The Deal With Solar

In the Tesla post, we established what would need to happen for us to get to a totally sustainable future world:

1) We need almost everything we use to be running on electricity.

2) We need almost all of our electricity to be produced from sustainable sources.

We’ve talked about #1 a lot already in the Tesla post. There are other parts of #1 beyond cars—think of what still works in your house when there’s a blackout. For many people, that’s their car, their heat, their stove, and their water. Water faucets are usually powered by gravity, so they’re off the hook, but the other three things should all eventually end up running on electricity. Musk is so optimistic about the future of #1 that he thinks all transportation except space-bound rockets will go electric in the future.

But how about #2? The fact is, #2 is an even more important challenge to conquer than #1—why? Because even if we perfect #1 and everything runs on electricity, if that electricity isn’t produced cleanly, it has limited effect. Electric cars can’t fix the grid or shut down fossil fuel power plants—all cars can do is get themselves ready for the Electric Era by growing up and ditching gas. That’s their role in this puzzle.

To conquer #2, we’ll need to pour money and innovation into sustainable energy production the same way we’ve poured ourselves into coming up with creative new ways to extract fossil fuels.

In order for an energy source to be sustainable, it has to be both renewable and clean, which I’m not sure everyone realizes are different things—i.e. A) renewable so it won’t run out and B) clean so it won’t throw garbage into the atmosphere. Some good options:

Nuclear: Nuclear is controversial, but built well (i.e. don’t do it dumbly) and in the right places (i.e. not on a fault line), nuclear power plants could play a large, healthy part in our future.

Hydro: Building new dams is hard in 2015 because everyone gets too mad at you, but there are a lot of existing dams, so hydro will continue to play its part.

Wind: Wind has huge potential for growth as a key source of the future.

But none of those are the energy source of the future. Check out this chart and take your best guess:1


Super awkward photoshoot when they made this diagram because solar is just a weird size compared to the others. And if the chart isn’t already silly enough, note that those are spheres—if they were two-dimensional circles, wind would have the diameter of a marble, coal would be the size of a bowling ball, and solar would be the size of a small house.

Musk calls the sun “this handy fusion reactor in the sky, where you don’t have to do anything—it works, it shows up every day, and it produces ridiculous amounts of power.”

And ridiculous is the right word—the sun radiates more energy to the Earth in a couple hours than all of humanity consumes from all sources each year.

We also sometimes forget that the sun already powers almost everything on Earth. The sun powers the weather, causing wind, precipitation, and warmth to exist; it’s the original source of all fossil fuel energy; and it powers the whole biosphere and its murder/theft cycle.1 Harnessing solar energy just cuts out various middlemen and goes straight to the source.

So we have this endless energy source and on the other side, most of our energy needs can be met with electricity. That makes solar panels—the connection between these two things—the most important thing ever. Right?

Musk seems to think so. Because he has nothing to do, he and his two cousins started SolarCity in 2006 with the mission of “accelerating mass adoption of sustainable energy.” Sound familiar? SolarCity’s mission and Tesla’s mission are two halves of a single effort to get us into the yellow zone.

SolarCity is like the Dell Computers of solar power, bringing together disparate parts of the supply chain to make an otherwise difficult process easy for the consumer. The hard part with solar power isn’t getting the panels—they’re increasingly cheap and easy to come by—it’s all the other stuff, like mounting the panels on the roof and wiring them to the grid. By acting as a one-stop-shop for homes and businesses who want to run on solar, SolarCity has become by far the largest solar installation company in America (and they’re about to get bigger—they’re currently building an advanced panel production factory in Buffalo, NY, which will be triple the size of the largest US solar plant). SolarCity is kind of like a distributed utility, or as they sometimes call it, “infrastructure as a service.” And they, and other companies like them, have given families and businesses an alternative to their local electrical utility.

Powerwall Blue Box
Solar power is all great until every night when it goes away, like an inconvenient waterfall that only runs during the day. And what would you do if you lived next to a waterfall that only ran during the day, but you knew you were going to want water at night too? You’d get a cup, or a bucket, or a barrel, and you’d collect a lot of water during the day to have at night. With solar power, the bucket is a big stationary battery. The battery hooks up to the roof solar panels, and collects energy all day for use at night. Incredibly obvious, right?Conveniently, one of Musk’s side projects is that $5 billion delicious-looking Gigafactory I showed you before. Musk says 30% of the Gigafactory capacity will be dedicated to making their stationary battery—called the Powerwall—making Tesla kind of a battery company as much a car company. For larger customers, Tesla makes a huge battery called the Powerpack, and for huge customers, Powerpacks can be grouped by the hundreds. Musk envisions entire cities running off of a nearby solar field and an accompanying block of batteries.Musk explained this all in a keynote presentation last month.

