The views of potential customers of the SPS project are very important in the decision on how to proceed. The customers for the SPS will be the electric utilities of the world. In general, electric utilities are large companies with strong finances, based on their large, high quality cash flows and with extensive technological skills centered on electrical engineering. If electric utilities became convinced that SPSs offered a profitable source of electric power, they could more or less finance their development. The initial priority for SPS research should therefore be to interest the electric utilities, and progressively to convince them that SPSs could be a commercially competitive source of electric power.

However, electric utilities do not have space engineering expertise, and it is not likely that they will develop such expertise - a task which would be very expensive and of uncertain benefit. From this viewpoint, most SPS-related experiments that have been proposed by the space industry, such as generation and transmission of microwave power between satellites in orbit, are of little interest to utilities. They cannot use them; they cannot assess them technologically; and they cannot draw any useful conclusions from them about the likely future cost of SPS power.

By contrast, it would be of considerable interest to utilities to demonstrate the transmission of power from an orbiting satellite to a rectenna on the ground. Consequently the objective of the proposed "SPS 2000" project (1) is to demonstrate the generation of useful electric power on Earth from a microwave beam transmitted from an orbiting solar power satellite, using available space technologies and launch services. The "SPS 2000" project has a number of key design features which have major implications for the design of the ground segment.

"SPS 2000" Key Design Features

1. Equatorial Orbit: The Space segment is to operate in an equatorial orbit, in order to provide rectennas with more frequent deliveries of power.
2. LEO satellite: The satellite will operate in LEO, approximately 1000 km altitude, in order to reduce launch costs, and to minimize the size of the transmitting antenna required.
3. 3 km diameter microwave beam "footprints": The microwave power transmission system (MPTS) will be sized to deliver power within a "footprint" some 3 km in diameter (North-South) when the transmitting antenna is 100 m in diameter.
4. Power output from 1MW to 10 MW. The space segment will have a power generation capacity of approximately 1 MW(rf) initially and will be expanded to as much as 10 MW(rf).
5. Space segment to be upgraded: The space segment will be upgraded progressively in order to increase the value of the energy delivered to each rectenna.

The operation of the SPS 2000 system would generate both electric power and information about the system's operating characteristics. This raises interesting questions about the optimum balance between these two objective. Although the provision of commercially competitive electric power is not the initial objective of the "SPS 2000" project, this is the ultimate objective. Consequently economic considerations are important in the design of the ground segment, and will be of particular interest to electric utilities.

The main implication of the selection of an equatorial orbit for the SPS 2000 space segment is that it all be on or near to the equator. The maximum distance from the equator that would be possible will depend on the design of the MPTS, in particular the microwave beam steering capability. It is planned that the transmitting antenna will point to the Earth's center and the microwave beam will be electronically steered to an angle of 30 degrees to the vertical, which will limit the siting of rectennas to within a few hundred kilometers of the equator itself. However, as the latitude of the rectenna increases, the duration of power transmission from the satellite decreases, reaching zero at about 600 km from the equator.

The countries in which rectennas could in principle be sited are any or all of the following: Ecuador, Colombia, Brazil, Gabon, Congo, Zaire, Uganda, Kenya, Somalia, Indonesia, and some Pacific Islands. It is noteworthy that none of these countries is a major industrialized country, and that their equatorial regions are mostly low-income, agricultural areas where electricity is not readily available. One or more of these countries could obtain major benefits from playing an important role in the development of the SPS.

To the extent that the rectennas were seen as valuable facilities, economically, scientifically and/or politically, there might be many candidate nations. For producing maximum engineering information it would be desirable for several rectennas to be built and operated. This creates the possibility of significant economies of scale and learning through mass production; of political momentum through international collaboration; and of United Nations support for a project of interest to both industrialized and industrializing countries.

LEO satellite

The selection of an altitude of approximately 1000 km entails that transmission of power to a rectenna will be of quite short duration, some 200 seconds continuously if the microwave beam can be steered up +/-30 degrees to the vertical. It also entails that the angular direction of the microwave beam will be changing quite rapidly, approximately 0.3 degrees/seconds.

In order to simplify the initial SPS design it is not planned to transmit power during the local night-time which would require on-board power storage. Consequently for a 1000 km circular orbit there would be some 8 transmission periods/day to each rectenna, giving approximately 1600 seconds total power delivery per day.

A wide range of experiments will be possible using the ground station (2). These will include tests of the system's steady-state operating characteristics under different conditions, and tests of its response to the range of transients to which it will be subject - start up, shut-down, load-following, unplanned outages and so on. Extended tests will also be possible on the system's reliability, and on its side-effects such as electromagnetic interference under different conditions.

