Presented at SPS-97: Space and Electric Power for Humanity, 24-28 August 1997, Montreal, Canada
Published in Proceedings SPS '97 Conference, Montreal, Canada, Canadian Aeronautics and Space Institute, Canada, and the Société des Électriciens et Électroniciens, France, ISBN: 0-920203-18-3, pp. 327-328 (1997).
The classic satellite solar power system, proposed by Glaser in 1968 , locates the power generator in a geosynchronous orbit. This orbit has a number of disadvantages. Orbits higher than geosynchronous, on the other hand, have a surprising number of unexpected advantages.
In this paper the advantages of locating a power satellite in a halo orbit around the Earth-Sun L-2 Lagrange point are discussed. Such a satellite could be uses synergistically with a ground-solar installation, allowing a simple path to development of the system. The Earth-Sun L-2 point is 1.5 million kilometers from the Earth, and remains fixed over the midnight point as the Earth rotates. Halo orbits around the Lagrange point are only weakly unstable, and satellites such as ISEE have demonstrated that such halo orbits can be maintained with a minimum of stationkeeping fuel consumption. A solar power satellite in a E/S L-2 halo orbit would require no rotating joint between the microwave transmitter and the solar array, making a considerable simplification of the design possible. A design is proposed using integral inflatable solar concentrator/microwave antennae.
A Natural Synergy: Ground-based Solar and Space Solar Power
Proponents of SPS occasionally disparage the future use of ground-based solar energy, possibly considering ground-based systems to be a competitor. Nothing could be further from the truth: ground-based and satellite-based solar power systems are natural allies. Ground-based solar power use will not merely fund the development of the PV technology required for space solar power, it will actually create the demand for space solar power by manufacturing a ready-made market that can best be filled by space power systems .
A significant risk element for any satellite power system is the photovoltaic array. This was identified in the NRC review of SPS  as one of the most critical areas where extrapolations of cost and performance were made. Experience with ground-based solar power is a necessary step to mature the required technology, define and trouble-shoot the manufacturing methods, and move photovoltaics down the learning curve to low-cost production. Among the photovoltaic technologies, many different approaches are still in consideration. A significant result of ground-based power will be the competition among various technologies to shake-out the lowest cost approach.
Current photovoltaic module production is about 100 MW(peak)/year. Cumulative production of several tens of Gigawatts is required before photovoltaics reaches the technological maturity required for finalizing a SPS design. The faster the demand for terrestrial PV grows, the more rapid the technology maturation will be.
Many ground sites exist in the U.S. with over 300 clear days per year. In the industrialized northern hemisphere nations, ground-based solar power is viable due to a match between daytime peak requirements and production [4,5]. During summer, when power-requirements are typically highest, the peak power requirement is in the daytime, due to loads imposed by air-conditioning. This added load comes at the same time as daytime workplace loads, leading to a considerably higher demand for power during the daytime than at night. The cost of generating this daytime peak power is much higher than that of generating base power, since the generating capacity is idle at night. Thus, the liability of solar power systems, that they generate power only during the daytime, is comparatively unimportant. Several analyses have shown that for generation fractions of up to about 20% of the current U.S. production, photovoltaic power generation can provide primarily peak power without cutting into baseline power.
Above about 20% penetration of the electric market, photovoltaic generation begins to displace base capacity instead of daytime peaking power. However, 20% of the U.S. power generation capacity is a huge amount of power by conventional economic standards, and large amount of growth in the solar power industry is possible before power systems come near this limit.
When solar-generation capacity reaches a level that saturates demand for peak power, utilities will search for a solar-energy system that provides continuous power. At this point the SPS system should be ready to step in. A SPS system has all the advantages of ground solar, plus an additional advantage: it generates power during cloudy weather and at night. The marketing argument is simple: a SPS receiver looks and operates just like a solar array: like a solar array, it receives power from space and converts it to electricity. The ground locations for the SPS rectennas have already been bought and paid for: SPS rectennas can be placed at the same location as the ground solar facility, since a rectenna can be easily made transparent to sunlight. In fact, it is not hard to imagine that an advanced solar array might even be built with the rectenna elements built right into the solar array, an array designed so that it can convert both ground-solar and space-solar energy.
