In the richer countries, electricity supply is of the order of 1 KW per person, though some countries use substantially more. If this is taken as a target for electricity supplies, China alone will need more than 1000 GW - some 10 times its current generation capacity. Figure 1 shows the impressively rapid growth of electricity supply in China in recent years (1). A world population of 10 billion people, which is expected to be reached by the middle of the 21st century, will need some 10,000 GW of generating capacity. Thus, depending on the plant lifetime, up to several times this amount will need to be installed during the next century. Consequently, in order for most of the world population to have a reasonable standard of living, energy sources will be required that are capable of expansion at an average rate of more than 100 GW per year through the next century.

With normal economic growth, the actual rate will grow through the century by an order of

magnitude from less than this figure to several times more, but 100 GW per year is representative of the rate of construction required to solve humans' energy problems.

In recent years the ever-increasing scale of human industrial activities has started to threaten the global environment. Consequently the quality of life on Earth in the future will increasingly depend on using energy sources that are more environmentally benign than those used during the development of the older industrialised nations. Some researchers claim that a "fully industrialized world" is unachievable, based on the view that the energy supplies that would be necessary, the rate of raw material utilization that would be involved, and the accompanying environmental damage would not be sustainable. In a similar vein, long-term projections of world energy supply have been published that foresee average energy utilisation per person in the poorer countries 100 years from now which is only a fraction of that in the rich countries.

However, these projections do not seem acceptable as images of the future for poor countries, and particularly not for those growing rapidly. Now that it has been shown that sustained rapid progress is possible it seems likely that such growth will become increasingly common around the world. A major benefit of this is that economic development and rising living standards are the most effective means of reducing population growth. It can also be argued that a major cause of the economic growth in the advanced countries during the period after WW2 was the low and falling price of oil during the 1950s and 1960s. For these reasons the development of environmentally benign energy supplies that are capable of rapid expansion to a very large scale is of enormous importance. It will also be of enormous commercial value. Construction of 100 GW of electricity generating capacity per year at some 200 \ / Watt will represent a market of some 20 trillion \ per year. Revenues from this capacity will grow at several trillion \ per year, and there will also be a growing maintenance and refurbishment market.

Of course such rapid expansion of energy supplies will require capital investment on a very large scale. And in their high rate of personaland corporate saving the fast-growing Asian countries also seem to be a better model for developing countries than the rich countries of Europe and America. Although average incomes in the latter are much higher than in Asia, the proportion of income that is consumed is also much higher, and savings are correspondingly lower.

In order to power the expected economic growth, humans will need all of the energy sources to which they have access; the use of coal, oil, gas, wood fuel and nuclear energy will continue for many years to come. However, for reasons that are well known, it is also clear that none of these sources is capable of providing sustainable energy supplies on the scale required to provide all humans with an acceptable standard of living in an environmentally benign manner.

Microwave energy transmitted from space to Earth apparently has the potential to provide environmentally clean electric power on a very large scale, and with the potential for very rapid growth. In the 1970s the US Department of Energy considered a system of solar power satellites ( SPS) of 300 GW capacity, suitable for supplying the USA. This work was very valuable in clarifying both the potential of SPS, and the research that needs to be done. However, a system of 300 GW is far too small from the point of view of world economic development, since it represents no more than 3% of the 10,000 GW of capacity that is needed in the next century.

During the 1980s most SPS research was performed in Japan. Currently the major project is the "SPS 2000" project to demonstrate the actual transmission of 10 MW of power from space to Earth using near-term technology (2). Figure 2 shows the present SPS 2000 satellite configuration. In order to be demonstrated in the near future, the satellite will operate in 1100 km altitude low-Earth equatorial orbit, from where it will deliver power to receiving antennas (rectennas) within a few degrees of the equator, as shown in Figure 3.

Figure 2: SPS 2000 satellite concept.

Figure 3: Sites for SPS 2000 rectennas.

In the following the feasibility of solar power satellites is not considered; it is assumed that transmission of electric power to Earth from space is in principle an attractive energy source, capable of supplying continuous electric power to Earth in an environmentally benign way, on an effectively unlimited scale. The following is a brief review of some of the implications of using SPS to increase electricity output capacity on Earth by 100 GW per year.

It has been suggested that if humans import such large amounts of power, they will alter the energy balance of the Earth, and in particular will add to "global warming". However, any such effect would be small compared to the heating effect of adding carbon dioxide to the atmosphere. The solar energy intercepted by the Earth is some 180 million GW, of which only approximately half, or some 100 million GW, is absorbed, due to the reflection of sunlight from the Earth. Humans' total electricity production today is of the order of 1000 GW, which is therefore some 0.001% of the solar energy absorbed by the Earth. If this increases by a factor of 10, it will still be only of the order of 0.01% of the Earth's insolation, which is too small to have a significant global warming effect. It is also notable that because of the high efficiency of rectennas (some 90%), the heat added to the environment by SPS is less than half that created by even the most efficient thermal power stations.

