Orbiting satellites, whether their purpose is communications, weather observation, as an aid to navigation, the mapping of natural resources, or for space research, all require a reliable source of electrical power, as do the stations at present working on the surface of the moon and on Mars. In the short term this electricity may be available from previously charged storage batteries but for prolonged use it has to be produced during the mission.

Fuel cells have proved to be satisfactory generators of electricity but the necessary fuel gases must be carried on the flight. Solar cells, powered by sunlight, are a widely used source of electricity on space missions but suffer from a number of disadvantages. The sunlight is collected by antennae which must be correctly orientated to receive it, and these are readily damaged by intense heat, by radiation and by meteorites. In addition the cells cannot function in sun shadows, or in remote areas where the sun’s energy is too weak, although the power they produce elsewhere can charge batteries for use in such areas. Nuclear generating systems have none of these disadvantages and two different types have successfully provided electrical power during space missions. Radioisotopic thermoelectric generators (R.T.G.s) have been used by the United States of America on no fewer than 19 space missions since 1961 and have proved to be a useful source of both electricity and heat. Nuclear reactors, the second type of nuclear generating system, have so far been little used by the U.S.A. in their space programme although they may be developed to meet growing demands for much larger power sources.

The two functional parts of a R.T.G. are the heat source, and the converter which turns this heat into electricity. The heat is produced by the natural decay of an unstable radioisotope, and the material generally used for space applications is plutonium dioxide (plutonium-238), which takes 88 years to lose half its radioactivity. Although the thermal output of the fuel is continually dropping, this relatively long half life means that the generator can have an operational life of several years, and as no moving parts are involved high reliability can be expected.

The heat produced is converted into electrical energy by its action on a number of thermocouples—over 300 of them in the newer modular units. In early devices these couples were made from telluride materials but better conversion is now obtained with silicon-germanium, while selenium alloys are being considered for future use.

It is obviously necessary to contain the plutonium-238 during both normal operational and potential accidental conditions. These may include blast-off or launch pad abort, a mission life which could last for several years, possible descent—with associated high temperature heating on re-entry, followed by impact with the earth and exposure to a variety of terrestrial environments before recovery. Clearly, all components and materials must be completely reliable but at the start of the U.S. space programme no single alloy possessing all the necessary properties was available. The concept of multiple layer encapsulation, where one alloy compensates for the shortcomings of another, was therefore adopted. At the same time development of more suitable alloys was undertaken with the aim of producing a single alloy cladding which would fulfil all the stringent requirements of safety, reliability and performance.

In 1961 a R.T.G. was first used to provide 2.7 watts of auxiliary electrical power on a U.S.A. navigational satellite and since that time there has been a steady increase in the power available from such generators. The units placed on Mars in 1976 by Vikings I and II are both powered by two R.T.G.s, each rated at 35 watts and still exceeding this output.