
Nearly 60 years after it was first proposed, space-based solar power is attracting private capital and its first commercial commitments. Power beamed from orbit could reach any site on Earth with land for a receiver, without a grid connection or a fuel supply. Whether that becomes a business depends on a short list of assumptions, and nobody has yet shown that they can all hold at once. This is our view of what they are, how realistic each one is, and what we want to see from the companies working on them.
In 1941, Isaac Asimov imagined a space station that collects sunlight and beams the energy down to Earth. In his short story Reason, a robot runs the station and refuses to believe Earth exists. Twenty-seven years later, Peter Glaser turned the fiction into an engineering proposal. In 1968, he proposed in Science collecting sunlight in orbit and beaming it to Earth, and he patented the design in 1973.
.png)
Space-based solar power (SBSP) collects solar energy in orbit and beams it as microwaves or laser light to receivers on Earth or other spacecraft, which convert it into electricity. These applications serve different customers and face different engineering challenges. This article covers both, with a focus on space-to-Earth (the space-to-space segment had a much less certain set of assumptions on the market size).
If it works, SBSP can put clean power almost anywhere, day or night. A site needs land for a receiver, but no grid connection and no fuel supply, so SBSP can compete with any power source on the ground and avoid grid bottlenecks. It matters most for defence, backup power, remote sites where the grid is unreliable and places the grid cannot reach fast enough, where even a price well above grid power can win. It can also power satellites in orbit. The challenge is delivering enough of it, reliably and at a competitive price.
Reusable launch has transformed space economics. The Space Shuttle used for the International Space Station cost approximately $54,500/kg to reach low Earth orbit, compared with around $2,700/kg for Falcon 9 and $1,500/kg for Falcon Heavy. Starship's proposed $100–150/kg remains a target, but it is no longer out of reach.
The tech is beginning to be proven. In 2023, Caltech's SSPD-1 mission tested solar cells, a deployable structure and a microwave beaming array in orbit. In 2025, DARPA transmitted more than 800 W of laser power over 8.6 km in a 30-second ground-based demonstration.
Surging demand for power across all sectors, driven mostly by data centres and decarbonisation, needs firm power in places the grid cannot reach or deliver fast enough. The IEA expects data centres alone to use about 1,200 TWh a year by 2035, half of all additional grid requests are coming from data centres desperate for power today; the grid queues are getting longer and longer, extending to over 6 years for new capacity.
Space based solar is one candidate, and investors are responding. Star Catcher raised a $65 million Series A led by B Capital to sell power between satellites. Overview Energy raised a $20 million seed round from Lowercarbon and Prime Movers Lab, and Reflect Orbital raised a $20 million Series A from Lux Capital and Sequoia. Aetherflux raised $50 million in 2025 to develop space-based solar power, then rebranded as Cowboy Space and raised $275 million last May. Its focus has expanded to include launch vehicles and orbital data centres.
In a space-to-Earth system, solar arrays in orbit generate electricity, and a transmitter sends it to the ground as microwaves or laser light. A microwave beam lands on a rectenna, a large field of antennas that converts it back into electricity. Laser systems use photovoltaic panels instead.
.png)
For decades, SBSP has referred to gigawatt-scale plants in geostationary orbit (GEO), where satellites stay above the same point on Earth and receive sunlight almost all year. NASA's assessment puts a 2 GW plant at more than 14 times the mass of the International Space Station. High upfront costs, slow progress, and long timelines have pushed developers towards other designs.
.png)
Newer designs split the plant into smaller pieces, in one of two ways. Modules can be assembled in orbit into one large plant, or a constellation can spread the capacity across many separate satellites. Lower orbits shorten the distance the beam travels, but a satellite there spends less time above a given receiver. Sustained service then needs a constellation, so another satellite can take over as each one moves away.
.png)
The destination of the power also shapes the technology. Space-to-space systems supply satellites or lunar assets, which need less power and suit compact laser receivers. Space-to-Earth systems move far more power, and most use microwaves, which pass through clouds and rain with little loss, while clouds absorb or scatter laser light. Overview Energy is one exception and plans to beam near-infrared light onto existing solar farms.
Can electricity from orbit compete on cost?
Not on the grid, and not yet. NASA puts space solar power at 12 to 80 times the cost of terrestrial alternatives per unit of electricity. In Caltech’s comparison, utility solar costs about $44/MWh and gas peaking plants about $169/MWh. Off-grid, the price is higher: the US Army pays about $400/MWh for power at its Arctic installations. That is where the first sales are most likely.
