Space reactor systems have been studied since the early days of nuclear power in the late 1950s. However, only one US reactor (SNAP-10A) (Harman and Susnir, 1964) and a few Russian reactors have ever been in space. There is now some renewed interest in nuclear power for space missions, in the US and also Europe.

In general, for space applications, fast reactor gas or liquid metal cooled designs, operating at high temperature are the most appropriate to meet the various requirements and in particular, launch constraints. Clearly also reliability is important and this depends on the status of the possible technologies.

Space applications include, planetary base applications, e. g. for Mars or the moon, nuclear propulsion and radioisotope power systems (RPS). For the former, possible designs include the lithium liquid metal cooled concepts, SP-100 in the US (Sapir et al., 1987) and the ERATO system in France (Carre et al., 1987), these generating power in the range 100-500 kWe. Gas cooled systems include the Sandia National Laboratories Dual Purpose design (Lipinski et al., 1999), and a Russian Project 1172 gas-cooled design (Andreev et al., 2000). A low-power PWR water-cooled system has also been investigated by Technicatome.

For propulsion, many of the reactor concepts under consideration have been developed from other applications. In general many reactor systems that have been developed to supply electrical power, can be employed as a power source in a nuclear electric propulsion (NEP) systems. The SP 100 and ERATO system could be adapted. The UK 200-SNPS was a particle bed system, designed for earth orbit electrical power supply, but could be adapted. The Enabler NERVA (Livingston and Pierce, 1991) was primarily aimed at nuclear thermal propulsion (NTP), where the energy source heats the propellant directly (as opposed to NEP where electrical power from the reactor is used for accelerating the propellant). The Russian TOPAZ-2 liquid metal (NaK) cooled system (Voss et al., 1991) or more advanced TOPAZ concepts could be used. There are also combined cycle (NEP&NTP) nuclear propulsion and other advanced concepts under consideration.

Some of the reactor designs are such that the same generic design can be used for both planetary base and propulsion applications. An example of one such is the ESCORT Derivative reactor (Feller and Joyner, 1999), designed for in-space propulsion and power (25 kWe) and to supply 160 kWe for 10 years on the surface of Mars.

Finally RPS consisting of a nuclear radioisotope heat source and power conversion, have been developed. This technology started in the SNAP programme in the 1950s and culminated in the General Purpose Heat Source (Angelo and Buden, 1985) module flown on the Galileo and Ulysses spacecraft. RPSs typically generate a few kilowatts.

Space nuclear reactor programmes are being supported by the National Aeronautics and Space Administration (NASA) (Nuclear Reactors in Space) and the European Space Agency (ESA) programme. A review of space nuclear power and propulsion for future space exploration is given in (Bond and Sweet, 2003). A particular interest at present is the benefits of nuclear power systems for Mars exploration (Sweet et al., 2002). In particular, work is on-going to examine the feasibility of different reactor systems, including the feasibility of a small gas-cooled, particle bed reactor, to power a Mars mission.