Pour some cold water on space photovoltaics

Recently, the concept of "space-based solar power" has suddenly become a hot topic in the capital market and the technology circle. Elon Musk's statement that "the United States should build 200GW of solar power, with 100GW on the ground and 100GW in space" sent the photovoltaic sector in the A-share market soaring, and the concept stocks hit consecutive limit-up boards. Domestic enterprises have also announced their plans to enter the field, with perovskite, gallium arsenide, and lightweight modules all being brought into the picture. Media headlines are getting more and more eye-catching: "A trillion-dollar blue ocean", "The next photovoltaic revolution", "Rescuing the over-competitive ground-based photovoltaic market"... It all sounds very exciting, as if the world's energy problems could be solved by space-based solar power tomorrow.

But today, we're going to pour some cold water on this hot concept: Space-Based Solar Power (SBSP) is still at a stage where "it sounds great, but it's extremely difficult to implement". It's still a long way from true large-scale commercialization. From multiple perspectives such as technology, economy, and engineering implementation, the bottlenecks it faces are not something that can be solved with a little more effort; they may not even be solved with ten times the effort. Let's break them down one by one.

1

Technologically: Sounds like science fiction, but full of bottlenecks in practice

The core idea of SBSP is to place solar panels in the Geostationary Earth Orbit (GEO), about 36,000 kilometers above the Earth. There is no day and night, no clouds, and no atmospheric attenuation, so the solar panels can generate electricity at full capacity almost 24 hours a day, 365 days a year. Then, the electricity is converted into microwaves or laser beams and transmitted back to the ground receiving station (rectenna), and then converted back into electricity and fed into the grid.

It sounds perfect, but in reality, every step has fatal problems.

Terribly low energy transmission efficiency

The conversion efficiency of ground-based photovoltaic modules can now reach over 25%, and in the laboratory, perovskite tandem cells can even approach 35%. However, the overall system efficiency of SBSP is only about 13% - 35% efficiency of solar cells → 90% for DC-DC conversion → 70% for DC-RF microwave conversion → 90% for antenna radiation → 98% for atmospheric transmission → 78% for the ground rectenna → 90% for the final DC-DC conversion for grid connection. In the end, 87% of the solar energy sent into space is wasted on the way. Another fatal problem with microwave transmission is beam spot diffusion. When a 2.45GHz or 5.8GHz microwave beam is transmitted from GEO to the ground, the diameter of the receiving station needs to be several kilometers to tens of kilometers to capture most of the energy. The intensity at the center of the beam spot is about 230W/m² (equivalent to 1/4 of the midday sun), but it attenuates severely at the edges, resulting in low overall collection efficiency. Not to mention interference factors such as rainy days, atmospheric turbulence, bird flocks, and airplanes passing through.

Enormous structure scale, extremely difficult to assemble and maintain

In NASA's reference design, a 2GW system requires a solar array with an area of 11.5 - 19 square kilometers and a total mass of 5,900 - 10,000 tons, equivalent to the mass of hundreds of International Space Stations. It's impossible to launch such a large structure all at once. Thousands of rocket launches are needed to send the modules into space, and then on-orbit robots will assemble them autonomously (ISAM). Currently, the on-orbit assembly technology has not even fully surpassed the level of the ISS, let alone assembling a structure 1,000 times larger. After assembly, regular maintenance is also required: radiation damage, micro-meteorite impacts, thermal expansion and contraction. Robots need to be sent into space every 10 - 15 years to replace parts or make repairs. In the space environment, every subsystem such as the precision of the robot arm, autonomous decision-making, energy supply, and debris avoidance needs to achieve unprecedented reliability. Not to mention the problem of orbital debris. The GEO orbit is already crowded, and adding a few giant structures several kilometers in size will exponentially increase the risk of collisions. Once out of control, it will be a disaster.

Safety and environmental hazards cannot be ignored

Although the designed intensity of the microwave beam is not high, there are still disputes about its long-term impact on the ecosystem, birds, insects, and the human body. The receiving station covers a large area (with a diameter of 6 - 10 km), and the surface temperature will rise slightly. How can it be compatible with agriculture and the ecosystem? The laser transmission scheme is even more extreme. With a high power density, accidental shooting or deflection could cause catastrophic consequences. Internationally, issues such as spectrum allocation, orbital slotting, safety standards, and cross-border liability division are all points of contention.

