Satellite solar power stations ( SPS) have the technical potential to grow into a major new source of electric power, and should therefore be of interest to electric utilities around the world. However, SPS research proposals have not made it easy for utilities to be involved. For example, it is not easy for utilities to use power transmission experiments between orbiting satellites to assess the potential of the SPS as a terrestrial energy source. Yet if the SPS project is to progress, utilities must play a major role.

The "SPS 2000" project, devised by the SPS Working Group at ISAS (1), has been designed to provide utilities with an experimental facility with which they can directly study the potential of the SPS as a terrestrial energy source, and therefore contribute directly to the evaluation and design of the SPS.

The overall configuration of the space segment is a triangular prism with a length of 800 m and sides of 100 m, as shown in Figure 1. The main axis lies in the north-south direction, perpendicular to the direction of orbital motion. The transmitting antenna on the horizontal under-surface faces the Earth, and the other two sides of the prism carry solar arrays. Energy storage for operation during eclipse is not to be provided initially (1).

This configuration was selected to permit gravity-gradient stabilization and to facilitate deployment in orbit. High operating efficiency through use of advanced technology is of lower priority than enabling each 100 m section, generating 1 MW, to be orbited as a single, self-deploying payload (1).

The transmitting antenna will be a phased array operating at 2.45 GHz. and capable of steering the microwave beam up to 30 degrees either side of the vertical. Consequently, from its orbit of 1000 km altitude the satellite will be able to transmit power to an equatorial rectenna for some three minutes during each orbit. In order to receive all of the power transmitted, a rectenna would need to be 3 km in diameter, but a smaller structure might be used initially. Low-cost rectennas could be valuable for power supply to poor equatorial communities, but one or more would be used primarily as electric utility SPS research centres, and would be provided with a range of associated research equipment and facilities (2).

Before describing the tests that will be possible on both technical arid economic aspects of SPS 2000, it is useful to describe the sequence of events during each satellite pass.

The timing and geometry of each overhead pass by the SPS 2000 satellite will be known precisely. Typically it will be above the local horizon for some 19 minutes during each pass, which will start with the reception of an omnidirectional beacon on the satellite. This will enable the ground station staff to monitor several parameters of the satellite's behaviour before transmitting a signal to the satellite to align the microwave power beam precisely on the rectenna

Some 8 minutes after the satellite's appearance over the western horizon its microwave beam will be at full power, at an angle of 30 degrees to the local vertical. The footprint of the beam will be an ellipse some 4 km EW by 3 km NS travelling east along the equator towards the rectenna at some 6.8 km/sec. When the centre of the beam reaches the centre of the rectenna the transmitting antenna control system will start to alter the angle of the microwave beam to the antenna in such a way as to keep the beam centred on the rectenna during the nominal power delivery period of some 3 minutes.

A rectenna on the equator will be either circular or elliptical EW, and in either case the expected pattern of power increase and decrease will be known in advance, depending on such factors as the geometry of the elliptical beam footprint overlapping the rectenna; the microwave beam power density cross-section; the distribution of microwave power-handling capacity across the rectenna and the variation in satellite RF power output as the angle between the Sun and the satellite solar arrays changes. It will therefore be possible to monitor very accurately the performance of many parts of the system.

Once the period of power delivery is over, the satellite beacon will continue to be monitored until the satellite goes below the eastern horizon. The approximately 90 minutes between satellite passes will be available for analysis of the data generated during the previous pass, and for preparation for the following pass.

During the years following initial implementation the space segment will be enhanced either through additional sections being added to create a single larger satellite, or the addition of independent satellites, which would reduce the periods between power transmission tests.

An exhaustive list of the experiments possible with the SPS 2000 rectenna that would be of interest to utilities would be very long, and would require the collaboration of experts in several different fields to compile. The following is a list of the main types of experiment that will be performed.

It will be possible to measure the efficiency of the SPS 2000 rectenna in converting RF power to DC electricity under different conditions. Understanding of the system behaviour could be increased progressively by adding more sensors to the system.

The "basic" experiment will be to measure the SPS 2000 system's normal operating characteristics. it is of course important to attain high RF-DC conversion efficiency, but costs must also be considered, in order to determine rectenna designs that produce electric power at the lowest cost. Normal operating characteristics will also be measured under different environmental conditions - temperature, humidity, precipitation, state of the rectenna surfaces, etc.