For users of the power delivered by the SPS 2000 ground segment, continuous electric power will be more valuable than short bursts of power, and so from the point of view of power generation, it will be desirable for the rectenna system to include some energy storage capacity. Such a system would be required to absorb energy for about 200 seconds and discharge for about 2 hours.

Thus, for example, if a rectenna generated 100 kW (E), it could store (100 E/18) kWhr, and deliver approximately (100 E/18 x 2) kW continuously, where E is the efficiency of the electrical storage-discharge cycle perhaps 2 kW (e) continuously with an efficiency of the storage cycle of 72%. On the same assumptions, 1 MW (e) from the rectenna could produce 200 kW (e) continuously. Today, commercially available batteries cannot be charged as fast as this. Consequently alternative energy storage technologies, such as water pumping would be necessary. However, as the space segment is upgraded, the performance required of the storage system is reduced (see below).

In general, average rectenna power output will be

(P.T.Y.E.N/24) kW(e),

where

P (kW rf) is the rf power transmitted to the rectenna,

T (hrs) is the length of continuous power delivery to the rectenna,

Y is the efficiency of rf-DC power conversion at the rectenna, and

N is the daily number of satellite passes in daylight.

The initial design is intended to minimize the development cost of the space segment, which is much more expensive than the ground segment. Subsequently, in order to reduce the cost of electricity produced, it will be desirable to perform a trade-off between many factors, both technical and economic.

A nominal SPS 2000 unit with 100 m diameter antenna would have a footprint of 3 km on the ground. For an equatorial rectenna, the microwave beam terrestrial "footprint" of 3 km diameter determines the minimum area of 7 square km. During the period of power transmission the area illuminated by the main lobe of the beam will start as an East-West ellipse with major axis of 4 km, shrink to a circle of 3 km diameter, and increase to an East-West ellipse again, the maximum East-West diameter being determined by the maximum angle of inclination that is feasible (see Figure 1).

For an initial beam power of 1 MW (rf), the average power density would be only 141 mW/sq m, which is less than 0.1% of that in the DOE SPS Reference System. The actual power density at different points of the rectenna will depend on the microwave beam power density profile, and will influence the optimum size of the rectenna

For a commercial SPS system, the size of a rectenna is determined by economic considerations: The amount and hence the value of power received per sq m of rectenna surface decreased with distance from the center of the beam. The edge of the rectenna is where the cost per sq m of the rectenna is equal to the present value of the energy that the rectenna will produce per sq m over its lifetime, less the cost of microwave supply.

With an average requirement of approximately 7,000 sq m/kW (rf), and so of approximately 350,000 sq m/kW(e) continuous, the ground system is clearly not cost-effective in commercial terms, even if the microwave power is delivered without charge. This would require the rectenna to cost approximately 1 Yen/sq m (see below).

The objective of operating the SPS 2000 rectennas is not only to generate electricity but also to generate information. Nevertheless, it may be that the optimum rectenna size would be considerably less than the overall beam footprint (see below).

From above, the continuous rectenna power will be

(P.T.Y.E.N/24) kW(e).

If the value of 1 kWhr is V Yen, the annual electricity revenue would be

(P.T.Y.E.N/24) 8760.V Yen.

that is

(365.P.T.Y.E.N.V) Yen.

The present value of the lifetime revenue is therefore:

(365.P.T.Y.E.N.V.D) Yen,

where D is given by ((1+i)n - 1)/i(1 +i)n, where i is the interest rate, and n is the operating lifetime. The limiting value for this (when the lifetime is indefinite) is 1/i.

Typical values for the variables might be: P = 1000, T = 1/18, Y = 0.5, E = 0.8, N = 8, V = 10, D = 5. These give a value of 3.2 million Yen or approximately 0.5 Yen/sq m.

The central portion of a rectenna would have a power density perhaps 10 times this average, giving perhaps 5 Yen per sq m. Even at this level of revenue the rectenna would not be commercial. It seems possible that even for demonstration purposes, only the central portion of the microwave beam footprint would be collected. The practical cut-off point would depend on many factors, but might be as little as 0.5 km from the beam center. Because the satellite would be travelling from West to East, the effective East-West extension of the rectenna would need to be increased if all the power available were to be collected. However, the economic value of the outer regions of an East-West ellipse would be even less than the value of the central circle due to the even shorter period of power delivery.