Solar energy is growing quickly, but Musk thinks the growth rate is going to explode in the coming years, predicting that solar will become the largest world energy source by 2031.2 His goal is for unsubsidized solar power to cost less than grid electricity from coal or fracked gas, and he thinks that over time, more and more families, businesses, and cities all over the world will ditch the grid and go energy independent. Musk also suggested that the solar panel + battery unit might “leapfrog” power lines in developing parts of the world—i.e. they’ll never get power lines at all because they’ll go solar before power lines ever arrive—kind of like the way mobile phones leapfrogged landlines for many people in developing nations.

When you realize how little of the world you’d need to cover with solar panels in order to power all of humanity—especially since most of the panels would go on rooftops and not take up extra land—the more obvious a solar future seems:2


I always think it would be cool to go back to 1994, knowing what I know now about the internet, and watch it all unfold up close. But the thing is, 2015 is the 1994 of some other important development—and rather than wonder what 1994 would have been like, I should be figuring out what Earth-altering movement is currently in its nascent stages.

I think we may be on the cusp of a revolution in solar power. Good time to start following the story.


This was a mini-post that’s part of a larger series on Elon Musk and his companies:

Part 1: Elon Musk: The World’s Raddest Man

Part 2: How Tesla Will Change the World

Part 3: How (and Why) SpaceX Will Colonize Mars

Part 4: The Cook and the Chef: Musk’s Secret Sauce


If you’re into Wait But Why, sign up for the Wait But Why email list and we’ll send you the new posts right when they come out. Better than having to check the site!

If you’re interested in supporting Wait But Why, here’s our Patreon.

  1. The few exceptions are geothermal energy, which is powered by the Earth’s hot core; ocean wave energy, which is powered by the moon’s gravity; and nuclear energy, which extracts energy from heavy elements like Uranium, which originally got their stored energy from ancient supernovae.

  2. Random year to pick.

  • Vinay Kapadia

    Love the Elon Musk posts! I just got solar panels installed on my house last Monday (they’re not even powering my home yet, still awaiting inspection from the electric utility). Best part is that here in Washington, its better than free. I took a loan to be able to get the panels, and the power company gives me a check every month for how much I produce (regardless of consumption), and that money makes the loan payment, and I have stuff left over! Super excited about the future of solar!

  • newnodm

    “SolarCity is like the Dell Computers of solar power, bringing together disparate parts of the supply chain to make an otherwise difficult process easy for the consumer.”

    All residential solar companies do this. I know you like Musk, but there is nothing special about solarcity. They are the highest price in most markets. Most people come out far ahead buying instead of leasing.

    Solarcity charges about triple per kilowatt compared to what solar systems cost in Germany and Brazil.

    • Rikaishi Rikashi

      I think Solar City was the first company to roll out many plug-and-play and financing innovations.

      Comparing costs to other nations is apples to oranges given the local variables involved.

      Leasing a product is always more expensive then buying outright. The point with solar leasing is to make it attainable for people on tight budgets.

      • newnodm

        Solarcity quotes a system purchase price along with the lease. It is high, and it is also what the buyout is based on. Overall a terrible deal.

        I’m not in the solar industry. I have never done business with solarcity. I have seen a lot of their proposals. I can’t believe anyone who reads through the lease actually signs the document.

        • Dan D

          After Tim’s first post about Elon mentioning SolarCity, we had a representative come out to our house. Everything looked fantastic and the numbers were enticing, but we simply wanted some numbers to compare with other companies’ figures and estimates. The next thing we knew, we were in a contract. We had three days to cancel it. Mea culpa. Looking at SolarCity on Yelp didn’t help, either, as they are rated pretty low compared to other companies. We are planning to get a Tesla and eventually a Powerwall, so having SolarCity involved would be part of the puzzle. I was shocked to read some of the reasons people posted low ratings for them. We’re still going solar (and we know it’s best to buy instead of lease), but it may not be with SolarCity.

          • newnodm

            What was the price per watt?