There are a range of transient conditions which rectennas will experience and which, in order for the SPS to be useful to utilities, they will need to tolerate while continuing to deliver power predictably:

It will be possible to start power delivery to the SPS 2000 rectenna in different ways: Low power delivery centred on the rectenna and building up to a maximum; full power beam centred on the rectenna but initially unphased, becoming progressively narrower while aimed at the rectenna; sequential switch-on of different segments of the transmitting antenna while aimed at the rectenna; switching the direction of the full-power, phased beam onto the rectenna; and other modes. The behaviour of the rectenna and its subsystems will need to be measured and understood during all these possible cases, in order to select the optimum modes of operation and to detect any system weaknesses.

Similar considerations apply as to the case of start-up transients.

During commercial electric power system operations a variety of system transients occur, and consequently it will be necessary for a rectenna to operate stably during the range of transients to which it will be subject. Thus the behaviour of the SPS 2000 rectenna will need to be monitored during power reception when the load to which it is delivering power is changed, both through switching in and out different sections of the rectenna, and through altering the load being drawn through the power conditioning system.

A particular case of change-of-load transients will be when an SPS is used for "load-following". The potential value of the system would be considerably increased if rectennas could be used efficiently for load-following (3). Experiments with the SPS 2000 system could be used to evaluate the feasibility of this, and the design requirements to achieve it. In general, reducing the power drawn from a rectenna either requires the power output of the satellite solar arrays to be reduced, or for power to be "dumped" somewhere else in the system, such as an increase in EF power reradiated from the rectenna The latter would be more or less acceptable depending on its characteristics.

Another type of transient through which rectennas must operate stably is the range of component failures to which they will be liable. In order to reach a target level of overall reliability, a commercial rectenna would be designed with components of specified degrees of reliability, and would be maintained in such a way as to achieve the target reliability. It will be possible to test the SPS 2000 system to ensure that it has graceful failure modes.

A geo-stationary SPS would on occasion pass through the Earth's shadow and lose all or part of its power output. Being predictable, this will of course be planned for. Operating strategies to be followed to accommodate this could be practised with the SPS 2000 system.

Utilities are required to achieve specified degrees of quality and reliability in their power delivery. Thus they are willing to pay higher prices for supplies of power from more reliable sources than for supplies from less reliable sources. In assessing the SPS, utilities will therefore need to determine the level of reliability attainable, and the cost of achieving specified levels. Once the SPS 2000 system has passed through its "teething troubles" and detailed operating procedures have been established, it should be possible to use it to collect long-term cumulative data on the ystem's operational reliability.

RE power density levels around a commercial SPS rectenna would have to meet accepted health and electromagnetic interference (RF "noise") standards. This RF would comprise direct, reflected, and reradiated power. The levels of both RF power and RE "noise" generated by the SPS 2000 satellite will be below international health limits, but it will be valuable to monitor RF levels around the SPS 2000 rectenna in different frequency-bands under different conditions.

RF "noise" around the rectenna would need to be monitored during normal operation in a variety of different environmental and load conditions.

Short-term changes in the load being drawn from a rectenna could cause large variations in the RF power levels around the rectenna, which would need to be measured and understood.

Microwave intensity levels beneath the rectenna would be extremely low, but would vary to some extent during system operation, arid between different rectenna designs. It could therefore be valuable to monitor these.

Rectennas could produce a significant amount of waste heat under certain operating conditions. In general for the SPS 2000 system this effect will be small.

In view of the great value of such a wealth of information about the SPS, it is clearly important that both the ground and space segments of the SPS 2000 system should be designed to provide answers to as many of utilities' concerns as possible, within the project's technology, time and budget limitations.

The SPS 2000 system would evolve progressively in operation over several years. The original rectenna research centres would evolve through components, sensors, control capabilities, antenna segmentation and other features changing progressively. Additional rectennas would also be built, including ones at non-equatorial sites as either higher altitude equatorial satellites or non-equatorial satellites were orbited, arid ones at offshore sites since utilities in many countries would need to site rectennas offshore (4). New rectenna capabilities would be designed in the light of operating experience of earlier models.

The SPS 2000 space segment would also evolve, successive parts of the space segment having different designs, changed in the light of operating experience. They could also be placed in successively higher orbits to give longer periods of power transmission (5) and into non-equatorial orbits in order to give non-equatorial rectennas experience at an early stage. As lower cost launch systems become available, such further models will evolve into commercial facilities.

