Space-based solar station: helping to deliver net zero
Image credit: Space Energy Initiative
Could satellites send solar energy to Earth? E&T talks to two men who say ‘yes’.
One of the most pressing questions facing the world today is that of how to ensure a reliable source of energy that comes without carbon emissions or the ‘intermittency’ associated with existing sources of renewables. According to Martin Soltau it is a question with an answer that is clearly visible every day: the Sun.
Soltau, currently business manager for aerospace with engineering consultancy Frazer-Nash, is not arguing for increasing areas of the Earth’s surface to be covered in solar panels. Instead he is a proponent of a much more radical idea. This is that massive satellites that could ultimately measure several kilometres across be put into orbit to convert sunlight into electricity in a highly efficient manner free from atmospheric interference and then beam it down to Earth in the form of microwave radiation for conversion back into usable electric power.
Nor is Soltau vague about what such a satellite might involve. The core figures he cites are a weight of 2,000 tonnes, a diameter of 1.7km and geosynchronous orbit at a height of 35,786km above the Earth generating a little above 3GW of power. That would enable a high-frequency microwave beam somewhere within the ‘atmospheric window’ of 1-10GHz to be directed with extreme precision down to a ground station guided by a pilot beam emitted from the ground.
The receiving station itself or rectenna would measure 6.7km x 13km and would absorb power at a rate of around 245W/m2. It would be fabricated, states Soltau, from “millions of dipoles” supported by a lightweight net structure that could easily be strung between the towers of an existing offshore windfarm to minimise its visual impact. That suggested location would, he adds, have the additional benefit of enabling the satellite-generated power to be fed into the grid by an existing connection.
The result, given an overall system efficiency of around 60 per cent, would be the supply to the grid of some 2GW of continuous power. The lost energy would be almost entirely that of heat emitted from the satellite itself. “There would be virtually no loss in transmission,” states Soltau. “You might get 2 per cent loss in heavy rain.” In other words, a satellite-based system of that scope could be “the equivalent of a nuclear power station”.
Soltau is also keen to point out that the area covered by that rectenna would still only be about 40 per cent of that required for a conventional terrestrial solar array capable of generating a comparable input to the grid. Even then there would still be a further much more fundamental contrast between space-based solar power and any terrestrial counterpart. “Space-based solar power can form part of the baseload supply because it will provide power continuously, whereas terrestrial solar is intermittent,” he states. The maximum beam intensity would still only be about one-quarter of maximum intensity of sunlight at the Equator. So there would be no danger, for instance, to birds or other wildlife or indeed people who found themselves exposed to it.
The idea of space-based solar power (SBSP) is, Soltau insists, not at all fantastical but with the appropriate investment entirely practicable. Indeed, March 2022 saw the formal launch of the Space Energy Initiative (SEI), a UK-led partnership of some 50 organisations across government, industry and academia specifically to further an ambition that could, he says, become reality as soon as the next decade. A commercial venture, Space Solar, has now been set up with Soltau acting as CEO as well as continuing in his Frazer-Nash role.
Soltau explains that the advent of the SEI was facilitated by the publication just over a year ago of a research project into the feasibility of SBSP commissioned from Frazer-Nash by the UK government’s Department for Business, Energy and Industrial Strategy. He says that initial project “looked at the technical and economic feasibility of the concept and asked whether it could it be developed in time to help deliver substantial amounts of energy by 2050”. In both cases the answer, he states, was an emphatic “yes”.
What has changed to make that possible is not any great advance in the technology for the in-space generation and transmission of electric power, nor in the on-Earth systems for its capture and supply to the grid. Soltau notes: “It has long been considered technically viable; it has just been seen as too ambitious economically.”
Instead the key recent change, Soltau argues, has been the massive reduction in launch costs brought about by the latest generation of reusable rocket vehicles. He pinpoints in particular those produced by the SpaceX company founded by US entrepreneur Elon Musk. “The cost of launches has come down by about 90 per cent with the advent of reusable launchers like Falcon Nine and SpaceX Starship,” he observes. “We would need 68 Starship launches to get all the parts for a 2GW satellite into orbit.”
The satellites themselves would then be fabricated robotically while in orbit, a task facilitated by their being highly modular in form so that they would comprise large numbers of identical parts. “There are new concepts for solar-powered satellites that are lighter in mass and more modular than those in use today,” Soltau confirms. “But they still all use today’s technology, so they don’t need to break any laws of physics or use new materials, though it would still be a huge engineering integration challenge.”
