accurate precision tools.”

Rafferty of Union Carbide told the committee about the process employed at the gas diffusion plant at Oak Ridge, Tennessee. Metal uranium was converted to a gas and then U235 was separated from U238 by means of “extremely delicate barriers or screens.” The barriers are the key to the entire process, and the Oak Ridge plant used over five million of them to produce fissionable material for atomic bombs. Rafferty estimated that the Soviets would require at least ten years to build this kind of plant without the basic knowledge only the Americans possessed. It would take the Soviets five years simply to develop the barrier itself, he believed.

As each industrialist testified, Groves became more and more confident of his estimate. Three years later, in a June 1948 issue of the Saturday Evening Post, Groves explained his rationale to the American public. He wrote that the Soviet Union would be incapable of building an atomic bomb in less than a decade even if the United States had sent the “complete blueprints of the Manhattan Project to Russia on V-J Day.” He emphasized the extent of the industrial effort, noting that the gaseous diffusion plant at Oak Ridge required 12,000 construction drawings. Blueprints for the rest of the Oak Ridge project would cover approximately 500 acres if they were spread out on the ground, and combined with the gaseous diffusion plans, they would weigh more than 230 tons. This does not even include all the plans that would be necessary for Soviet duplication of Hanford or Los Alamos, nor for the tens of thousands of special-design drawings made by American industrial firms throughout the United States.

Of course, Groves would be proven wrong just a little over a year later. His mistake was a classic case of mirror-imaging: presupposing the Soviets would do things exactly the way the United States did. The Soviet Union – and Soviet intelligence – had other ideas.


Dr. Vince Houghton is the Historian and Curator of the International Spy Museum. He has a PhD in Diplomatic and Military History from the University of Maryland, where his research centered on U.S. scientific and technological intelligence (nuclear intelligence) in the Second World War and early Cold War.