Tons of Mirrors βοΈπ°οΈπͺπβ‘
Ben Nowack is Solving Solar Energy's Nighttime Problem
EXO explores the invisible realm that lies just beyond human perception. These are often dense areas of inquiry, that call for editorial deep-dives that take forever to write. This cumbersome production timeline comes with a couple consequences: EXOβs publishing cadence sits somewhere between lop-sided and non-existent. Also, too many great stories come across my desk never to see the light of day.
To increase publishing frequency and cover more ground, Iβve developed a fleet of lighter-lighting formats. The first is Interface, conversations with folks hellbent on making out-there ideas a reality. This is that!
Now, perhaps a dangerously sleek graphicβ¦
This Tweet caught my eye last week:
My first reaction was who on planet fricking Earth is crazy enough to try to pull this one off? So I found out.
Ben Nowack is a 26-year old inventor and entrepreneur. After a mechanical engineering degree, a raft of gigs, and an appearance on Mythbusters he is now the CEO of Tons of Mirrors. Tons of Mirrors is using satellite-mounted reflective surfaces to redirect sunlight to earthbound solar panels at night.
Nowack didnβt invent the idea of using space mirrors to alter localized lighting conditions. A Senate Subcommittee on Energy Production and Supply proposed a similar concept in 1977. In 1988, the Russian Znamya Project successfully deployed a 65-foot mirror into orbit, sending a five kilometer-wide roving beam of light traipsing across Europe. In 2017, a Norwegian bookkeeper named Oscar Kittilsen erected large rotatable mirrors on a mountain above the town of Rjukan to combat seasonal depression. Chinaβs βthree moons" project is building giant orbiting reflectors to replace streetlights. The University of Glasgow is also developing an orbiting solar reflector concept called SOLSPACE.
Nowackβs coming at the challenge from a different angle though, with refreshed technology and a claim to make sunlight at least 90.7x cheaper than previous proposals.
EXO: Whatβs your backstory?
Ben Nowack: I've always been a builder and a maker. I built a fusion reactor in high school. I built an Underwater ROV when I was 14. I worked with Peter Brown from the Science Channel when I was a sophomore. We built these cameras that would livestream video from underwater. He flew me to Hawaii to install them with a local tribe. We were also putting tags on sea turtles and beaming the information up to a radio network. That got me a job at SpaceX as a freshman in college. I was part of a team ensuring the Dragon 2 capsule was safe for astronauts.
But even doing the SpaceX work wasnβt really enough. On the weekends, I would stay up late and work on other projects. After that I worked at a couple of startups. One of them was writing software for a folding bike helmet company that ended up not working out so well. I was working in New York, going to school in Boston, and flying to China every other week. Then I delivered medical products with drones in Rwanda. Then I worked with Zipline making fixed-wing drones. When I was there I had this idea. I took it to the head of the engineering team. He said, βit sucks to lose you, but you should really do this company thing, because you got one chance in your life to do it.β
What was the initial idea?Β
I had an interesting way to solve the real issue with solar power. Itβs this unstoppable force. Everybody's installing so many solar panels everywhere. Itβs really a great candidate to power humanity. But sunlight turns off, it's called nighttime. If you solve that fundamental problem, you fix solar everywhere.Β
Where did the idea come from?Β
I was watching a YouTube video called The Problem with Solar Energy in Africa. It was basically saying that you need three times as many solar panels in Germany as you do in the Sahara Desert and you can't get the power from the Sahara to Germany in an easy way.
I thought, what if you could beam the sunlight and then reflect it with mirrors, and put that light into laser beam vacuum tubes that zigzag around the curvature of the Earth.Β It could be this beam that comes in just like power companies, this tube full of infinite light. That was the initial idea. But the approach was completely economically unworkable.
I was like, this is not going to compete with solar in ten years. I should just completely give up and do something else. Then I was on a run two days later and thought what if I put that thing that turns sunlight into a beam in orbit then you donβt have to build a vacuum tube anymore. And it's so much more valuable because you can shine sunlight on solar farms that already exist. Then I developed several more technologies which I know for a fact no one else is working on. That made the model even more economical.
Are these just like regular household mirrors, but fixed to a satellite?
If you did that, the light would go to too many places. The sun is a certain size. It's not a point, it has a distance across. The light from one side of the sun would bounce off your mirror, and the light from the other side would also bounce off your mirror. If you used a perfectly flat mirror, every single microscopic piece would have this angle of diverging light coming from it. By the time the reflection hit Earth, youβd get a 3.6 kilometer diameter spot, which is gigantic. There are only 10 solar farms that big.
So I did the math, and figured out that if I could hit a 500-meter spot instead of a 3,600-meter spot, then Iβd be able to hit 44 times more solar sites per orbit.
So whatβs the solve?
The only way to do this is with a collimator.
If you picture a candle, the light is going to spring out in every possible direction. If you put a mirror on one side of the candle, now it's spreading out more one way than everywhere else. If you put the mirrors all around the candle, youβll get more lighting on a single very specific point. Itβs about how parallel the light rays are. A laser is very collimated light.
There was one patent for a collimator in orbit filed back in 2005. It was basically a giant reverse telescope that shined light down on Earth. But it would be super-expensive to make a telescope that big because it would have to be optically perfect. So I spent three months trying to figure out how to make it cheap. Then one day I suggested to somebody that we could maybe tile it. It took me another two weeks to figure out how to actually tile it.
You lost me with these tilesβ¦
The James Webb Space Telescope takes light from a very small star very far away and blows that image up. I'm doing precisely the opposite. It's the same exact mirrors, you just turn it the other way. If you did that the James Webb way and you want to make something as bright as the sun on the ground with a collimator, the big lens has to be a kilometer-and-a-half across, which is gigantic.