Electricity generation from sunlight using photovoltaic cells, both on the earth and on solar power satellites, is different from traditional energy sources (other than wood fuel) in utilizing systems that comprise very many units of a small number of different components - solar panels, structural members, transmitting and receiving antenna panels - in very large quantities. It is thus inherently suitable for mass-production. The studies of SPS performed by the US Department of Energy in the late 1970s envisaged the production of 10 GW of SPS capacity per year. However, as seen above, for much of the coming century the world population will need electricity supply growth of more than 10 times this scale. Thus it is interesting to consider in outline the possibility of supplying such rapid growth with SPS.

Many designs of SPS are still competing, but for simplicity we follow the design of the SPS 2000 satellite of the SPS Working Group (2). This uses amorphous silicon photovoltaic cells for electricity generation, and 2.45 GHz microwave power transmission to Earth, using solid-state microwave generating modules. We assume average solar cell efficiency of approximately 10%, which is higher than that currently available, on the grounds that the invention of multi-band-gap solar cells renders the probability of achieving efficiencies even higher than this within a few decades reasonably high (3). On this assumption ten square meters in orbit will produce 1.4 kW of electricity, and so ten square kilometers will produce 1.4 GW in orbit. We assume that the microwave power transmission and reception system has an overall efficiency of 50%. This is higher than that available today using solid-state microwave generators, but less than that demonstrated in the 1970s using magnetrons. Thus 10 square kilometres in orbit will produce 0.7 GW at the rectenna on Earth, and so 100 GW on Earth will require some 1400 square kilometers of solar arrays in orbit.

Thus the production of 100 GW of SPS capacity per year will require production of some 1400 square kilometers of solar arrays per year, or some 4 square kilometers per day. The area of transmitting antenna modules required would be approximately an order of magnitude less than this. The total rectenna area required to receive 100 GW will depend on the intensity of the microwaves received at the rectenna. In order to economise on land use, it is probable that several satellites will transmit power to each rectenna, and so the total rectenna area will be considerably less than the satellite solar arrays. Thus, as a representative figure, we assume that some 6 square kilometres of solar arrays and antenna panels would need to be produced per day. If made in strips 5 m wide, this would require some 1200 km to be produced per day, of which some 800 kilometres would be solar panels. If produced at 10 plants using 24-hour production lines, the output speed would be some 1.4 metres / second, which is not very different from existing amorphous silicon cell production line speeds (though these are narrower today).

Production of SPS electronic components on such a scale might be called "Macro-Electronics". Being semi-conductor technology, it would be more readily automated than the manufacture of traditional thermal power stations. If the specific mass of SPSs is approximately 12 tons / MW, which is the target for SPS 2000, annual production of satellite parts for 100 GW output will have a mass of 1,200,000 tons. This is only a few percent of the mass handled by the automobile industry, and so is clearly not in itself difficult to achieve. The design of the factories, materials supply and transportation systems needed to achieve such rates of macro-electronic production would be an interesting exercise.

太陽発電衛星構造 - SPS Structure

The structural components of solar power satellites will be simpler than the electronic components, but very large numbers of units will be required, sufficient to support 1400 square kilometres of solar panels per year. Production and assembly of these will comprise many repetitive operations, and so will be very suitable for robotic construction systems. Thus, if SPS can supply power at competitive prices, construction will provide engineering companies with demand for light-weight structures on the scale of WW2 aircraft manufacturing, when, for the only time, tens of thousands of aircraft were produced per year. Another relevant precedent is the post-WW2 experience of production of the "Liberty ships" built rapidly in the USA to carry "Marshall Aid" to Europe. Using a new welded design the first ship reputedly took some two years to build, while the 200th ship took less than two weeks. Compared to these two examples, SPS production will be simpler, involving much larger numbers of fewer different components, and will therefore have the potential to reach cost levels little above the cost of the materials used.

Today perhaps only car manufacturers and electronic consumer-product makers have experience of such large scale manufacturing. SPS units will therefore probably be built by consortia of these companies and construction companies, who have experience of managing the production of such large and spatially complex structures. Modern-day aerospace companies, with their tradition of small production runs of high-cost, hand-built products will face fierce competition if they are to have a share of this new commercial business.

In order for SPS to be commercially competitive it is clearly a prerequisite that fully reusable launch vehicles and orbital transfer vehicles are developed and put into regular, airline-type operation. Today only the proponents of fully-reusable SSTO VTOVL rockets claim to be able to reach launch costs of around 20000 \ / kg (4), which is the approximate level required for SPS to become competitive. These claims are controversial. Some makers of expendable launch vehicles even claim that SSTO VTOVL is impossible. However, the Appendix shows clearly that SSTO is possible using existing hardware. What is not yet known is the cost of making such vehicles fully reusable. This will remain controversial until such vehicles are built and put into operation, as proposed for SPS in the 1970s and 80s (5).

Once this is done, if a typical cargo launch vehicle has a payload to LEO of 50 tons, and the specific mass of an SPS is approximately 12000 tons / GW, construction of 100 GW of capacity per year will require 24000 flights to LEO per year, or some 70 flights / day. Although far beyond today's expendable launch vehicle industry, this is very small by comparison with modern-day air transportation, being the traffic rate of a single small airport.