In addition to the range of technical experiments described above, the SPS 2000 will also provide fundamental information for cost analysis. Once electric utilities have operational experience of all aspects of the ground segment and its associated systems - their design, construction, operation, maintenance, reliability, failure modes and repair - they will be able to assess the costs of each of these. Knowing these costs and the cost of competing energy sources, utilities will be able to calculate the maximum price that they could profitably pay for deliveries of microwave power to given specifications from a full-scale SPS, and will be able to quote prices for such supplies of "microwave fuel", as discussed in (3).

The cost to a utility per kWh of SPS bulk power can be expressed as the sum of Crk the rectenna capital cost contribution, Cro the rectenna operating cost, and Cmw the cost of microwave power delivered to the rectenna:

Cel = Crk + Cro + Cmw

Knowledge of Crk and Cro based on experiments with SPS 2000, and of the cost of competing power sources, will enable utilities to estimate Cmw, the maximum price that they could profitably pay to satellite operators for deliveries of microwave "fuel" for their rectennas.

Thus, to give a simplified example, a utility might learn from working with SPS 2000 that the use of a commercial SPS rectenna for supplying daytime power at a certain load factor would cost them 2 US cents/kWh. If they could obtain bulk power of similar quality at, say, 5 cents/kWh they could then offer up to 3 cents/kWh to satellite operators for supplies of microwave power to their rectenna.

This approach has important advantages over that followed in most SPS research whereby the satellite- rectenna combination is treated as a single economic unit. In particular, considering satellites and rectennas as independent commercial units decouples the load factors achieved on satellites from those achieved on rectennas. This is desirable since in reality they would be determined by different technical arid market influences, as discussed in (3).

High load factors are important for SPS satellites because they have much higher capital costs than rectennas, but low load factors may be acceptable for rectennas since daytime and peak electricity (that is, power with a lower load factor than base-load power) are of greater value to most utilities than base-load power.

For example, an SPS might not be competitive as a baseload power source for which the cost of alternative sources is relatively low, even if both satellite and rectenna were operated at load factors of 0.8 (see point A in Figure 2). It might also be uneconomic if operated with a load factor of only 0.67 (typical of daytime demand) on both satellite and rectenna due to high satellite costs. Figure 2 shows how such a system might nevertheless supply daytime power profitably with a load factor on the rectenna as low as 0.6 provided that the satellite load factor was 0.9 (ic involving delivering power to more than one rectenna in succession). The shaded area in Figure 2 illustrates the range of load factors over which the system could be both profitable and competitive for supplying daytime power.

Once utilities are able to offer prices for supplies of microwave fuel, the space engineering companies will be "in business". That is, there will be a commercial market for any company that can meet the necessary cost targets. Those that can will be starting to tap a world-wide market with potential to grow to hundreds of $ billions per year. Possible roles for government in helping the space industry to meet these commercial targets will also be clarified by the existence of a market price for microwave power from space.

As an inherently international project, "SPS 2000" will be a unique source of research information for the electricity industry world-wide. In addition to providing electric power, the ground segments have the potential to resolve many of the concerns, both technical and economic, on which utilities must be satisfied before they can offer a price to satellite operators for supplies of microwave power from space.

If SPS can achieve economic viability, it will nevertheless be a dramatically innovative form of power supply. Utilities would probably need to follow an incremental path to developing and evaluating the concept. SPS 2000 offers such an approach.

Consequently the "SPS 2000" project provides the opportunity to change the SPS proposal from being a far-future possibility of very uncertain value into a project with clear commercial prospects once the space segment reaches defined cost targets. It thus provides the utilities of the world with the opportunity to evaluate the SPS in a way not possible by any other means, and thereby to create a new candidate for large-scale electric power generation in the future.

1. M Nagatomo, 1991, "An Evolutionary Satellite Power System for International Demonstration in Developing Nations", Proc SPS 91 Power from Space Symposium, Electricite de France
2. P Q Collins, 1991, "Design considerations for the SPS 2000 ground segment", Proc SPS 91 Power from Space Symposium, Electricite de France
3. P Q Collins and R Tomkins, 1991, " A Method for Utilities to Assess the SPS Commercially", Proc SPS 91 Power from Space Symposium, El. de France
4. ESA, 1980, " Study of Infrastructure Considerations for Microwave Energy Ground Receiving Stations", ESTEC 4382/80/NL/PP(SC)
5. R Akiba and H Yokota, 1987, " Systems Analysis of Energy Storable Power Station (ESOPS) in Medium Altitude Equatorial Orbit", Proc Pacific ISY Conference, pp 66-9