The true cost of “solar power at night” with Reflect Orbital

Here on Earth, humanity’s global energy needs only seem to increase over time. A combination of increasing populations, the widespread development of heating and cooling, a reliance on modern electronics, and the introduction of new energy-intensive technologies (such as the blockchain, smart technology, and artificial intelligence) are among the factors driving our rising energy needs. Sure, we can always build more power plants, but what about the simple solution of increasing the efficiency and production of already-existing plants, particularly the ones that only see part-time usage: wind and solar. Wind power doesn’t work when the air is still, and solar doesn’t work during the night.

Or can it?

That’s something that the US-based startup, Reflect Orbital, wants to change. The idea is to produce, as they put it, “sunlight on demand” — and light pollution by design — with plans to launch thousands (or possibly hundreds of thousands) of reflective satellites that would use mirrors to beam sunlight, collected in low-Earth orbit, onto the locations of already-existing solar power plants on Earth. By shining this light onto the power plants, even at night, they could cause these plants to generate extra power.

However, there are a series of costs that the company itself doesn’t consider: from costs to human safety and human health to ecology, astronomy, and the basic safety of the low-Earth environment. Here’s what the true cost of “solar power at night” would truly be.

This diagram shows the energy budget of Earth, with incoming and outgoing radiation (values are shown in W/m^2). Satellite instruments (CERES) measure the reflected solar, and emitted infrared radiation fluxes. The energy balance determines Earth’s climate and temperature. When the Sun is directly overhead, atmospheric absorption is minimal, allowing for the best surface measurements of incident solar radiation on Earth.

During a clear, sunny day, when the Sun is close to directly overhead (and, therefore, when its rays filter down through the atmosphere to reach the surface with the maximum possible intensity), every square meter will receive about 1360 watts of power incident upon it: 1360 W/m². The energy from that light can be:

• converted directly into electricity, with a device like a photovoltaic solar panel,
• or reflected onto a tank of water, causing the water to boil, creating steam and turning a turbine, which then creates electricity,

and that electricity can then either get distributed and be used on-demand or can be stored in a battery for use later.

The more angled the Sun is, however — i.e., the farther from the zenith it is — the less power-per-square-meter the ground will receive, as the incoming sunlight both gets spread over a larger portion of the curved Earth and must pass through more and more of the Earth’s atmosphere, reducing the incident power. Once the Sun dips below the horizon, which is the case in all the hours between dusk and dawn, no solar power is generated at all.

But what if there were an array of mirrors in space, capable of reflecting that light back down onto Earth, even when it’s night? That’s the key idea behind the proposal of Reflect Orbital.

This aerial photograph of San Francisco at night shows the effect of artificial lighting as well as the light pollution coming from the sky itself. While most light pollution on Earth now comes from luminous sources on the ground, a mirror-based project like Reflect Orbital could create greater amounts of artificial luminance than even a major city like San Francisco can.

The start would be one test satellite named Earendil-1, 18 meters in diameter, which would be launched in 2026. This proof-of-concept would position this satellite hundreds of kilometers above Earth’s atmosphere — at a purported altitude of 625 km above the ground — and would then reflect the sunlight it collected down towards Earth, where it would shine on one particular, targeted patch of the ground: where infrastructure for generating solar power already exists.

By 2030, they’d seek to have 4000 satellites in orbit, of larger size (about 54 meters, apiece, in diameter, or nine times the area of the test satellite, according to the company’s founder) all working together to shine and focus their reflected light onto various solar power plants around the world. The goal is, with so many satellites, to deliver at least 200 W of power — or just 15% of the midday Sun’s total power — to existing solar farms.

And that’s just the beginning. Their founder, in an interview here, discussed the possibility of having a total of 250,000 satellites in orbit, some 600 km up. To understand the impact that these various satellite proposals would have on Earth:

• the one 18-meter test satellite,
• the 4000 54-meter planned satellites,
• or the 250,000 satellites of 54-meters apiece,

we can do a little math, and scale the results of their balloon-borne test flight from 2024.

In summarizing the results from this test flight, founder Ben Nowack brought a flat, square mirror of roughly 2.5 meters in diameter (6.25 m², or just under 2% of the area of the 18-meter diameter test satellite they’re planning to launch in 2026) and reflected that sunlight to a location 242 meters away. In one instance, they measure 516 watts-per-square-meter of reflected sunlight: a huge amount that’s nearly half of what the midday Sun produces!

However, if we attempt to scale this the way the Reflect Orbital team intends, we find that putting

• an 18-meter diameter reflector (about 7 times the diameter, or 50 times the area, of the test mirror)
• or even a 54-meter diameter reflector (about 22 times the diameter, or 467 times the area, of the test mirror)
• at a distance of around 625 km from the target (more than 2500 times the distance of the test mirror),

that the incident power would not be around half that from the midday Sun, but only about 0.04 W/m² for the 54-meter mirror and about 0.004 W/m² for the 18-meter mirror.