Ground-based solar is not actually the competitor to space solar power-- it is the natural precursor. Existing solar power satellite concepts, however, are not designed to take advantage of this synergy. Given the synergy between ground solar and space solar, is is possible to design a solar-power satellite to best make use of the synergy? Such a satellite concept would be designed to fill in for ground-based solar for applications where ground based solar is inadequate.
Reinventing the SPS
The classic satellite solar power system, proposed by Glaser in 1968 , locates the power generator in a geosynchronous orbit. This orbit has the advantage of appearing stationary over a given location of the Earth, but has a number of disadvantages. Lower orbits have often been proposed, but have a significant disadvantage of allowing only intermittant view to a given ground site. Orbits higher than geosynchronous, on the other hand, have not been well investigated, but have a surprising number of unexpected advantages.
For a power system which acts to complement solar daytime power with nighttime satellite power, an obvious choice would be to look for an orbit which maximizes the use of the power system on the night side of the Earth; preferably a satellite which "hovers" over the night side of the Earth, and is thus able to feed power to ground installations as they rotate away from the sun. As it happens, there exists a set of orbits with exactly the required parameters: halo orbits around the L-2 Earth-Sun lagrange point.
An alternate possibility, the use of a solar sail to allow the satellite to hover above the night side of the Earth (the "statite" concept), was analyzed, but the required mass to area ratio was beyond the capability of existing materials.
The Earth-Sun Lagrange point, Earth-Sun L-2, is technically known as the Trans-Earth Lagrange point. The E/S L-2 point is 1.5 million kilometers from the Earth, and remains fixed over Earth midnight as the Earth rotates. A satellite would, in fact, be placed into a halo orbit around the L-2 point instead of exactly in the L-2 point. A halo orbit can be chosen to avoid being shadowed by the Earth. There exists a large family of such halo orbits around the Lagrange point, and the spacing is such that literally trillions of satellites could be placed in such orbits with little chance of any intersection of orbits.
Unlike the well-known L-4 and L-5 Lagrange points, halo orbits around the L-2 Lagrange point are, in fact, unstable orbits. However, the instability is an extrordinarily weak one, and both the L-1 and L-2 Earth-sun Lagrange points have been used as locations for scientific satellites. Satellites such as ISEE have demonstrated that such halo orbits can be maintained with a minimum of stationkeeping fuel consumption.
Such a satellite could be used synergistically with a ground-solar installation, allowing a simple path to development of the system.
A solar power satellite in a E/S L-2 halo orbit would require no rotating joint between the microwave transmitter and the solar array, making a considerable simplification of the design possible. Since the Earth and the sun are in the same location in the sky, viewed from the satellite, the same dish could be used both as a solar-power collector and also as the microwave antenna. This would reduce the size and complexity of the satellite by a considerable amount. A design is proposed using integral inflatable solar concentrator/microwave antennae.
 P.E. Glaser, "Power from the Sun: Its Future," Science Vol. 162, pp. 957-961 (1968).
 G. Landis, "An Evolutionary Path to SPS," Space Power, Vol. 9 No. 4, pp. 365-371 (1990).
 U.S. Office of Technology Assessment, Solar Power Satellites, 1981.
 T. Hoff and C. Jennings, "Match Between PG&E's Peak Demand Period and Insolation Availability," Proc. 18th IEEE Photovoltaic Specialists Conference, pp. 235-239 (1985).
 N.W. Patapoff, Jr., "Two Years of Interconnection Experience with the 1 MW at Lugo," Proc. 18th IEEE Photovoltaic Specialists Conference, pp. 866-870 (1985).
 D.D. Sumner, C.M. Whitaker. and L.E. Schlueter, "Carrisa Plains Photovoltaic Power Plant 1984-1987 Performance," Proc. 20th IEEE Photovoltaic Specialists Conf., pp. 1289-1292 (1988).