With both the K-25 and Y-12 plants suffering setbacks in the spring of 1944, Oppenheimer urged Groves to approve the construction of a thermal diffusion plant. The U.S. Navy had researched this method for three years and was already building a pilot plant of 100 columns in Philadelphia. After reviewing Oppenheimer’s suggestions, Groves decided in late June 1944 to approve construction for what would become the S-50 Thermal Diffusion Plant.
Closely patterned on the Navy pilot plant in Philadelphia, the S-50 plant consisted of 2,142 uniform columns, each 48 feet high. Manufacturing this plant to exacting specifications within 90-days would be no small feat. Indeed, 21 firms turned down the assignment before the H. K. Ferguson Company, an engineering firm in Cleveland, accepted the challenge.
The construction of the plant demanded a high level of precision. It required nearly perfectly round columns with a uranium hexafluoride layer spacing of only 0.010 inches (3 sheets of paper) thick! In order to meet the nearly impossible deadline, operators, electricians and welders scrambled to complete the project and even used passenger trains to transport construction materials. In the end, the contractors beat the deadline and completed the S-50 plant in just 69 days.
The theory behind investing in a third plant was that the enrichment process might work best if the three plants were used in a series. In practice, this proved to be correct. The uranium product was slightly enriched at S-50 (one to two percent U-235) and this was fed into the K-25 plant. The gaseous diffusion process raised the enrichment to about 20 percent. This was fed into the Y-12 plant for the final enrichment cycle. Through this serial approach, the first atomic bomb received its enriched uranium.
The S-50 production plant required an enormous amount of energy and was shut down in 1946. The K-25 plant was most effective.