Because you need approximately 200 W/m² of incident power to make a solar power plant produce any amount of useful energy, that implies you’d need thousands of these satellites working together to produce any sort of meaningful power: about 5000 of the 54-meter variety and about 50,000 of the 18-meter variety.

This illustration shows an unfurled solar sail: a large, thin, reflective surface. Replacing this solar sail with an Earth-facing mirror would lead to catastrophic and irregular light pollution across Earth’s surface at night, with a 54-meter diameter mirror outshining the full Moon by more than a factor of 10.

You also run into the problem of orbital dynamics: to keep satellites up in orbit, they have to move quickly around the Earth: about 7.5 km/sec (or 4.7 mi/sec), or fast enough to orbit the entire planet once every 90 minutes. To a static location on the ground, as computed by astronomers Michael J.I. Brown and Matt Kenworthy, that means each satellite will be within 1000 km of a given location on the ground for no more than 210 seconds (or 3.5 minutes) per orbit. And you’d need thousands of them, all reflecting sunlight to the same location on the ground, to generate any power at all.

Even with all 250,000 satellites that Reflect Orbital’s founder posits could exist — a number that’s dozens of times greater than all active satellites, combined, and that even exceeds all the active and inactive satellites plus all large pieces of space junk put together — this vast array could only deliver about 20% of the midday Sun’s light to approximately 80 locations on Earth at once. This would enable the energy production of the equivalent of 16 additional fully functioning solar power plants at night, in addition to all the solar power plants that function normally during the day.

That’s the full, most ambitious value-add that Reflect Orbital has to offer.

This array of solar panels is now a solar power plant in Minnesota, alongside an old farm field. While a sufficient amount of light could be focused onto the panels even at night, resulting in the production of some power, that light would also fall onto the adjacent crops, disrupting their natural day/night cycle.

But then, there are the costs to their plan as well. Let’s start by examining just the test satellite: the 18-meter sized Earendil-1, which remember, would be only about ⅑^th the scale prototype of a fully functioning mirror that was part of the proposed satellite megaconstellation array. A satellite of this size, composed solely of a mirror, would work to reflect sunlight down to the ground. Immediately, we must ask questions like:

• how large of a “spot” of light would it generate on the ground,
• how bright would that spot of light appear,
• what would happen to someone looking with the naked eye at the satellite,
• and what would happen to someone looking through a telescope if this mirror passed into its field of view?

The first thing we have to start at is the nature of the Sun: it’s not a “point” of light the way a distant star appears to us, but rather a disk to us, one with an angular diameter of about half-a-degree. At an altitude of 625 kilometers, an 18-meter reflector — even if optimally curved to focus that light down onto the Earth’s surface — would create a spot of light just-under 6 km across on Earth’s surface. This size won’t change as we scale the size of the mirror, but larger mirrors (as well as more mirrors) would increase the brightness.

Each satellite that we can see is only visible from Earth because of reflected sunlight. Even if a mirror were perfectly designed to focus that sunlight down to a point on Earth’s surface, however, it would still correspond to a size on the Earth’s surface that reflected the Sun’s angular size of about half-a-degree. For a satellite altitude of around 600 kilometers, that translates into an illuminated circular area on Earth’s surface of around 6 km in diameter.

For one 18-meter satellite, this corresponds to a power-per-unit-area that we already calculated: about 0.004 W/m². Compared to the midday Sun, which produces 1360 W/m², it’s only 1/340,000^th of the Sun’s power. However, the Sun appears 380,000 times as bright as the full Moon, meaning that even for the first test satellite Reflect Orbital is planning on launching, being in the path of that mirror’s reflected light means the satellite will appear brighter than the full Moon. Only, instead of that light coming from a large, extended disk spanning half-a-degree on the sky, like the full Moon does, that light would all be concentrated into an 18-meter-long mirror at a distance of 625 km: about 6 arc-seconds, or ten times smaller than the limit of what the human eye can resolve.

If you looked through a telescope at that mirror as it was reflecting light down onto you, you’d see something that was of comparable brightness to looking directly at the Sun through a telescope: enough to fry a human retina, or the optics and optical coatings within the lenses of a telescope even if no human were looking through it. With larger, 54-meter mirrors, the problem becomes ten times worse, as each spot of light is ten times brighter while still appearing point-like to the unaided human eye.

Moreover, the 6-or-so kilometers surrounding each solar plant wouldn’t be the only areas affected, because the mirrors must rotate (and hence, the reflected beam of light that they generate moves as well), causing every generated spot-of-light on the ground to slide across the Earth’s surface to the next, adjacent power plant. Everyone not just within a few kilometers of a solar plant, but who lives in the spaces between solar plants, would be at risk.