However, if one considers the probable pattern of development of a commercial SPS industry, it seems likely that such a high rate of launch from Earth will not be needed. Initially all SPS components will be manufactured on Earth, and orbital operations will involve mainly orbit transfer and assembly. However, due to the losses in microwave power transmission and reception, the cost of the electricity used to generate microwaves on an SPS would be some 50% of the cost as delivered on Earth. Thus, if SPS electricity costs are competitive with other terrestrial power sources, the orbital price of electricity at that date would be some 50% of the market price on Earth. Consequently it is likely that energy-intensive industrial operations involved in SPS manufacture will increasingly be performed in space.

This will create a large market in Earth orbit for raw materials such as silicon and aluminium, which will be manufactured into SPS components in orbit. At a later stage these may be supplied from extra-terrestrial sources such as the Moon and asteroids, from which delivery costs could fall below launch costs from Earth. Without trying to predict a time-table for such developments, it is possible to envisage a range of different scenarios driven by the rapidly growing commercial demand for electric power on Earth. It has been argued that technologically such developments are not particularly difficult or expensive by comparison with modern mining projects, and that they will surely occur when there is a need for them (6).

Thus, once launch costs are sufficiently low for SPSs to be able to supply electricity at competitive prices, it seems likely that international commercial consortia will manufacture and operate SPSs, selling power profitably to electricity supply companies on Earth. Later companies will also invest in the development of extra-terrestrial materials on a commercial basis, without the need for further government subsidy.

In order to assess the above possibilities in more detail, the next step required is to verify that SPS is in fact a good candidate for such large-scale electricity supply in the future. This requires work in at least four different directions.

First, the SPS system must be tested in practice. This task is the objective of the SPS 2000 project currently being planned to deliver power from space to Earth in the near future (2).

Second, fully-reusable launch vehicles with launch costs to LEO of approximately 20000 \/kg must be developed. Among government projects today, only McDonnell Douglas' "Delta Clipper" rocket is claimed to be able to reach this goal within the near future (4). Thus fully-reusable SSTO VTOVL vehicles seem attractive candidates for further study for SPS cargo transportation.

Third, the mass-production of SPS components must be studied further in order to prepare SPS designs suitable for rapid rates of production.

Fourth, suitable rectenna sites must be identified and prepared, involving analysis of their economic, environmental and social impacts. The SPS 2000 satellite is planned to operate in a low equatorial orbit of 1100 km altitude, in order to demonstrate power transmission from space to Earth as economically as possible. Currently rectenna sites are being sought within 3 degrees of the equator, but even sites at latitudes as high as 4 degrees such as the Zeng Mu An Sha or Spratley islands, may be possible.

A later project to follow SPS 2000 will give a wider range of countries the chance to gain experience of building and operating rectennas. The country with the largest energy needs is China, which has plans for very large increases in energy supply (1). China also has a large land area, and so seems to be an excellent candidate for rectenna sites. Detailed studies of possible sites in China, both in general, and in southern districts which could be reached by an SPS in low Earth orbit, would therefore be valuable (8).

All people in the world wish for a good standard of living. Achieving this will entail a massive increase of energy supplies with low environmental impact around the world. After the advanced nations have adjusted to the post cold-war situation, substantially more resources will be available for consumer-led economic growth, particularly the skills of engineers and scientists who were previously employed in weapon system development. However, in order to obtain these benefits for society, the motivation of young people to study technical subjects must be maintained.

Takeuchi has described that the future should not be seen as the spontaneous result of economic and social trends, but rather as the result of choices that we make (9). Thus we should urgently study SPS in parallel with any other energy sources that seem to offer the possibility of expansion by more than 100 GW per year while causing minimal pollution.

With the end of the cold war, the aerospace industries of the leading countries are currently faced with substantial over-capacity and the need to contract. Yet, to date, they have shown little interest in SPS outside Japan. Using the advanced engineering skills released by the end of the cold war in order to solve the growing energy problems of the rapidly advancing countries on a commercial basis seems an attractive way to create a good future for humankind.

1. E Osawa, 1992, 中国の電気事業, Macro Review Vol.5 No.1 pp 41-47
2. M Nagatomo and K Itoh, 1991, "An evolutionary satellite power system for international demonstration in developing nations", Proc. SPS91 pp 356-363
3. A Suzuki et al, 1988, " SPS is the next goal of commercial solar cells", Space Power Vol.7 No.2 pp 131-143
4. W Gaubatz et al, 1992, " Single stage rocket technology", Proc. 43rd IAF Congress paper no. IAF-92-0854
5. D Koelle, 1981, " SPS transportation requirements - economical and technical", Space Solar Power Review Vol.2 pp 33-42
6. D Strangway, 1979, " Moon and asteroid mines will supply raw material for space exploitation", Canadian Mining Journal Vol.100 no.5 pp 44-52
7. P Collins, 1993, " Concerning a solar power satellite to follow SPS 2000", 12th ISAS Space Energy Symposium pp 160-166.
8. K Takeuchi, 1993, " Perspectives of advanced-technology society", in " Perspectives of advanced-technology society" 1992, RCAST, Tokyo University, pp 6-13