All space vehicles will require some source of electrical power for operation of communication equipment, instrumentation, environment controls, and so forth. In addition, vehicles using electrical propulsion systems like ion rockets will have very heavy power requirements.

Current satellites and space probes have relatively low electrical power requirements-of the order of a few watts. Bolder and more sophisticated space missions will lead to larger power needs. For example, a live television broadcast from the Moon, may require kilowatts of power.1 Over the distance from Earth to Mars at close approach, even a low-capacity instrumentation link might easily require hundreds of kilowatts of power.2 The net power needs of men in space vehicles are less clearly defined, but can probably be characterized as "large."3 Electrical propulsion systems will consume power at the rate of millions of watts per pound of thrust.

Power supply requirements cannot be based on average power demands alone. A very important consideration is the peak demand. For example, a radio ranging device may have an average power of only 2 watts, but it may also require a 600-watt peak. Unfortunately most foreseeable systems are severely limited in their ability to supply high drain rates; consequently they must be designed with a continuous capacity nearly equal to the peak demand.

A third important consideration is the voltage required. Voltage demand may be low for motors or high for various electronic applications. Furthermore, alternating current may be required or may be interchangeable with direct current. Transformations of voltage and/or direct to alternating current may be effected, but with a weight penalty.4


The power supply most readily available is the battery, which converts chemical to electrical energy. Table 1 summarizes the ultimate and the currently available performances of a few selected battery systems. The theoretical performance figures refer to cells in which all of the cell material enters completely into the electrochemical reaction.5 These theoretical limits, are, of course, unobtainable in practice because of the necessity for separators, containers, connectors etc. The hydrogen-oxygen (H2-O2) system refers to a fuel cell.

1 Crain, C. M., and R. T. Gabler, Communication in Space Operations, The RAND Corp Paper P-1394. February 24. 1958.

2 Research in Outer Space, Science, vol. 127, No. 3302, April 11, 1958, p. 793.

3 Ingram, W. T., Orientation of Research Needs Associated With Environment of Closed Spaces Proceedings of the American Astronautical Society, Fourth Annual Meeting, January 1958.

4 Smith, J. G., Transistor Inverters Supply 400-Cycle Power to Aircraft and Missiles Aviation Age, vol. 28, No. 6, December 1957 p. 112.

5 Potter, E. C., Electrochemistry, Principles and Applications, Macmillan, 1956.



Hydrogen and oxygen, stored under pressure, take the p]ace of standard electrodes in a battery reaction, and about 60 percent of the heat of combustion is available as electrical energy.6 The figures listed under "Currently available performance" in table 1 refer to long discharge rates (in excess of 24 hours) and normal temperatures.

TABLE 1,-Electrochemical systems
Type of cell Limiting
per pound)
Currently available

per pound
Currently available

per cubic inch

Batteries as prime energy sources do not give really long lifetimes; they do not operate well at low ambient temperatures or under heavy loads. Batteries are best suited to be storage devices to supply peak loads to supplement some other prime source of energy.

Other factors to be considered with regard to chemical batteries are: (1) They are essentially low-voltage devices, a battery pack being limited to about 10 kilovolts by reliability limitations; (2) a high-vacuum environment and some forms of solar radiation may have deleterious effects; and (3) many batteries form gas during charge and have to be vented, which is in conflict with the need for hermetic sealing to eliminate loss of electrolyte in the vacuum environment of space.


Solar energy arrives in the neighborhood of the Earth at the rate of about 1.35 kilowatts per square meter. This energy can be tapped by direct conversion into electricity through the use of solar cells (solar batteries), or collected 7 to heat a working fluid which can then be used to run some sort of engine to deliver electrical energy.

Solar cells are constructed of specially treated silicon wafers, and are very expensive to manufacture. They cost about $100 per watt of power capacity.

Table 2 gives the performance at 80°F. of some current silicon cells produced by the Hoffman Electronics Corp. These cells generally operate at about 9 percent efficiency on the overall solar spectrum. On a weight basis they deliver about 30 watts per pound (bare). Possible hazards to solar cells include the impingement of micrometeorites (which might have an effect similar to sandblasting) and various forms of solar radiation. Experiments on high-speed impact phenomena have not yet clarified the extent of damage to be expected when impacts do occur.8 As for radiation effects, it has

6 Hydrox Fuel Cells-An Electrical Power Source for Civilian and Military Applications Patterson-Moos Research Division of Universal Winding Co.. Jamacia, N. Y.