Once a first 1 MW sub-unit of the "SPS 2000" space segment is operating, it is proposed to upgrade it progressively by adding more 1 MW sub-units up to a maximum of 10 MW(1). Increasing power output in this way would provide further experience of system design and operation, and would also reduce the cost of power produced: For a given rectenna cost, the cost per unit of electric power produced would be inversely proportional to the rf power delivered to the rectenna

NB the intensity of the microwave radiation outside the central section, even for 10 MW(rf) transmission, would lie within existing public exposure limits.

The space segment could be upgraded in other ways:

1. Power could continue to be delivered in short bursts, but more frequently, by siting more satellites in the same orbit. Ten satellites spaced evenly around the orbit would require only nine minutes storage capacity at the rectenna to provide continuous power. Such satellites need not be produced by a single organization; a number of different groups might use different designs, thereby introducing an element of competition.

2. Alternatively, longer periods of continuous power could be delivered by formation flight of multiple satellites close together in the same orbit, and coordination of their power transmission to rectennas. For example, ten satellites would provide 30 minutes power every two hours, requiring 1 hours' storage capacity in order to provide continuous power at each rectenna

3. The periods of power delivery could be lengthened by increasing the satellite altitude. This would reduce the number of rectennas that could each receive the maximum duration power from a single satellite, the extreme case being the 1:1 GEO satellites of the US Reference System.

   Raising the altitude also increases the distance North and South of the equator to which the satellites could deliver power. In addition, raising the altitude increases the area required of the satellite transmitting antenna and/or of the rectenna In order to reduce the cost of power it is then necessary to increase the satellite power output, as proposed, for instance, in the Energy Storable Orbital Power Station (ESOPS) with an output of 70MW, and transmitting antenna diameter of 200m (3).

4. The periods of power delivery could be extended through the night by using energy storage on the satellites. For example, the ESOPS proposal includes 45 minutes thermal storage capacity, permitting power delivery on every satellite pass (3). The optimum allocation of storage capacity between the ground and space segments will depend on many factors, including economic considerations.

5. The rf beam parameters (including power density profile and wavelength) could be altered to achieve an improved power distribution pattern at the ground. The first SPS 2000 space segment is designed for minimum development cost, but later units might be used to evaluate different antenna systems.

6. The power delivered could be increased by delivering power to a rectenna from two satellites close together in the same orbit simultaneously, as proposed in (4). If this possibility is feasible in practice it would greatly increase the potential value of a rectenna to an electric utility, by reducing the capital cost per KW of capacity by 25-33% (5). It would thereby increase the price that utilities could offer to satellite operators for supplies of rf power. The SPS 2000 system would enable exhaustive experiments to be performed.

These possible developments all lead towards the final objective of delivering continuous, high-density rf power, and all have implications for the design of the ground segment. In particular, if such changes were planned to occur within approximately 5 years of the initial operation of the system, it would probably be economic to design the ground system initially to accommodate them. For example, if the power density in the microwave beam was to be increased, it would be appropriate for the rectenna to have a higher power capacity than if this change was not planned, even at the cost of having lower rf-DC efficiency initially.

The power storage and related systems could be altered fairly easily at a later stage, being discrete systems at the edge of the rectenna. However, the rectenna itself, covering some millions of sq m, would be less easy to upgrade. The detailed design of the rectenna surface electronic circuitry is very dependent on the planned power density. Consequently if this was to be increased by several hundred percent, this would have a major effect on the optimum design.

It would, nevertheless, be possible to design a rectenna for ease of upgrading. For example, if made easily moveable like a lightweight mesh, the panels might initially have power handling capacity in inverse proportion to their distance from the center of the rectenna Then, in order to upgrade the rectenna, all panels would be moved radially outwards and reconnected appropriately, and new, higher-capacity panels would be sited in the center.

Rectenna System Detailed Design Considerations

The minimum cost rectenna would consist of a single flexible layer in the form of an open mesh that could be installed simply by unrolling long sheets and interconnecting them appropriately. The mesh could consist of 2mm diameter plastic cable carrying rf-DC printed circuits and covered with a protective layer. Such an open mesh would pass rain and sunlight, but it would have an efficiency of only perhaps 10% since the efficiency of rectenna power reception is very dependent on the presence of a reflector plane (see Figure 2).

A simple reflector plane could be added to such a mesh by attaching a metallised open plastic framework behind the active power reception plane. This would increase the mass by some 200%, and the cost by perhaps 100% (being much simpler than the front surface), but would increase the bulk for transport and deployment by perhaps 1000%. The efficiency would be increased to perhaps 50% (see Figure 3).