A single proposed Reflect Orbital satellite, even the smaller (18-meter) prototype, will illuminate a ~6-km area on Earth’s surface and shine with an intensity brighter than the full Moon. If an astronomical observer looked at it through a telescope or even binoculars, it could lead to permanent eyesight damage, similar to looking at a partial solar eclipse through a telescope.

Again, a single Reflect Orbital satellite, even at the full scale of 54-meters in diameter, is far too low-power to activate a solar power plant; it would take thousands of them all acting at once. With thousands of such satellites up there, even engaging in amateur astronomy as a hobby — looking at the night sky through telescope or binoculars — would suddenly become an extremely risky activity. Looking at even one of these satellites, even briefly or accidentally, through a magnifying device like a telescope or binoculars could cause permanent eye damage, similar to looking at a magnified partial solar eclipse.

Pilots in the air and drivers on roads are also at risk from an array of very bright satellites such as this: both in terms of distractions and in terms of the disruption of their night vision with bright flashes of lights. Similar to how laser strikes on aircraft are disallowed for safety reasons, these satellites would pose particularly increased dangers near airports.

And the exposure to bright artificial flashes of light, even for milliseconds, can not only disrupt human circadian rhythms, leading to less and lower-quality sleep, but has also been linked to certain types of cancers in humans.

The collision of two satellites can create hundreds of thousands of pieces of debris, most of which are very small but very fast-moving: up to ~10 km/s. If enough satellites are in orbit, this debris could set off a chain reaction, rendering the environment around Earth practically impassable. It is estimated that a single large-area satellite, one 50 meters in diameter, would encounter over 100 micrometeoroid strikes each year in low-Earth orbit.

Ecologically, we’re poised to further disrupt the already-disrupted natural day-night cycles of light and darkness that all life on Earth has adapted to over billions of years. Unlike the light pollution of cities or the low-intensity light pollution of artificial satellites, however, the continual, periodic array of bright flashes from space that Reflect Orbital satellites would create has the potential to disrupt something new: the last untouched, unpolluted, pristine ecosystems on Earth, as well as the agricultural crops grown around the world to feed most of humanity.

Studies have already shown that pollination and plant growth cycles would be disrupted by the amount of light that Reflect Orbital satellites would produce, in addition to the disruptions of bird migrations and the behaviors of animals under both laboratory and wild conditions. But the potential catastrophe of famine, brought about by an untested planet-wide experiment to only slightly increase our electrical production capacity, seems like a cost-heavy, benefit-light endeavor.

In addition, the full-scale plans for Reflect Orbital, of 250,000 total satellites, will not only catastrophically increase the risk of satellite collisions and Kessler syndrome, but at an altitude of 625 kilometers, each reflector can expect to encounter hundreds of micrometeoroid and debris impacts annually: degrading the reflector and creating larger, more diffuse beams, which will then be less-useful for their designed purpose of powering solar plants.

The debris from many sources, including rocket stages and de-orbiting satellites, winds up re-entering Earth’s atmosphere and depositing its contents into our atmosphere. The more satellites we put up into low-Earth orbit, today, the greater the amount of inevitable pollution we’re going to have to deal with and mitigate in our atmosphere in the future.

Environmentally, these satellites will also pose an incredible new type of pollution: the pollution that will arise when these satellites re-enter Earth’s atmosphere. Already, we have 1-to-2 Starlink satellites re-entering Earth’s atmosphere daily, adding significant amounts of aluminum to our upper atmosphere and affecting cloud seeding, with unknown downstream effects. With large mirrors, we can fully expect some reflective metal like silver or aluminum to be part of the mirror’s construction, and what goes up into low-Earth-orbit will soon come back down: depositing all of those metals into our upper atmosphere. The long-term effects have not even been quantified, but we can be certain where these materials will inevitably end up: in our atmosphere.

And finally, although most will likely argue this is the “least important” of the consequences for planet Earth, there will be a catastrophic effect on both ground-based and space-based astronomy. Any satellite as bright as the Reflect Orbital satellites would disrupt ground-based optical astronomy observing programs due to the scattering of light along the beam, restricting deep observations to only the brightest stars and planets: less than 10% of all astronomy research. Even a brief exposure to the direct light from such a satellite could damage the very sensitive (and expensive) camera equipment used, as it’s designed to study faint celestial objects. And yes, we still need ground-based astronomy, as it performs superior science in a number of ways to what space-based astronomy is good for.

The most unfortunate part of the Reflect Orbital proposal is how unnecessary it is. We already have solutions to all of the problems that it purports to solve, with none of the many significant consequences that an array of giant mirrors would bring to Earth. But sadly, the days of the “wild, wild west” still continue when it comes to the environment surrounding our planet: low-Earth orbit. Unless governments from across the world, or the US’s FCC, take decisive action, this new type of pollution and the consequences that come from losing our dark night skies will affect us all: not just astronomers, but food producers, consumers, and all humans who rely on Earth for our sustained survival.

The author acknowledges Samantha Lawler and Matt Kenworthy for valuable information in constructing this piece.