NASA's 2024 assessment estimates electricity costs of $610–1,590/MWh for its two reference designs. A UK government-commissioned study projects £335–595/MWh for a first-of-a-kind system in 2030, falling to £87–129/MWh for a mature plant in 2040. Caltech's model reaches $94/MWh after another decade of technological development. These estimates describe different systems, assumptions, timelines, and financing conditions, making direct comparisons difficult but they show what could be possible.
Launch is only one input. Satellites need to become lighter and cheaper to manufacture, transmission losses must fall, and equipment must operate reliably for decades. NASA's most optimistic scenario reaches $30–80/MWh only when several (almost all) favourable assumptions are combined.
Financing matters as much as hardware. In the UK study, a lower cost of capital alone cuts the cost of electricity by 42%. Unlike terrestrial solar, an SBSP plant requires substantial investment before delivering its first electricity.
For investors, the immediate question is whether smaller systems can generate revenue before the economics of utility-scale electricity are proven. Diesel sites can pay for the first systems, but each needs only megawatts, and there are few of them. To return a venture fund, a company has to reach the next tier, data centres and other grid-constrained sites, where demand runs to gigawatts.
The first customers will be those who pay for access to power where alternatives are expensive or unavailable. Two markets stand out: satellites that need more energy than their onboard solar arrays can provide, and terrestrial sites that depend on costly fuel deliveries.
In space, additional power could extend satellite lifetimes or increase computing capacity without adding larger solar arrays and batteries. Star Catcher reported seven commercial power purchase agreements by May 2026 and demonstrated 1.1 kW of optical power transmission on the ground. Its first orbital mission launched in October 2026, although commercial power delivery remains unproven.
On Earth, remote mines, islands and military installations could justify higher electricity prices than national grids. This offers a potential first market for space-to-earth transmission, provided the technology can compete with diesel generation and other local alternatives.
Meta has reserved up to 1 GW from Overview Energy, tied to a 2028 orbital demonstration, and Virtus Solis has signed a 20-year PPA with Brae Systems for 100 MW. But commitments are not proof of delivery. History provides a warning: Solaren (US) signed a 200 MW power purchase agreement with PG&E in 2009, promising electricity by 2016. The project never delivered, and the contract was dropped after financing failed.
SBSP’s climate value depends on the energy it replaces, and the first buyers replace the most carbon-intensive power there is. Diesel generators emit about 0.8 tonnes of CO₂ per MWh, more than three times the 0.24 tonnes of the 2035 grid in the IEA’s Stated Policies scenario. A gigawatt of space power delivered over a year would avoid close to 6 Mt of CO₂ at diesel sites, against 1.7 Mt on the grid. At full potential, SBSP supplies the firm power that gas and coal still provide to grids and data centres. The IEA expects data centres to use about 1,200 TWh a year by 2035, which at the 2035 grid mix is roughly 290 Mt of CO₂ a year (but given the grid constraints a lot of this could be from gas generators, see our other report on LDES).
Space-to-space power follows a different logic. Its climate value depends on how much power use moves to orbit, on which applications cheap power in orbit enables, and on how many launches it saves. A satellite that buys power on demand can carry less solar array and battery, and launch is the largest source of emissions (64 to 72%) in NASA’s lifecycle estimate for space solar power.
The design in orbit matters technically, but SBSP will succeed or fail on what it delivers on Earth. It has to deliver enough power where customers need it, at a price that beats their local alternative. For early customers at remote sites, that alternative is diesel.
Several parts of SBSP already work on their own. Space-grade solar cells convert 30–34% of sunlight into electricity, compared with about 24% for commercial solar panels on Earth. In orbit, Caltech's MAPLE array beamed about 0.25 W to receivers inside its own spacecraft, and its signal was detected on Earth. Nobody has yet delivered sustained, useful, (cost effective) power from orbit to the ground. For that, six things have to hold:
.png)
Until an integrated demonstration delivers sustained power from orbit with measured end-to-end efficiency, the economics of commercial SBSP remain theoretical
For Europe, SBSP is a question of strategic autonomy. ESA's SOLARIS programme has investigated its technical and economic feasibility, while European industry already has expertise in space photovoltaics, deployable structures and satellite engineering.