These two designs could be installed relatively easily even on unprepared land, thereby causing minimal environmental disruption. For some purposes however, a more expensive, maximum efficiency rectenna design would be preferable, such as the circular microstrip antenna devised by Itoh et al (6).

In the DOE SPS Reference System design high rectenna efficiency was achieved by interconnecting many rf dipoles and using high power diodes with high efficiencies. This approach increases the directional sensitivity of the rectenna.

In order for the "SPS 2000" rectenna to receive power efficiently from 30 degrees East and West of the vertical, it will be necessary for series connections between dipoles to be mainly in the North-South direction. The rectenna design proposed by Itoh et al (6) has the capacity to receive power efficiently from a wide range of East-West directions.

Figure 3: Low Cost Rectenna with Reflector Plane.

For any given rectenna site there is a broad choice between a low-cost but low-efficiency approach and a high-efficiency but high-cost approach. The former would demonstrate the generation of some power at the lowest total cost; the latter would produce the most power and the most information for development purposes.

The optimum approach in any case would depend on the objectives of those financing the project. Economies of scale would be maximized by building and operating several similar rectenna systems; engineering experience would be maximized by using several different designs. A possible compromise would be to design two "standard" rectenna systems, at least initially:

Type A rectennas would be designed with the intention of being upgraded progressively to become large and stable sources of commercial electricity generated by a future geostationary SPS system. They would therefore have high efficiency and high power handling capacity.

Type B rectennas would be designed to remain local sources of relatively low power, with energy storage systems, at minimum cost. The "hatsuden mahoh no jutan" described above would clearly cost much less than a rigid structure.

In addition to either of these approaches, engineering research will require the installation of numerous sensors, data collection equipment and control systems, in order that utilities should be able to assess the potential of the SPS to the greatest extent possible. Such research would be valuable with both Type A and Type B rectennas. The potential scope of this research is discussed in (2). In brief, the SPS 2000 ground segment, and the experiments that it makes possible, have the potential to resolve essentially all of the concerns on which electric utilities must be satisfied before they can offer a price to satellite operators for supplies of microwave power from space.

It is clear that many factors will have to be taken into account in the design of the SPS 2000 ground segment. Although SPS 2000 rectennas will initially be more valuable as sources of engineering data than as sources of electric power, they will nevertheless have the potential to earn commercial revenues from the sale of electric power from an early stage.

The research for which the rectennas will be used will be both technical and economic. This is, while technical results such as determining the rectenna design with the highest efficiency will be important, ultimately it will be the most cost-effective design, that is the design giving the minimum cost per kilowatt-hour delivered, which will be preferred by utilities. Consequently, not only the technical results of experiments performed with different ground system and sub-system designs will be of interest, but also the costs of these different experiments.

Following the comprehensive programme of experiments that will be possible with the SPS 2000 ground segment, electric utilities will be in a position to estimate reasonably accurately the cost of construction of a rectenna to a given specification, its operating costs, and the cost of integrating its output with their distribution grids. This will enable utilities to calculate the price which they would be prepared to pay potential satellite operators for supplies of microwave power to a given specification.

The ground stations of the SPS 2000 project will be very important research facilities, being of unique interest to electric utilities in every country. Ultimately, as the space segments are upgraded to provide continuous, high-density rf power, the ground segments will become major sources of electricity. The SPS 2000 project therefore offers the equatorial countries the opportunity to play an important role in the development of the SPS, which promises to become a major, non-polluting energy source for the world.

1. M Nagatomo and
2. I Kiyohiko, 1991, "An Evolutionary Satellite Power System for International Demonstration in Developing Nations", Proceedings of SPS 91 Power from Space Conference
3. P. Collins, R Tompkins and M Nagatomo, 1991, "SPS 2000: A Commercial SPS Test-bed for Electric Utilities", Proceedings of 26th IECEC
4. R Akiba and H Yokota, 1987, " Systems Analysis of Energy Storable Orbital Power Station (ESOPS) in Medium Altitude Equatorial Orbit", Proceedings of Pacific ISY Conference, pp 66-69
5. P Collins and R Gelsthorpe, 1980, " Increasing Rectenna Output by using a Pair of Satellites", Electronics letters, Vol. 16, No 9, pp 311-3.
6. P Collins, 1985, " Economics of Satellite Solar Power Stations", PhD thesis, London University.
7. K Itoh and Y Ogawa, 1991, " An Inland Rectenna using Reflector and Circular Microstrip Antennas", Proceedings of SPS 91 Power from Space Conference Volume 10, Numbers 3 & 4