The practical financial implication of this change in the overall economics is spelt out in the Frazer-Nash report. The total development cost required to make a genuinely commercial system a reality by the more ambitious date of 2040 would be £16.3bn. Soltau adds that work would be likely to need a kick-start with about £300m of probably public-sector cash over the first five years, after which largely private investment should be able to take over.
In fact, Soltau says, the SEI is thinking in terms of advancing even that shorter timescale. “We think we can get to 2GW in orbit in 2035 if we could give the go-ahead from beginning of 2023,” he says. “Then because of the way they are built on a production line we could commission one new 2GW satellite every year.”
A major reason for that optimism is that the SEI already has a preferred satellite design. Moreover, it is confident it is a design that will not only provide the efficiencies promised by the basic concept of space-based solar power, but also of the flexibility and scalability in deployment to make full-scale commercial implementation a genuinely practicable proposition.
This is CASSIOPeiA (Constant Aperture, Solid-State, Integrated Orbital Phased Array), a concept originated by small UK company International Electric based at Harwell, Oxfordshire, and perhaps more specifically by chief engineer Ian Cash. In terms of its basic geometry, CASSIOPeiA uses what Cash describes as a unique central helical array of photovoltaic cells onto which sunlight is directed by banks of mirrors positioned at either extreme of the structure. But, as he explains, what follows from this configuration is a whole series of attributes.
“Essential points about it include that it is solid state and so has no moving parts,” explains Cash. “But it can also do what the textbooks say you can’t do with a phased array, which is steer the beam through a full 360 degrees.”
Cash is happy to provide more details. “The mirrors are angled at 45 degrees to reflect sunlight down onto the array, which is split into a very large number of layers. For a geostationary application there would be about 60,000 layers and the microwave beam would be emitted from the gaps between layers,” he says. “The innovation in the array design is that each element comprises multiple transmitting elements that can steer their radiation pattern through 360 degrees. So, when you put lots of them on this helical geometry you can steer the whole beam through 360 degrees.”
According to Cash this provides his design with a crucial performance advantage over rival concepts. “The key thing about CASSIOPeiA is that it is Sun-facing not Earth-facing,” he states.
Cash explains that a fundamental challenge in the design of any satellite intended for an SBSP application is “to overcome orbital mismatch between having to look at a particular point on the Earth while always looking at the Sun”. He says that in previous designs this issue has been resolved “by having a large flat microwave phased array transmitter that always faces the Earth and a series of rotating or otherwise articulated mirrors that reflect sunlight onto solar cells on the rear of transmitter”. It is, he concedes, “a good design but it has high mass and large axes, which means it has got a high moment of inertia”.
This would not be the case even with a full-size CASSIOPeiA satellite involving a helical array about 1.7km in diameter and 1.7km in height with an overall length including the mirrors of about 10km. “The mirrors would only be about 5 per cent of the mass, with most of the mass concentrated on the helical array,” says Cash.
This difference in geometry also has important implications for the type of orbit that is possible. As Cash explains, putting a satellite with moving mirrors into an elliptical orbit with necessarily variable speeds would generate “a huge momentum issue as the mirrors are manoeuvred”.
This is precisely where CASSIOPeiA, with its complete lack of moving parts, provides a distinct competitive edge because its design makes it suitable for lower altitude, variable speed elliptical orbits as well as much higher geostationary ones. In turn this means that relatively small satellites can be deployed in such orbits as an intermediary step towards the much larger ones intended for geostationary applications.
Cash says that what is generally regarded as a low Earth orbit (LEO) at a height of about 550km would not be a useful place for power beaming because of the short contact time it would allow between the satellite and a constant point on the ground. Typically, he says, that might be just “12 minutes with a repeat time that could be more than one day”. The consequence would be a requirement for a very large number of satellites, “maybe thousands”, which is what happens with LEO communications satellites.
An elliptical orbit, in which a satellite spends the greater part of the time at the highest point or apogee because at that point it is moving relatively slowly, is a practicable alternative. It means, says Cash, that in effect a satellite “could loiter over a particular area”. Moreover, at the lowest point of the orbit, or perigee, it is moving much faster than a satellite at same height in a circular orbit and so would spend comparatively little time in that non-productive location.