Other regions are moving towards demonstrations. The US combines private investment with defence-backed research, while China has proposed orbital tests and Japan continues to develop wireless power transmission. Europe has much of the technical foundation, but still faces a funding gap between research and commercial deployment.
This requires more than developing individual components. The ability to integrate solar arrays, wireless transmission, orbital structures, and ground infrastructure will be critical. Moving from research to orbital demonstrations will also require public funding and established industrial partners, as venture capital alone is unlikely to finance the earliest large-scale projects.
At World Fund, we believe SBSP must prove it can be a business in low Earth orbit (LEO) before any plan to move to massive structures in GEO. We look for companies that build world-class hardware, understand how energy is bought and sold, and can navigate a highly specialised aerospace ecosystem.
We need to see credible commercial demand, ideally through a priced agreement or paid pilot. Their cost models must account for realistic launch access, system mass, transmission efficiency, and the price customers are willing to pay.
We remain cautious about projects requiring billions in capital before their first sale and components whose commercial opportunity depends entirely on SBSP succeeding. Large pre-commercial demonstrations may require public funding and established industrial partners rather than venture capital alone.
.png)
Power from orbit is a big opportunity with many assumptions that need to materialise for it to be successful. The ones listed in this article are the ones we should consider before getting excited.
The question for space-based solar power is no longer whether it can be imagined, or even launched. It is whether every link in the chain can work at the same time: launch at $500/kg or less; solar arrays that are far lighter and cheaper; beams steered precisely at scale; at least a third of the power reaching the ground; assembly in orbit that runs itself; and land and spectrum on the ground. Each link has advanced on its own, but none has been shown working together, and today’s best cost estimates still put power from orbit above the price of diesel at a remote military base.
The first winners will close that gap with a smaller system in LEO that delivers measured power to a customer with no better option, long before anyone builds a gigawatt plant in GEO. The company that does this first will have turned a 60-year-old idea into an energy business. It will also be well placed to compete for the far larger market of data centres and grid-constrained sites.
If you are working on any link in this chain, or on finding the first paying customer, and you have results you can measure, please get in touch at robin@worldfund.vc and craig@worldfund.vc.

Nearly 60 years after it was first proposed, space-based solar power is attracting private capital and its first commercial commitments. Power beamed from orbit could reach any site on Earth with land for a receiver, without a grid connection or a fuel supply. Whether that becomes a business depends on a short list of assumptions, and nobody has yet shown that they can all hold at once. This is our view of what they are, how realistic each one is, and what we want to see from the companies working on them.
In 1941, Isaac Asimov imagined a space station that collects sunlight and beams the energy down to Earth. In his short story Reason, a robot runs the station and refuses to believe Earth exists. Twenty-seven years later, Peter Glaser turned the fiction into an engineering proposal. In 1968, he proposed in Science collecting sunlight in orbit and beaming it to Earth, and he patented the design in 1973.
.png)
Space-based solar power (SBSP) collects solar energy in orbit and beams it as microwaves or laser light to receivers on Earth or other spacecraft, which convert it into electricity. These applications serve different customers and face different engineering challenges. This article covers both, with a focus on space-to-Earth (the space-to-space segment had a much less certain set of assumptions on the market size).
If it works, SBSP can put clean power almost anywhere, day or night. A site needs land for a receiver, but no grid connection and no fuel supply, so SBSP can compete with any power source on the ground and avoid grid bottlenecks. It matters most for defence, backup power, remote sites where the grid is unreliable and places the grid cannot reach fast enough, where even a price well above grid power can win. It can also power satellites in orbit. The challenge is delivering enough of it, reliably and at a competitive price.
Reusable launch has transformed space economics. The Space Shuttle used for the International Space Station cost approximately $54,500/kg to reach low Earth orbit, compared with around $2,700/kg for Falcon 9 and $1,500/kg for Falcon Heavy. Starship's proposed $100–150/kg remains a target, but it is no longer out of reach.
The tech is beginning to be proven. In 2023, Caltech's SSPD-1 mission tested solar cells, a deployable structure and a microwave beaming array in orbit. In 2025, DARPA transmitted more than 800 W of laser power over 8.6 km in a 30-second ground-based demonstration.
