The first experimental reactor—a graphite cube about 8 feet (2.4 metres) on edge and containing about seven tons of uranium oxide—had been set up at Columbia University in July 1941. By the end of that year, reactor work had been transferred to the University of Chicago, where Arthur Holly Compton and his cryptically named “Metallurgical Laboratory” were considering related problems. On December 2, 1942, the first self-sustaining nuclear chain reaction was carried out under Fermi’s supervision in Chicago Pile No. 1, a reactor that Fermi had constructed in a squash court under the bleachers of Stagg Field, the university’s football stadium. It had now been proven that the controlled release of atomic energy was feasible for the production of power and the manufacture of plutonium.
In February 1943 construction began on a pilot uranium enrichment plant located on the Clinch River in the Tennessee Valley, about 15 miles (about 24 km) west of Knoxville, Tennessee. The Clinton Engineer Works (later known as Oak Ridge) occupied a 70-square-mile (180-square-km) tract of land and came to employ roughly 5,000 technicians and maintenance personnel. For the project’s full-sized reactors, however, a more isolated site would be necessary. Groves had expressed concern about the pilot reactor’s proximity to Knoxville, and the larger reactors would have significantly greater power needs than could be accommodated in the Tennessee Valley.
In January 1943 Groves had selected a 580-square-mile (1,500-square-km) tract in south-central Washington for the project’s plutonium production facilities. The location was desirable for its relative isolation and for the availability, in large quantities, of cooling water from the Columbia River and electric power from the Grand Coulee Dam and Bonneville Dam hydroelectric installations. The creation of what came to be known as the Hanford Engineer Works required a significant displacement of the local population.Residents of the towns of Hanford, Richland, and White Bluffs were given just 90 days to vacate their homes, and the Wanapum Native American people were forced to relocate to Priest Rapids, losing access to their traditional fishing grounds on the Columbia. At its peak in the summer of 1944, the huge complex at Hanford employed more than 50,000 people.
For the final stages of the project, it was necessary to find a location that was even more remote than Hanford for the purposes of both security and safety. A site was chosen by the Manhattan Project’s scientific director, J. Robert Oppenheimer, on an isolated mesa at Los Alamos, New Mexico, 34 miles (55 km) north of Santa Fe. Beginning in April 1943, scientists and engineers began arriving at the Los Alamos Laboratory, as it was then called. Under Oppenheimer’s direction, this team was tasked with developing methods of reducing the fissionable products of the Clinton and Hanford production plants to pure metal and fabricating that metal into the components of a deliverable weapon. The weapon had to be small enough that it could be dropped from a plane and simple enough that it could be fused to detonate at the proper moment in the air above the target. Most of these issues had to be addressed before any significant stores of fissionable material had been produced, so that the first adequate amounts could be used in a functional bomb. At its peak in 1945 more than 5,000 scientists, engineers, technicians, and their families lived at the Los Alamos site.