The primary difficulties with the dream of Space Solar Power (SSP) for earth, are the extreme launch costs of solar power satellites to Geosynchronous Earth Orbit (GEO), and the absence of an evolutionary path to SSP. This makes the cost-to first- power unacceptably high. We present a 3-stage approach to SSP, and lay out the problems and opportunities. The key idea is to use space assets initially for global transmission and distribution, rather than generation, establish the infrastructure, and then add space-based power generation to a revenue-generating power grid. In the first stage, a Space Power Grid is established with satellites 1,200 km above Earth, distributing earth-generated beamed microwaves to other satellites and ground receivers. This boosts the earth-based alternative power industry (wind and solar-thermal) by providing transmission between any points on earth, and accommodating fluctuations in demand and supply. It also satisfies strategic needs for emergency power delivery. In Stage 2, direct power conversion technology will augment the existing space power grid with space-generated solar power. In Stage 3, large ultralight GEO reflectors will beam light to the direct-conversion collectors, and multiply the power through the grid. The need to put expensive, heavy solar arrays in high orbit is avoided, along with the objections to atmospheric transmission of visible light. The system would gain significantly from the development of low-mass, high-efficiency conversion equipment for direct conversion of broadband solar energy to beamed microwaves.
End-to-end efficiency of a space based power grid could reach 50%
At present, the end-to-end efficiency of this process alone does not compare favorably with earth-based transmission of energy, in existing markets. With 70% at conversion, and 10% for each atmospheric pass, even with essentially 100% waveguides and in-space transmission, the end-to-end efficiency is limited to roughly 50%, compared to about 90% for transmission over high-voltage lines. However, this masks the value of the approach in opening up worldwide markets, smoothing power fluctuations, avoiding loss of the “excess” power of ‘green’ energy plants, and enabling power plants in remote areas and connecting them to new development in other remote areas. A more detailed examination of the economics and policy aspects of the concept must wait until a later paper, where we expect to show how the inclusion of these large-system aspects, typically neglected in engineering concept development, make all the difference here.
Top-Level Breakeven Cost Estimation
Several of the issues listed below are beyond the scope of this paper but may determine market feasibility.
1. Average power generation cost around the world a couple of years ago was between $0.04 and $0.08 per KWH.
However, this was dominated by the low cost of hydrocarbon fuels. Hydrocarbon costs have tripled.
Transmission costs are typically cited at 0.02 per KWH in the US, but may be substantially higher elsewhere
due to much higher loss percentages and land costs. Land values and environmental impact costs of power lines go up.
2. Interest rates on long-term debt fluctuate considerably.
3. The cost of carbon-based energy sources may be expected to rise further as environmental concerns impose
strict limits and penalties. The global need to switch to non-hydrocarbon sources justifies some level of public
(government) expenditure on infrastructure to enable ‘green’ power.
4. Several nations would trade their ‘green credits’ and buy power from nations that sell power generated through
whatever means – if they could get that power.
5. At present, high efficiency is not a driving consideration in microwave power beaming, since antenna size, mass
and cost are more important for transmission at low power values. A value for achievable conversion efficiency
at high power level is hard to find. Most experts appear to assume a 70% practical limit using present-day
technology, for open-system conversion, with 30% dissipated as heat at the conversion point. In a power plant
context, it may be assumed that a good portion of this can be recovered – we assume that enough can be
recovered to compensate for the 20% loss of transmission through the atmosphere and bring the loss rate to the
7 to 8% now lost in the transmission process. We expect that in future, dedicated plants to generate microwave
beams can be developed to minimize the conversion loss from raw sources. Thus, the transmission approach
that we propose only transmits 70% of the power generated.
The above items make it difficult to assign a value per KWH for power transacted through SPG at this stage of the
concept. For simplicity, we assume that the space-based power transmission stage has a value of $0.02 per KWH
transacted. To break even over the 20-year replacement period of SPG-stage satellites, the initial 36 satellites must
have at least a capacity of 136MW each. This assumes that all power beaming from earth is on the sunny side (only
18 sats in view) and that each ground station has a choice of 4 satellites for its beams at any time. A market size of
5,500 GWH per year of power transmission must be achieved. This appears quite reachable for an initial
consortium agreement. We project SPG launch cost of $6600 per kg to a 1,215km high orbit. Assuming satellite
mass of 1,500kg, and cost per satellite of $22M, there is enough left from an initial $1.7B bond issue to fund one
replacement satellite per year.
