Transcript
Ask Jigar
Jamie Nolan: Welcome to Ask Jigar, our weekly segment where Jigar answers your questions about energy, and honestly, anything else you're thinking about. Huge thanks to our sponsor, Octopus Energy, for making this segment possible. And a quick reminder that anyone who sends in a question that Jigar answers on the show gets an Energy Empire hat from our merch shop, so send them in.
Rooftop Solar at $2,100 a Kilowatt
Jamie Nolan: All right, let's get to it, Jigar. Our first question comes from Chad Comeault, who says: Jigar, I just put 12 kilowatts of unsubsidized solar on my roof in Portland, Oregon, for $2,100 a kilowatt. That's about the same installed cost as your average utility-scale onshore wind farm today. What gives? How did rooftop solar and utility-scale wind converge on cost, and what does that mean for the United States?
Jigar Shah: It's a great question, and something people have been pondering a lot over the last 12 months. Last year, Tesla put out a paper describing exactly how you could reduce the cost of solar down to this $2,100 a kilowatt. There have been a lot of other efforts. One of the companies that supports people who want to do it themselves is a company called Signature Solar, which lets you buy all the equipment yourself and have it shipped direct to your home. Some people actually decide to build it themselves. You have to be very handy to do that safely, so please be safe out there. And others will hire somebody local with an electrical license to help them finish it. But in Australia, they install residential solar for half that cost. In Germany, it's also cheaper than $2,100 a kilowatt. It's important to note that we can hit that number. And if we hit that number, along with battery storage, then pretty much everybody in the United States could generate their own power for 10 cents a kilowatt hour, which is cheaper than what anyone pays across the country.
Jamie Nolan: Well, let's just say that's not what I paid. That's a hell of a cost. So congratulations to you, Chad.
Do We Actually Need All This Compute?
Jamie Nolan: My next question is from Austin Nielsen-Penner, who says: Jigar, why is data center load growth treated as a given? ChatGPT has been out for almost four years, and I've stopped hearing about exciting new uses. History is full of speculative technology manias. Why does almost no one in energy question whether we actually need all of these gigawatts of compute?
Jigar Shah: Well, I appreciate the question. And I certainly have been questioning whether we need all this compute, here on Energy Empire as well as on the other podcast I'm active on, which is Open Circuit. But at the end of the day, what you find is that we're in a cycle right now where everybody thinks that they're going to be a winner, and that when they're a winner, they're going to be short on compute. So they need to go in hard and get as much compute as they can. Places like Texas have 470,000 megawatts of requests for additional load into a grid that only has a 91,000 megawatt peak. It's crazy town. So it is very clear that most of that is not going to get built, and that we don't need that much compute, at least not in the next five years.
But I do think this time is a little bit different, in that when someone builds a data center and they end up going out of business, I do think someone else is going to pick up that compute and use it for their needs. So I don't think we've overbuilt yet. The other thing I would say is that now that we're moving away from training and into inference, you're seeing a bunch of people talk loudly about not needing 1,000 megawatt data centers anymore, and that we can actually use 20 megawatt data centers, which are way easier and way cheaper to integrate into the grid. So if you're going to push on something, push on making data centers smaller and using the underutilized part of our grid. That would actually bring down everybody's electricity bill.
USED EVs AND VEHICLE-TO-HOME
Jamie Nolan: Great. Our next question comes from Kathy Malone, who says: Jigar, back in March you talked up Tesla home powering. So let me point you to what GM is doing. I just bought a 2024 Equinox EV coming off lease for half the price of a new one, since residual values tanked when the incentives went away. And GM made these cars compatible with their EV-to-home hardware. I'm saving up for that next. I'm in love, and I will never go back to gas. So here's my question: are used EVs plus vehicle-to-home the best bargain in energy right now?
Jigar Shah: Yes. Is that our shortest answer yet? Basically, if you buy a brand new Tesla Powerwall 3, which I did recently, it probably gets installed for roughly $900 a kilowatt hour. If you have a car that uses 100 kilowatt hours of power, which is most large EVs, 900 multiplied by 100 would be $90,000. And most people spend less than $90,000 for their car. So if you get a used EV for $20,000, that means you're paying $200 a kilowatt hour for that battery. You're getting the car for free just by buying the battery. It's a screaming good deal.
Now, the problem is that the hardware to connect the car to your home, to run your home as a vehicle-to-grid or vehicle-to-home setup, is expensive right now. But there are a number of firms working on radically reducing that cost. I think by 2030, pretty much everybody who buys an electric vehicle will also buy a vehicle-to-home or vehicle-to-grid system, to make sure that their home can run off their car.
Jamie Nolan: Awesome. I'm excited. I can't wait for that. That was not available when I got my charger, so I will look forward to adding that to my house.
Why Solar Matters
Jamie Nolan: Our final question comes from Sanai Parikh, who I believe is related to you, Jigar Shah.
Jigar Shah: My niece.
Jamie Nolan: Amazing. Well, she says: Jigar, why is solar energy so important in this world, and what small everyday changes can we make that actually matter in the long run? So we got a two-parter.
Jigar Shah: Oh my gosh. Now I have to really think hard, because she's going to confront me about this. When I was 16, I remember reading about all these different energy technologies and being wowed by solar. And the reason is that basically everything that we do is to create steam. You burn coal to make steam. You burn natural gas to make steam. You process uranium and have fission reactions to make steam. And then you turn a generator, and there's magnets inside of it, and that's how you get AC power. Even in wind turbines, you spin a generator with magnets to generate power.
Solar does not work like that. Solar is a radically new way to make electricity. It is photons coming from the sun that dislodge an electron sitting on a semiconductor wafer. And then those electrons are all collected, and you get DC power, not AC power, and then you go through an inverter to convert it into AC power. That is so cool. And the reason it was so important in the 1980s, for instance, is there used to be weather stations and telecom towers where you had to helicopter in diesel fuel to keep the station going, and solar was way cheaper than helicoptering in diesel fuel. And today it just keeps getting cheaper and cheaper and cheaper. Now 50% of all of the new electricity added to the grid worldwide, I think it's closer to 70% now, came from solar power last year. We are so lucky to have it. And if you want to generate your own power, it's really the only choice you have. So I am super excited about solar power, and super excited about you using solar power. Now you've got all these different formats, like balcony solar, all sorts of stuff. So super exciting.
Jamie Nolan: Amazing. And that's all for Ask Jigar this week. Remember, you can send in your own questions. If we choose your question to read on air, you'll get an Energy Empire hat. The link for submitting questions is in the show notes.