For CASSIOPeiA, Cash envisages an initial application involving a four-hour elliptical orbit with a ground track across the tropical regions of the Earth. He says that would entail “five satellites in four-hour orbits with an apogee of about 12,000km delivering to four ground stations. Perigee would be the same as LEO.”
That flexibility would also facilitate the deployment of a small-scale prototype in an elliptical orbit to prove the concept before any commercial application was attempted. Indeed, Cash indicates that this might be possible on a surprisingly short timescale.
“If we got the go-ahead tomorrow then we could get a representative CASSIOPeiA satellite delivering power to the ground in five to six years,” he states. He suggests that such an initial test deployment would involve a mildly elliptical orbit varying between 550km and 2,000km with perhaps 20 minutes of contact time at apogee and six minutes at perigee, delivering 180MW of power.
“It obviously wouldn’t be a commercial satellite but it would demonstrate all the necessary technologies to prove we can do what we say we can do,” he says. Providing that first deployment went well, Cash estimates that “at most nine years would be needed to get first commercial production and maybe 12 years to get large-scale production”. The first commercial application would then be that of five satellites in elliptical orbits to four locations.
In other words, genuinely commercial space-based power generation could be a reality in the middle of the next decade.
Not too far beyond the next decade the prospects for space-based solar power seem almost astonishing. The reason, as Ian Cash explains, is its ability to use high-concentration photovoltaic (HCPV) chips. This technology is impractical for terrestrial use not least because of the very narrow acceptance angle for sunlight of just a few degrees it requires, but in space-based application fixed mirrors can maintain the required accuracy with absolute reliability.
Then the specific advantages the technology offers can be exploited. Firstly, a single layer of HCPV is twice as efficient as conventional PV with laboratory testing in 2020 achieving “47.1 per cent efficiency”. Secondly, HCPV also provides “microscale concentration at each layer”.
Cash says the best current terrestrial solar power generation is “one Sun at about 23 per cent efficiency”. HCPV cells with 250 layers on a satellite with two banks of mirrors could make feasible power generation equivalent to that from 500 Suns, he says. Moreover, “for the same annual energy delivered, HCPV requires 1/7,400 of the PV material area compared to a solar farm such as Cleve Hill in the UK”.
Cost calculations for space-based power
How expensive might space-based solar power be – at least initially? Martin Soltau says that the Frazer-Nash study developed a cost model for the Levelised Cost of Electricity (LCOE) that five 2GW satellites in operation by 2040 might incur. This would cover all aspects of production, launch, assembly, operation and decommissioning.
The figure of £50/MWh would be more than that for large-scale solar, offshore wind and onshore wind at, respectively,
£33/MWh, £41/MWh and £44/MWh. But it would be significantly less than that for gas turbine with associated carbon capture, nuclear PWR and dedicated biomass at, respectively, £82/MWh, £96/MWh and £98/MWh.
Crucially, though, Soltau says this estimate assumes a very high “investment hurdle rate” compared with other renewable technologies of 20 per cent because of the present low maturity of the technology. But as the technology matures the LCOE for space-based solar power will inevitably be reduced, he says.
Global SBSP Distribution via CASSIOPEIA In Highly Elliptical Orbits
CASSIOPeiA proponent Ian Cash says its solid-state design and constant Sun-facing attitude make it very suitable for highly elliptical orbits (HEOs).
A small constellation of four satellites on 8-hour HEOs (same ground track), each massing 1,400 tonnes, would be able to dispatch and distribute 1.4GW simultaneously and flexibly to three regions spaced by 120° in longitude, 24/365.
This also works at 180-420MW scale (300-650 tonnes) in 4-hour HEOs, allowing, for example, Calgary, London and Sapporo, or Santiago, Cape Town and Sydney, to share 51 power-beaming hours each day, with the freedom for any one city in each group to receive 24-hour base load.
With this concept, even small nations or cities could partner together for a lower-cost entry-route into space-based solar power, with options other than a graveyard orbit at end-of-life.
These 4-hour and 8-hour orbits can be designed to avoid the harshest Van-Allen radiation, loitering in environments far less severe than geostationary orbits.
Sign up to the E&T News e-mail to get great stories like this delivered to your inbox every day.