**3. The early years**

The history of RPS began in the early years of the Cold War, when surveillance satellites were a major impetus for the early space race. The Manhattan project and the years leading up to it had yielded a wealth of knowledge on nuclear physics, particularly the radio-decay properties of actinides and other alpha particle-producing materials. The energy released from the radioactive decay of different elements had become well characterized, and it was recognized early on that radioisotopes could provide power for military satellites and other remote applications. An early study by the North American Aviation Corporation had considered radioisotopes for space power. Then a RAND Corporation report in 1949 evaluated options for space power, and concluded that a radioactive cell-mercury vapor system could feasibly supply 500 We (watts-electric) for up to one year. In 1952, RAND issued a report with an extensive discussion on radioisotope power for space applications, which spurred interest in applying the technology on satellites.

Recognizing the viability of nuclear power for reconnaissance satellites, the Department of Defense (DOD) requested in August 1955 that the Atomic Energy Commission (AEC) perform studies and limited experimental work toward developing a nuclear reactor auxiliary power unit for an Air Force satellite system concept. AEC agreed, but wanted to broaden its examination to both radioisotope and reactor heat sources. This marked the beginning of the SNAP program, which was structured into parallel power plant efforts with two corporations. Odd-numbered SNAP projects focused on RPS and were spearheaded by the Martin Company, while even-numbered SNAP projects using reactors were performed by the Atomics International Division of North American Aviation, Inc.

In these early days, efforts focused on dynamic energy conversion. The work of the Martin Company progressed through an early SNAP-1 effort that used the decay heat of Cerium-144 to boil Mercury and drive a small turbine in a Rankine cycle. In early 1954, a new simpler static energy conversion method was conceived by Kenneth Jordan and John Birden of the AEC's Mound Laboratory in Miamisburg, Ohio. Having been frustrated in their efforts to use radioisotope heat sources to generate electricity via steam turbines, these two researchers considered using two metals with markedly different electrical conductivities to generate electricity directly from an applied heat load. This thermoelectric method was patented by Jordan and Birden, and has remained the basis for all RTGs to the present day. In 1958, work began on two thermoelectric demonstration devices at Westinghouse Electric and 3M, while AEC contracts with other companies explored the development of demonstration thermionic units.

The project to develop a generator based on thermoelectric energy conversion was given the designation, SNAP-3. The 3M Company delivered a workable converter to the Martin Company in December 1958. Shortly thereafter, a complete radioisotosope-powered generator was delivered to the AEC as a proof-of-principle device, producing 2.5 We with a half charge of Polonium-210 (Po-210) fuel.

That SNAP-3 actually never flew in space, but it became an invaluable showpiece for RPS and the SNAP program. President Eisenhower, who had been keenly interested in developing nuclear power for U.S. surveillance satellites, was shown this breakthrough device in January 1959, when the SNAP-3 was displayed on his desk in the Oval Office (Fig. 5). Eisenhower used the opportunity to emphasize his view of "peaceful uses" of nuclear technology, and it afforded him an opportunity to issue a challenge to NASA to develop missions that could exploit the device's potential. The SNAP-3 continued its marketing role, and was shown at several foreign capitals as part of the U.S.'s "Atoms for Peace" exhibits.

Fig. 5. SNAP-3 presentation to President Eisenhower.