            • Dan D

              Rate per kWh

              16.90 ¢
              Before Rate Reduction Payment

              12.12 ¢
              After Rate Reduction Payment

            • newnodm

              Did they give you a purchase price? Normally that number is converter to price to panel watts for the installed system. This is normally how prices are compared between different vendor quotes.
              A good deal today in California would be something like a 5kw system at $3.20 per watt. That gives a purchase price of $16,000 installed. An installed system in Germany and Australia is about half that price.

            • Dan D

              $19,125 and then subtract $5,738 for the tax credit.
              Estimated System Size 3.750 kW
              Estimated First Year Production 5,845 kWh

            • newnodm

              $5.10 a watt

              That is extremely expensive. A good price in the U.S. would be a bit more than half that amount. Seldom does one hear about $5/watt in 2015, even in California.

              A 3.75kW installed system in Australia and Germany would not be more than $5000 today. Soalcity keeps 80% of the benefit of a solar install for themselves.

              Prices in Australia are going up due to demand because of very high electricity rates, but take a look:

  • Chris

    “…nuclear energy, which fuses heavy elements like Uranium” – oops. Nuclear energy is so far all *fission* – splitting apart the atom to get at its stored nuclear energy, not fusion (which will come later). Perhaps a mention of nuclear fusion on here would be good – it will probably eclipse (ha ha) solar within 50 years of its perfection…

  • Friedrich Stiebeling

    Nuclear is actually not only flawed, but not clean at all. There is over 12,000 tons of nuclear waste piling up every year, and there is no safe solution where to put it. This waste radiates for thousands of years (e.g. Plutonium 239 half-life period: 24.000y) and who can really know if a so called “permanent” disposal site is safe in the future? Geologically for instance, but in many other ways too: Who can assure that a human in 20k years will still know about that disposal site and understands the signs that it may be a bad idea to dig in that place? A lot of things can happen to humanity in that time, and it is way over our head to think we can foresee this. Just look back what happened the last 24k years.

    In addition to that, the production of nuclear energy is a huge bet. Tschernobyl and Fukushima show that if a meltdown happens, the impact is devastating. It even has an impact on a global scale, at least economically, and for a very long time. It is true that a chance of a meltdown for one reactor is very, very low because of the huge safety precautions. The problem is, that but you can not eradicate the chance of a meltdown totally, there will always be a small chance of the worst case happening. This infinitesimally small chance gets multiplied with every reactor being built, and there are already over 430 on earth to this day. This is just waiting for Murphy’s Law to apply again: What can happen, will happen.

    • Adam

      I did think I would disagree with your safety argument when I started reading the second paragraph — perceived risk of nuclear meltdowns in society is far higher than the actual risk, but it’s true that there is some appreciable risk of meltdown, otherwise we wouldn’t know the names Fukushima or Chernobyl. It might only be the names of 2 power plants out of the hundreds built since humanity started using nuclear power, but they still caused serious impacts.

      As for nuclear waste, I think it’s better than chucking out CO2 and SO2, NOx, CO etc., which coal, oil and natural gas do [although catalytic converters and flue scrubbing help with some of those gases] — we know that those gases aren’t safe in the short term, let alone long term. It could still be an issue, though, and dealing with it does increase the cost of running a power plant.

      But to compare, for example nuclear vs. fracking, I’m much more concerned about the unknown dangers of fracking than the known dangers of nuclear. Also, burning fossil fuels can cause serious harm to people in densely populated areas. It’s not just nuclear power that’s dangerous to human health.

      • Friedrich Stiebeling

        I agree that fossil energy is the main problem here, but to state nuclear energy as a viable, green alternative is very short-sighted. Tim Urban does a great piece of journalism here, but I think he should think again about the “nuclear option” 😉

        • Adam

          A thought that’s just come to me is that because the future is so unpredictable, nuclear waste might not turn out to be a problem. “Nuclear waste? We just react that with [sciencey-sounding chemical]”, people might be saying in a couple of hundred years, let alone 24,000 years. I agree nuclear has some serious downsides, and nuclear waste is certainly not something we should aspire to produce, but I think it’s still a possible short-term aid to getting the world away from fossil fuels.

          I think it’s incredibly important that there are multiple copies of records of where nuclear waste is being chucked, though: we don’t want people accidentally stumbling upon it in the future. I agree it’s a possible risk and I am still kind of conflicted over my opinion of nuclear.