Surging demand for power across all sectors, driven mostly by data centres and decarbonisation, needs firm power in places the grid cannot reach or deliver fast enough. The IEA expects data centres alone to use about 1,200 TWh a year by 2035, half of all additional grid requests are coming from data centres desperate for power today; the grid queues are getting longer and longer, extending to over 6 years for new capacity.
Space based solar is one candidate, and investors are responding. Star Catcher raised a $65 million Series A led by B Capital to sell power between satellites. Overview Energy raised a $20 million seed round from Lowercarbon and Prime Movers Lab, and Reflect Orbital raised a $20 million Series A from Lux Capital and Sequoia. Aetherflux raised $50 million in 2025 to develop space-based solar power, then rebranded as Cowboy Space and raised $275 million last May. Its focus has expanded to include launch vehicles and orbital data centres.
In a space-to-Earth system, solar arrays in orbit generate electricity, and a transmitter sends it to the ground as microwaves or laser light. A microwave beam lands on a rectenna, a large field of antennas that converts it back into electricity. Laser systems use photovoltaic panels instead.
.png)
For decades, SBSP has referred to gigawatt-scale plants in geostationary orbit (GEO), where satellites stay above the same point on Earth and receive sunlight almost all year. NASA's assessment puts a 2 GW plant at more than 14 times the mass of the International Space Station. High upfront costs, slow progress, and long timelines have pushed developers towards other designs.
.png)
Newer designs split the plant into smaller pieces, in one of two ways. Modules can be assembled in orbit into one large plant, or a constellation can spread the capacity across many separate satellites. Lower orbits shorten the distance the beam travels, but a satellite there spends less time above a given receiver. Sustained service then needs a constellation, so another satellite can take over as each one moves away.
.png)
The destination of the power also shapes the technology. Space-to-space systems supply satellites or lunar assets, which need less power and suit compact laser receivers. Space-to-Earth systems move far more power, and most use microwaves, which pass through clouds and rain with little loss, while clouds absorb or scatter laser light. Overview Energy is one exception and plans to beam near-infrared light onto existing solar farms.
Can electricity from orbit compete on cost?
Not on the grid, and not yet. NASA puts space solar power at 12 to 80 times the cost of terrestrial alternatives per unit of electricity. In Caltech’s comparison, utility solar costs about $44/MWh and gas peaking plants about $169/MWh. Off-grid, the price is higher: the US Army pays about $400/MWh for power at its Arctic installations. That is where the first sales are most likely.
NASA's 2024 assessment estimates electricity costs of $610–1,590/MWh for its two reference designs. A UK government-commissioned study projects £335–595/MWh for a first-of-a-kind system in 2030, falling to £87–129/MWh for a mature plant in 2040. Caltech's model reaches $94/MWh after another decade of technological development. These estimates describe different systems, assumptions, timelines, and financing conditions, making direct comparisons difficult but they show what could be possible.
Launch is only one input. Satellites need to become lighter and cheaper to manufacture, transmission losses must fall, and equipment must operate reliably for decades. NASA's most optimistic scenario reaches $30–80/MWh only when several (almost all) favourable assumptions are combined.
Financing matters as much as hardware. In the UK study, a lower cost of capital alone cuts the cost of electricity by 42%. Unlike terrestrial solar, an SBSP plant requires substantial investment before delivering its first electricity.
For investors, the immediate question is whether smaller systems can generate revenue before the economics of utility-scale electricity are proven. Diesel sites can pay for the first systems, but each needs only megawatts, and there are few of them. To return a venture fund, a company has to reach the next tier, data centres and other grid-constrained sites, where demand runs to gigawatts.
The first customers will be those who pay for access to power where alternatives are expensive or unavailable. Two markets stand out: satellites that need more energy than their onboard solar arrays can provide, and terrestrial sites that depend on costly fuel deliveries.
In space, additional power could extend satellite lifetimes or increase computing capacity without adding larger solar arrays and batteries. Star Catcher reported seven commercial power purchase agreements by May 2026 and demonstrated 1.1 kW of optical power transmission on the ground. Its first orbital mission launched in October 2026, although commercial power delivery remains unproven.
On Earth, remote mines, islands and military installations could justify higher electricity prices than national grids. This offers a potential first market for space-to-earth transmission, provided the technology can compete with diesel generation and other local alternatives.