2

Economically: Extremely high cost, and the return is far away

Technical difficulties can be overcome gradually, but if it doesn't make economic sense, it's a dead end.

Launch cost is still the biggest obstacle

Currently, the launch cost (the target price of Starship) to Low Earth Orbit (LEO) is about a few hundred to a thousand dollars per kilogram, and it's even higher to GEO. NASA estimates that for a 2GW system, the launch cost alone accounts for 71 - 77% of the total lifecycle cost, and the total capital expenditure (CapEx) is as high as $90 - 137 billion. Even if Starship reduces the cost to $100 per kilogram, including maintenance, assembly, ground station construction, and operational depreciation, the Levelized Cost of Energy (LCOE) is still as high as $0.61 - 1.59 per kWh under the baseline scenario. In contrast, the LCOE of ground-based photovoltaic + energy storage is expected to be only $0.02 - 0.05 per kWh by 2050, a difference of 10 - 80 times! To reduce the LCOE of SBSP to the level of ground-based solar power, the following conditions need to be met simultaneously: reduce the launch cost by another order of magnitude (to $50 - 100 per kilogram), increase the efficiency of solar cells to 50%, extend the hardware lifespan to over 15 years, achieve an 85% learning curve for assembly, and reduce the cost of maintenance robots to an extremely low level... Only when all these conditions are met can it barely compete. NASA's sensitivity analysis shows that under the optimal combination scenario, the cost can be reduced by 93 - 95%, but that's already an "ideal state".

The investment scale and return period are off-putting to everyone

The total lifecycle cost (including maintenance and disposal) of a 2GW system is as high as $276 - 434 billion, equivalent to building dozens of Three Gorges Dams. The return period may start from 30 - 50 years. Private capital is most afraid of projects with "high risk, long cycle, and huge upfront investment". The World Economic Forum itself admits that the biggest obstacle is not technology, but "the structure and risk preference of private capital". Government subsidies? Governments around the world can't provide unlimited financial support. In contrast, the cost per watt of ground-based photovoltaic has dropped to $0.1 - 0.2 RMB. The industrial chain is complete, the iteration is fast, and the risk is low. Who would invest in space-based solar power?

3

Current progress: Still far from "commercialization"

Optimists always like to say that "technology is advancing rapidly". It's true that Caltech conducted the MAPLE experiment in 2023, transmitting 1kW of power over 50 meters with an efficiency of 60%. The European Space Agency's SOLARIS project will make a decision in 2025. China is also planning lightweight modules and on-orbit demonstrations. However, these are all laboratory-level, small-scale verifications, and there is still a huge gap between them and GW-level commercial applications.

A NASA report clearly states that under the baseline scenario, SBSP will not be competitive by 2050; it may only be on par with ground-based renewable energy under extremely optimistic assumptions.

The so-called "trillion-dollar space-based solar power track" hyped up in the domestic capital market is more driven by narratives and capital games. Most of the so-called "space-based solar power" modules being developed by enterprises are actually high-efficiency batteries for ground use, or small-scale products for military or satellite use. The lightweight, radiation-resistant, and ultra-high-efficiency modules for GW-level space power stations are still far from mass production.

Conclusion: A beautiful dream, but a harsh reality

The concept of SBSP is indeed appealing: 24/7 base-load clean energy, power supply to any location in the world, and no restrictions from weather and geography... If it succeeds, it will open a new chapter in the history of human energy. However, at present, it's more like a "distant science fiction bet" - there are numerous technical bottlenecks, the economic calculation doesn't work out, and the engineering implementation is still a long way off.

Today, ground-based solar power has already achieved a "levelized cost of energy lower than that of coal-fired power", the cost of energy storage is continuously dropping, and the distributed + smart grid is rapidly maturing. For SBSP to overtake on a curve, it will require at least 20 - 30 years of technological leap, a more than tenfold reduction in launch cost, and the full coordination of international rules. During this period, there may be capital frenzies, concept bubbles, and policy trials, but the construction of the first commercial GW-level space power station may not happen until after 2050, or even later.

So, friends, when faced with this new story of "space-based solar power", it's worth asking: Is it the next big thing in solar power, or the next nuclear fusion? The answer may be more brutal than expected.

This article is from the WeChat official account "Researcher", author: Researcher Jun, published by 36Kr with permission.