7 Miller, O. E., J. H. McLeod, and W. T. Sherwood. Thin Sheet Plastic Fresnel Lenses of High Aperture, Journal of the Optical Society of America, vol. 41, No. 11, November 1951, pp.807-815.

8 Huth, J. H., J. S. Thompson, and M. F. VanValkenburg, High Speed Impact Phenomena, Journal of Applied Mechanics, vol. 24, No. 1, March 1957, pp. 65-69.


been estimated that solar cells should survive for many years in a solar environment.9 Surface cooling may be necessary for use of solar cells.

TABLE 2,-Performance of solar cells

(solar power available at the top of the Earth's atmosphere: 135 milliwatts per square centimeter)

Overall dimensions (centimeters) Active area  (square cen timeters)  ¹Voltage (volts) ¹Current (milliamps) ¹Power (milliwatts) ¹Milliwatts per square centimeter
3.195 (diameter)
2.86 (diameter)
1X2 (rectangular)
0.5X2 (rectangular)
¹Output in full sunlight,with a matched load

For satellite applications of solar cells there is a need to store energy for use during periods of darkness. This storage is the sort of application for which batteries are appropriate. As a rough indication of total weight, a combined installation of solar cells and storage batteries can be expected to weigh about 700 pounds per kilowatt of capacity.

There are a number of possible future improvements in solar cells. Of all the energy striking a solar cell, part is effectively used in producing electrical energy, part is reflected (about 50 percent), and part is actually transmitted through the cell, particularly in the lower wavelength end of the spectrum. Therefore one improvement would be to reduce the reflectivity of the cell; another would be to make the cells thinner.10 Another possibility might be to actually concentrate solar energy through a lightweight plastic lens. It would be desirable to develop cells for high-temperature operation.11 Temperature control problems for solar cells have been investigated in the U.S.S.R. for satellite applications. 12

The second possibility for utilizing solar energy is through heating of a working fluid. An example of a possible system of this sort is shown in figure 1. Here a half-silvered inflated mylar plastic sphere (about 1 Mil thick) 8.5 feet in radius, might serve as a collector,13 at a weight cost of only about 8 pounds per 30 kilowatts of collected thermal energy.14 An installation of this size would require roughly 100 square feet of radiator to reject waste heat (assuming a 10-percent overall conversion efficiency). This kind of system is roughly similar to a solar cell system as to weight for a given power capacity. Again meteorite effects present an unknown factor, in this case with respect to puncture of the collecting sphere and/or the radiator. The greater the actual meteorite hazard turns out to be, the thicker and consequently the heavier the radiator will have to become. The development potential of solar energy sources seems good.

9 Proceedings of the 11th Annual Battery Research and Development Conference, Power Sources Division, U. S. Army Signal Corps Engineering.

10 Jackson, E. P., Areas for Improvement of the Semiconductor Solar Energy Converter, presented at the Conference on Solar Energy at the University of Arizona, October 1955.

11 Halsted, R. E., Temperature Consideration In Solar Battery Development, Journal of Applied Physics, vol. 28, No. 10, October 1957. p. 1131.

12 Vavilov, V.S., V. M. Malovetskaya, C. N. Galkin, and A. P. Landsman, Silicon Solar Batteries as Electric Power Sources for Artificial Earth Satellites, Uspekki Fizicheskikh Nauk, vol.63, No.1a, September 1957, pp 123-129.

13 Such a collector is capable of producing considerably higher temperatures than can be utilized by existing materials

14 Ehricke, K. A., The Solar Powered Space Ship, American Rocket Society, paper No. 310-56, June 1956.


sketch of a device utilizing solar energy to heat a working fluid

Fig.1 - Solar-powered alternator unit


Nuclear power sources call be available either with a reactor or through the use of isotopes.

TABLE 3.-Specific power of pure isotopes

Isotope Specific power of pure isotope (watts per gram) Specific power of attainable isotope compound  (watts per gram)  half-life  Source
Strontium 90 
28 years
Fission product.
2.6 years 
- - - - -
138 days
Neutron irradiation 
of Bismuth.

