A steam-cycle power plant for high-power communications satellites

Several high-power communications satellites have been proposed which were based on the availability of an electric power supply which could provide 60 kW for one year (minimum) at the cost of 3000 lb. These were the initial specifications of the dual SNAP-8 nuclear powerplant. Since that time, however, it has become clear that the SNAP-8 will be unable to meet these specifications. The purpose of this paper is to present a conceptual design study for a steam-cycle power supply which provides the required specific weight (around 50 lb/kWe) together with reasonable expactation of obtaining the required one-year minimum lifetime.

The proposed cycle is the conventional Rankine steam cycle utilizing superheated turbine-inlet steam at 1200 psi and 1200°F. After removal of residual superheat in a recuperator, the saturated vapor at the turbine exhaust is condensed in a radiator at approximately 400°F and returned by a pump to the energy source.

At first glance, it does not appear possible to reject heat at such a low temperature without enormous radiator weights. Indeed, the radiator surface area is considerably larger than that required for the much higher temperature liquid metal or gas cycles. However, the unique combination of using steam at low temperatures permits the utilization of two design features which provide remarkably low radiator weight per unit area in the conventional flat fin-and-tube configuration. First, the high latent heat of the condensing steam allows only small volumetric through-put per kilowatt output, resulting in very small tuve diameters. Since the tubes and headers are the only major parts of the radiator which require meteoroid armor, the total tube and header weight does not become excessive. Second, because of the low temperature, it is permissible to use aluminum as the radiator material. Its high thermal conductivity therefore permits the use of quite thin, large-area fins between the tubes without suffering the conduction loss necessitated by the higher-temperature materials of other systems. Thus the fraction of radiator are occupied by the heavy, aluminum-armored tubes becomes quite small. In the sample design (30 kW) presented in this paper, the combination of small tubes and large fins results in a weight per unit radiating area of less then 0·25 lb/ft² of radiating area.

The energy source may be either a nuclear reactor, the most favorable configuration of which would be a single-pass coiled-tube design (although more conventional boiler-superheater reactors may be used with little weight penalty), or a solar-powered boiler-superheater. Other cycle components are a conventional turbogenerator, recuperator and pump. The specific weight of the sample (30kW) design is 70lb/kW, but upon scaleup to higher powers, improved turbine efficiency, radiator segmentation, and reduced fractional weight of the energy source and turbine can provide estimated specific weights competitive with systems using other working fluids, with no sacrifice in reliability.

With regard to reliability, the use of the conventional, long-operational steam provides not only the benefit of many years of experience, but also the capability of using conventional materials; i.e. stainless steels and aluminum, for all cycle components. In consequence, although development problems will undoubtedly arise, there appear to be no significant state-of-the-art improvements required. It therefore appears possible with the proposed steam cycle to provide broadcast satellite power requirements in time to phase in with anticipated communications-system availability.