Meta has reserved up to 1 GW from Overview Energy, tied to a 2028 orbital demonstration, and Virtus Solis has signed a 20-year PPA with Brae Systems for 100 MW. But commitments are not proof of delivery. History provides a warning: Solaren (US) signed a 200 MW power purchase agreement with PG&E in 2009, promising electricity by 2016. The project never delivered, and the contract was dropped after financing failed.
SBSP’s climate value depends on the energy it replaces, and the first buyers replace the most carbon-intensive power there is. Diesel generators emit about 0.8 tonnes of CO₂ per MWh, more than three times the 0.24 tonnes of the 2035 grid in the IEA’s Stated Policies scenario. A gigawatt of space power delivered over a year would avoid close to 6 Mt of CO₂ at diesel sites, against 1.7 Mt on the grid. At full potential, SBSP supplies the firm power that gas and coal still provide to grids and data centres. The IEA expects data centres to use about 1,200 TWh a year by 2035, which at the 2035 grid mix is roughly 290 Mt of CO₂ a year (but given the grid constraints a lot of this could be from gas generators, see our other report on LDES).
Space-to-space power follows a different logic. Its climate value depends on how much power use moves to orbit, on which applications cheap power in orbit enables, and on how many launches it saves. A satellite that buys power on demand can carry less solar array and battery, and launch is the largest source of emissions (64 to 72%) in NASA’s lifecycle estimate for space solar power.
The design in orbit matters technically, but SBSP will succeed or fail on what it delivers on Earth. It has to deliver enough power where customers need it, at a price that beats their local alternative. For early customers at remote sites, that alternative is diesel.
Several parts of SBSP already work on their own. Space-grade solar cells convert 30–34% of sunlight into electricity, compared with about 24% for commercial solar panels on Earth. In orbit, Caltech's MAPLE array beamed about 0.25 W to receivers inside its own spacecraft, and its signal was detected on Earth. Nobody has yet delivered sustained, useful, (cost effective) power from orbit to the ground. For that, six things have to hold:
.png)
Until an integrated demonstration delivers sustained power from orbit with measured end-to-end efficiency, the economics of commercial SBSP remain theoretical
For Europe, SBSP is a question of strategic autonomy. ESA's SOLARIS programme has investigated its technical and economic feasibility, while European industry already has expertise in space photovoltaics, deployable structures and satellite engineering.
Other regions are moving towards demonstrations. The US combines private investment with defence-backed research, while China has proposed orbital tests and Japan continues to develop wireless power transmission. Europe has much of the technical foundation, but still faces a funding gap between research and commercial deployment.
This requires more than developing individual components. The ability to integrate solar arrays, wireless transmission, orbital structures, and ground infrastructure will be critical. Moving from research to orbital demonstrations will also require public funding and established industrial partners, as venture capital alone is unlikely to finance the earliest large-scale projects.
At World Fund, we believe SBSP must prove it can be a business in low Earth orbit (LEO) before any plan to move to massive structures in GEO. We look for companies that build world-class hardware, understand how energy is bought and sold, and can navigate a highly specialised aerospace ecosystem.
We need to see credible commercial demand, ideally through a priced agreement or paid pilot. Their cost models must account for realistic launch access, system mass, transmission efficiency, and the price customers are willing to pay.
We remain cautious about projects requiring billions in capital before their first sale and components whose commercial opportunity depends entirely on SBSP succeeding. Large pre-commercial demonstrations may require public funding and established industrial partners rather than venture capital alone.
.png)
Power from orbit is a big opportunity with many assumptions that need to materialise for it to be successful. The ones listed in this article are the ones we should consider before getting excited.
The question for space-based solar power is no longer whether it can be imagined, or even launched. It is whether every link in the chain can work at the same time: launch at $500/kg or less; solar arrays that are far lighter and cheaper; beams steered precisely at scale; at least a third of the power reaching the ground; assembly in orbit that runs itself; and land and spectrum on the ground. Each link has advanced on its own, but none has been shown working together, and today’s best cost estimates still put power from orbit above the price of diesel at a remote military base.
The first winners will close that gap with a smaller system in LEO that delivers measured power to a customer with no better option, long before anyone builds a gigawatt plant in GEO. The company that does this first will have turned a 60-year-old idea into an energy business. It will also be well placed to compete for the far larger market of data centres and grid-constrained sites.
If you are working on any link in this chain, or on finding the first paying customer, and you have results you can measure, please get in touch at robin@worldfund.vc and craig@worldfund.vc.