If you want a mirror that's a kilometer-and-a-half across, and have it be the right shape - which is not flat, itβs a parabola - every part has to be extremely accurate and it's gonna be really big and floppy. And if you ever want to rotate it, youβre in trouble.
Instead of making a kilometer-and-a-half single parabola, I am making a kilometer-and-a-half of several million parabolas. These tiny little parabolas are like dimples on a plastic Diet Coke cup. At that size, they hold the shape of the parabola so well. It completely solves the problem and makes it super-linearly scalable.
What needs to happen to get from concept to application here?
It's really just money. I'm trying to raise $5 million to put one of these tiles on the International Space Station right now. I want it up there to make sure this spot size is what we expect it to be and that all the math checks out. The next step is building a bigger satellite that's a proof-of-profitability that serves legitimate customers on the ground. The next step after that is raise even more money and put a bunch more satellites up there and have a real constellation. Once weβre at that stage, weβll know how cheap the manufacturing gets, how expensive the set-up is, fixed costs, operating costs. Then we'll have a better idea of how this stacks up against fossil fuel plants. Making this cheaper than everything else, thatβs the challenge.
Are you trying to blind us all?
We're designing ours to be as bright as the sun or less. Even if you shoot light in the wrong spot, there's no way for it to be brighter than the sun. So yeah, there's not much of an issue there.
Whatβs the conversation like with solar farms?
Our conversations with those guys goes like this: βYou know, your solar farm shuts off after the sun has gone? Wouldn't it be cool if it didn't do that?β I've talked to some of the biggest solar farm owners in the US and they're like, βyeah, obviously, that's a godsend. You're telling me I can make more money with the infrastructure Iβve already bought, and pay off my loan faster?β
There are three cost tiers. They pay us for sunlight and then they sell the electricity. We sign a contract with them to utilize their infrastructure for a couple hours when they're otherwise not utilizing it and we own the electricity sale. The third tier is we buy solar farms and operate them ourselves.
Whatβs the timing here?
I'm trying to raise $5 million by October 1. Weβre then aiming to get on the mission thatβs going up to the International Space Station in April. My full-time job is going to become raising another $50 or $100 million for the next round, which is going to be the profitable satellite. Itβs a gigantic round. But that's what itβs going to take to get this done.
Anyone told youβre out of your mind yet?
Oh, yeah. Several optical physicists at the biggest aerospace companies like James Webb, Starlink and Keck Observatory. They said, βyou canβt make a smaller spot from orbitβ. They said, βlight spreads out, you canβt do it.β I said, βwhat if you use an Inverse Cassegrain.β For weeks, they had to think about it. One guy had been working on the problem for five years. Then they came back and said, βoh shit, yeahβ. Then they said, βit would take up eight times more area in space.β I said, βif itβs eight times cheaper it doesnβt matter.β And they said, βoh shit, yeah.β One guy said, βitβs impossible to make a Cassegrain linearly scalable.β So I solved that too.
Iβve talked to a lot of people. Iβm four or five stages ahead of anything thatβs been done. Thereβs very few people working on this and I know what theyβre all doing.
How is the lane so open? Is it a lack of imagination? Are the corp bros moving too slowly?
When you tell most people something, they donβt look into it to figure out why itβs wrong. My entire life Iβve been thinking about stuff and why itβs wrong for all these reasons. Then I sit down and do the math. Usually the things people tell you are just completely fucking wrong. I've worked at enough aerospace companies to see how things go. It's just groupthink.
A lot of times I'm wrong. I'm wrong way more often than I'm right. But I am also right more often than people because I'm trying more often.
You feeling pretty right on this one?
So far! I haven't figured out how to aim this thing as fast as I want to yet because every two minutes you have to switch to a new satellite. If you have to rotate the satellite that many times a day, the fuel costs get really expensive. It would be awesome if it was electrical and especially if it was an electrostatic distribution of voltage across a panel that was also responsible for aiming it.
My thinking process for that is go surfing for a couple days, don't think about it at all, and then come back to it with a fresh perspective.
How much time have you spent with the interplanetary motion aspect here?
That's been completely solved. Teledyne has this tech on the International Space Station called MUSES. Itβs a two-axis gimbal. They know exactly where the space station is, exactly where it's pointing, exactly where the Earth is, and exactly how it's rotating. They can hit a 30-meter spot from orbit. Google Earth does this. Every planetary imager has to know where it's pointing. MUSES has four spots for people like me, companies that can install their hardware and test.
How would you sum up the value proposition of Tons of Mirrors?
Today, with the solar panels that are out there, it's a $20 billion a year industry. What Iβm building is bigger than any of the markets they currently have. If this is the electric solution, and let's say in 200 years this replaces fossil fuels, itβs a $17 trillion market.
Are you stepping on anyoneβs toes here?
Thereβs scenarios where I'm working with one country and not another country. Itβs an enormous national security risk if China has access to electricity for 10 or 100 times cheaper than the US does. There's the fossil fuel guys that have recently been investing in solar panels. Thereβs also the RF beamdown guys. Thereβs the fusion reactor guys. This is stepping on their toes, because Iβm making electricity way cheaper than they can. Iβve heard horror stories about big fossil fuel plants from dozens of people - that if you threaten billions of dollars of revenue theyβre just gonna kill you. But I don't think the world works like that. I just think it's not a thing that's worth thinking about.
I want to do this as fast as possible and I want to make it as good as possible. That's all I want.
If he built a fusion reactor in high school the billions of fusion research dollars spent so far would be superfluous.
A "fusion reactor" in high school? Seriously?