          • That [sciency-sounding chemical] is actually a technique called “nuclear reprocessing”, and it’s used extensively in fuel cycles outside the US. It reduces the waste to a small fraction of that found in your average spent fuel rod, in addition to recovering most of the U-235 and Pu-239 for use as additional fuel. (The politics are nasty on this because that recovered fuel is a proliferation risk.)

            Furthermore, several advanced reactor designs can receive the reprocessed waste and burn it up as part of normal reactor operations. This is a problem that’s about 95% solvable. You’re always going to have a small amount of waste, but it can be almost arbitrarily tiny.

    • Rikaishi Rikashi

      The waste and safety concerns can be remedied by technology advances – fast breeders and gen IV designs. But fast breeders have already proven to be a weapons proliferation risk, gen IV designs cannot make a reactor bomb or sabotage-proof, and these generators are not expected to be ready for deployment till 2050 alongside fusion power.

      Meanwhile the nuclear industry has disintegrated, and the few construction or renovation projects that exist are mostly delayed and over budget or stuck in planning.

      Post-fukushima safety regimes have basically priced nuclear out of the market. With the cost of solar continuing to fall, and unlikely to stop falling until it bottoms out somewhere between 3c/kWh and 0c/kWh, it’s unlikely that nuclear will ever be a viable energy source again.

      • Gwen

        We’ve been saying the fusion power will be ready in 20 years since the 60s. Saying that 2050 is the expected timeframe for fusion power isn’t really accurate, as far as I know. If you have a better reason for that than I know of, I’d really appreciate it if you could relate it.

        • Rikaishi Rikashi

          I think 2050 was the assessment of scientists building the ITER reactor. But yes, that is still quite speculative, especially since fusion will not be “ready” until it can compete economically.

          • neroden

            As always, we have a great fusion reactor — it’s called the sun. We should use it.

    • Gwen

      Potential new generation nuclear reactors (thorium) would produce perhaps up to a thousand times less waste, and the waste produced would decay after a couple hundred years, which would be much more manageable. In addition, weapon-grade radioactive material would not be produced by the reactor.

      Chernobyl and Fukushima are definitely not indicators of devastating impacts from nuclear meltdowns. The United Nations Scientific Committee on the Effects of Atomic Radiation (UNSCEAR) released a report that stated that they had found no evidence to support any increase in cancer rates or birth defects from Fukushima. As reported by the WHO, the Chernobyl disaster could cause up to 4000 deaths, with only 50 of those having been caused by 2005, about twenty years later. According to Dr. Burton Bennett, the chair of the WHO committee, “we have not found profound negative health impacts to the rest of the population in surrounding areas, nor have we found widespread contamination that would continue to pose a substantial threat to human health, within a few exceptional, restricted areas [the roughly fifty workers that were exposed inside of the plant itself].” Obviously 4000 deaths are not good, but that’s hardly an indicator of mass danger.

    • 12,000 tonnes of nuclear waste is equivalent to about 24,000 spent fuel rods. A PWR fuel rod is about 4.5m x 20cm x 20cm, or 0.18 cubic meters. So you’re talking about 4320 cubic meters of waste. If you melted that down and made a solid cube out of it, it’d be 16.3 meters on a side. Not exactly an enormous amount of waste.

      Now, if you put all of that into a solid cube, it would melt instantly and become highly nasty, so we’re really talking about a practical volume that’s much larger than that. If we make it 10,000 times larger, now the cube is 351 meters on a side. That’s a pretty big space, but it’s a pretty big world, too. Compare that with the size of coal slurry ponds or the settling ponds for fracking fluid and you’re going to come to the conclusion that the waste problem isn’t so bad. And, of course, if you’re really worried about it, you reprocess the waste, which not only reduces its size by a large amount but produces plenty of additional nuclear fuel. Then you vitrify the remaining waste and it pretty much can’t go into the water table, even if the dry casks and the concrete containment of your waste depot are breached.

      Just because I get obsessive about stuff like this, I once went over the maps and did a calculation that totaled up all the areas contaminated by Chernobyl, Fukushima, and Kystym (the only three disasters to make any land uninhabitable) at more than 5 curies/km^2 (which is what the IAEA uses as a metric for “uninhabitable”), and it came out that, if you average the land lost over the entire 60 year history of commercial nuclear power, you lose about 500 square kilometers of land a year to nuclear accidents. Again, not nothing, but we’re talking about a square that’s 14 miles on a side. (Weasel words: A lot of governments use 2 curies/km2 as the habitability standard, and I was eyeballing the maps. My area estimate could be off by a factor of two, and going down to 2 ci/km2 would probably quintuple the area. So we could be talking about 5000 km2 of loss per year, or a square 44 miles on a side.)