NOTE.- Estimated fission product power from nuclear power industry:


Year: Cumulative installed reactor megawatts Approximate total beta and gamma kilowatts


Table 3 lists the specific powers available from a few selected pure isotopes. These figures refer to the fresh isotope (that is, to a newly produced isotope; as time goes on, the power available from a given quantity of isotopes will progressively decrease). Since it is difficult to obtain isotopes in a pure form, figures are also given for realizable compounds (oxides} . Isotopes are generally quite expensive and may not be realistically suitable for really large-scale operations. However, for satellite power demands, isotopes may offer some interesting possibilities. One direct way of using isotopes is in the form of a battery, through collection of beta particles (electrons) thrown off by an isotope such as strontium 90 (see fig. 2). This will produce a battery yielding several kilovolts, a gain over electrochemical batteries which are essentially low-voltage devices. However, strontium batteries, although they have a very long lifetime, yield a rather low number of watt-hours per pound as illustrated. Also, nothing can be done to alter the rate at which the isotope releases energy. Another possibility might be to use an isotope such as polonium-which is incidentally rather easy to shield, with reference to external radiation hazard-as a heat source and connect thermocouples from this source to a radiator. Such assemblies have yielded about 300 watt-hours per pound over their half-life. (Conventional metallic thermocouples are very inefficient; semiconductor varieties are better, but subject to radiation damage, etc.15 Isotope batteries are expensive; $375,000 for the one illustrated.16

15 Goldsmith, H. J.. Use of Semiconductors in Thermoelectric Generators, Research, May 1955, p. 172

16 Hammer, W. J., Modern Batteries Institute of Radio Engineers Transactions on Component Parts, vol. CP-4, No. 3, September 1957, pp. 86-96.


overhead sketch of proposed prometheum battery

sketch of side view isotope powered battery

Fig. 2 Utilization of isotope power

Figure 3 shows the Elgin-Kidde prometheum battery (originally designed to operate a watch). All of these batteries have very low efficiencies as regards converting isotope decay energy.


drawing illustrating the suggested composition and design of a prometheum battery

Fig. 3-Prometheum battery

Another possibility would be to use a polonium (or cerium) heat source essentially as a boiler. Rotating conversion equipment could then probably provide higher conversion efficiencies, but with possibly less reliability.

In summary, isotope power supplies, although they can probably outperform the electrochemical systems, are still quite limited in a number of respects. They dissipate energy at an unalterable rate determined by the half-life of the isotope (290 days for cerium 144). Consequently they must be designed for the expected peak demand. Also there is the problem of throwing away excess heat at the start.They are expensive, and there will probably never be enough material for any large-scale operation. Also, the hazard is quite high for a number of isotopes. (For example, polonium is a bone seeker.) Furthermore, this hazard is greatest just at launching when the unit is fresh. An isotope power supply using cerium 144, called Snap I, is in development. The Martin Co. is the prime contractor, with Thompson Products as subcontractor for the rotating conversion equipment.17

Nuclear reactors have a criticality requirement which sets a lower limit to weight, and they may require more shielding. However, they do offer essentially an unlimited lifetime and less hazard at takeoff. Reactors are attractive means for generation of large quantities of heat energy for very long times. In the foreseeable future the weight of a large Installation based on a reactor source is likely to lie mostly in the machinery required to convert reactor heat into electricity. A reactor power supply called Snap II is also in development. The prime contractor is the Atomics International Division of North American Aviation. It is to use the same conversion equipment as Snap I.18

17 Outer Space Propulsion by Nuclear Energy hearings before subcommittee of the Joint Committee on Atomic Energy, Congress of the United States, 85th Cong., 2d sess., January 22, 23, and February 6, 1958; Col. J. L. Armstrong p. 122.

18 See footnote 17.


A polonium heat source can be used to heat a thermocouple to generate electricity. Polonium encapsulated in a pellet and attached to one end of a thermocouple will heat the assembly to a temperature of about 1500°F. The unheated end of the thermocouple will radiate waste heat to space. It is estimated that such units in development at the Westinghouse Electric Co. as Snap III will deliver 3 watts for 6 months with a total weight of only 10 pounds. The cost of such a unit is expected to be about $25,000. The efficiency of the conversion from thermal to electrical energy is about 8 percent. No shielding from radiation would be required for most applications.19

19 See footnote 17, p. 53.