The most recent GPS satellite launches provide cost guidance on a government-developed dual-use system. The cost
per 2,036kg satellite is $44M, and the launch costs $50M (or $24,557 per kg). The orbit is over 1,0000km high.
The GPS model is inadequate as a reference, because of the huge user base and primary military application.
At the 200MW peak power level involved in the transmission through a satellite, even a 1% loss means 2 MW of
heat dissipation, aggravated by the small size and mass of these satellites compared to SSP designs. However, with
large thin receiver arrays attached by booms to the satellite, radiation transfer from the shadow side can suffice for
thermal management. The intermittent duty cycle, as opposed to the steady heating of SSP satellites in GEO, helps
in this process, but innovative means of recovering part of the power and efficiently rejecting the rest are needed.
Thus, increasing the efficiency of the satellite throughput is of utmost importance. However, waveguides today have
demonstrated near-100% throughput.
One key to advancing towards Space Solar Power is the technology of Direct Conversion using optical rectennae
that convert sunlight to DC, and possibly extensions of this where sunlight is directly converted to microwave. This
promises a large increase in efficiency (over 80% from DC as opposed to the present theoretical 40% of solar-cell
systems), and a large payoff from decreased converter mass from the present 1kg per kilowatt (Brown 1984). More
exciting is the prospect of converting sunlight to microwaves directly at 85% efficiency to microwaves. The technical barriers appear to be in nano-fabrication of antennae that can tune effectively to much of the solar
spectrum. Rapid progress is expected in this field in the coming decade (Berland 2003).
Space Power Grid Phase
An initial concept is to have 36 satellites orbiting 1,215 km above earth (determined to allow line-of-sight
transmission to a satellite 45 degrees away without atmospheric losses) and passing microwave beams between
earth-based locations in a real-time energy trade. Each craft can handle (receive and distribute) upto 200MW at 140
GHz. Each is assumed to be able to transact power with 4 other satellites and up to 100 ground stations. The craft
are sized to recover system deployment cost in 20 years from savings in costs of ground transmission, based on
current GPS satellite costs (as explained below). The satellite number will rise to 72 as locations far from the
equator join the system. Beaming losses to and from LEO are density-equivalent to those for 22 km of transit
through sea-level atmosphere. Hence the business case applies to new plants located over 22 km from their major
distribution hubs. Satellites in equatorial / tropical orbits will be capable of receiving more power from the ground,
and transmitting more power to outlying satellites. Satellites in inclined orbits will distribute more than they collect from ground.
Ground stations will be located at power plants to generate and beam microwaves, at ideal solar / wind collector
locations – dry high-desert locations where land is cheap and sunlight or wind abundant, and mountainous / coastal
ridge regions with high winds. The US Southwest, South Dakota, Hawaiian Islands, North African, Gobi, Thar and
Australian deserts and Greenland are examples envisaged. Receiving stations on the ground can be located almost
anywhere. There is no need to co-locate receiving stations with generator stations. Since receivers will be much
smaller in diameter than those envisaged for GEO-located SSP systems, a given utility company can place its
generators at optimal generator locations, and its receivers to best distribute power to end-user customers.
Direct-Conversion Augmented SPG
We project that Direct Solar Conversion to microwave beams will become feasible with 50% efficiency and reduced
mass by 2035. To replace current global production with solar energy at 50% efficiency, 5600 sq.km of solar
collector area in space (where solar intensity is 1GW/sq.km) is required. The breakthrough needed for this is nanofabrication technology, and the same will also permit better conversion and beaming efficiency. The SPG satellites
will then be replaced with 2,000kg–class Direct Conversion Augmented-SPG (DCA-SPG) satellites, with a 1km
diameter sun-tracking ultra light collector and converter on each adding 0.5 GW to the grid. As the number of
satellites increases, mean transmitting distances between satellites comes down, and hence the system efficiency
could be increased by going to lower frequencies with lower atmospheric losses for the same receiver size.
Full Space Solar Power Phase
GEO sun-sats launched in the 2040s will each have 100 sq.km ultra light collector/ reflectors that simply focus
sunlight onto the 1sq.km collectors of the DCA-SPG. Each is expected to add 50GW to the grid at 50% efficiency.
Thus the system of 72 LEO satellites and 72 GEO ultralight mirrors, with a 70% transmission efficiency, will
generate 90% of today’s global energy production. It is noted here that deploying large ultra-thin collectors with
high-intensity solar cell arrays is an alternative to any Direct Conversion technology, alleviating technological risk.
Since the system is in LEO, the launch cost is far below that of launching solar panels to GEO.