      • neroden

        That land loss is actually pretty bad from an agricultural-land point of view, because that land is lost basically forever. Even land which has houses built on it can be returned to agricultural use as needed.

        • Cs-137 and Sr-90 both have half-lives of about 40 years. Even if you assume contamination of up to 10 curies per sq km, you’re down to 0.6 ci/km^2 in 160 years. From an arability standpoint, we’re a) not talking about huge amounts of land and b) you get it back over time. At some point, even if you continue to lose land at the same rate to new accidents, land from old accidents is returned to service at the same rate as land is lost.

  • Gwen

    An important other point to mention is the possibility of nuclear fusion power (different from the current nuclear fission) in the future. Once we reach the point of sustainable fusion power, we can have essentially infinite power fed straight into the grid for almost no cost and with no waste – it’s the same as solar power except that the energy is generated on earth instead of in the sun, and it is collected through a much more efficient process that solar power. The problem is that we haven’t actually got nuclear fusion to work here on earth sustainably (it takes less energy to start it than it gives off).

    • Nicolas Guérin

      The usual fusion is made with hydrogen, which does not exist in large enough quantities on Earth. We have to “make” it in order to fuel fusion power plants. The problem is that it needs a lot of energy to be manufactured making it inefficient. The way fusion is made right now also produces waste. Hydrogen is more a battery, a way to store energy, than an energy source.

      • Sarudak

        You’ve got it all mixed up. In the sense of burning hydrogen (chemichally) you’re correct hydrogen is not an energy source. In the sense of nuclear fusion the energy released is orders of magnitude more than it takes to release hydrogen from it’s chemical bonds and fusion is very much an energy source (after all it’s what powers the sun)

  • R1ckr011

    I was extremely upset you didn’t take a hatchet to the “solar panels aren’t biodegradable” bullshit that I hear a LOT. v I was expecting that kind of treatment, the same way you destroyed the “long tailpipe” argument.

    Also: indeed fusion, but I think We can’t worry about THAT because there isn’t even a decent framework for making it more than .01% efficient. We can get excited in 2030 when they actually build the first fusion generator that Works. Until then, only solar, wind, and tidal make sense at current scales.

    Nuclear is much greener than any fossil fuel, since almost all plastics are horribly wasted in landfills, and so much is in the air. Nuclear will never have a “pool of poison” in the Pacific ocean. Tons of waste are cute to say when making a big deal over it, but can’t we simply use long-lived isotopes in different reactions to cause them to decay faster? Or instead of stockpiling them, spread them as far apart as possible? hell, would a couple grams of nuclear material encased in lead pellets spread worldwide REALLY cause that much of a problem?* There’s lead poisoning in groundwater, but once again, everything in chemistry is a matter oc concentration. 100 grams per square kilometer seems like a very safe distribution. Hell, leave them naked and use them as tracers for very small fluid flows on earth, mapping out much of the world’s waterways..

    *Randal from XKCD doesn’t seem to think nuclear’s that bad, and he’s got the science chops to talk about it. OOoooh, I smell a CROSSOVER 😀 My two favorite online people atm!!

    Not that it matters, there’s no political will to put the science behind nuclear and not enough science to put the politics behind nuclear. It’s a catch 22. Which sucks. We could have completely skipped over fossil fuels in 1940-60, pushing nuclear tech much higher and better. Guess it was just too scary, and that same fear of making mistakes is why nuclear proliferation is so awful at this point, IMHO.

    Is that an expensive project? surely, but a lot less problematic than having to guard stockpiles of worthless waste

  • Greg Alexander

    Although there may not be scientific and political will to research nuclear in the Western world, China will have a Liquid Thorium solution in 10-20 years. Far safer plant design, far cheaper, short half-life of waste, can’t be used for weapons, and plenty of thorium to mine (in China, but also elsewhere). China will make a lot of money on it, but at least it will be available.

    • CaPEITÃO América

      I really dig this new thorium reactors, seems like the solution or mitigation to all standard nuclear reactors problems. We only have to thank China for doing new stuff while the ocident keeps weeping about the high costs to start the technology.

  • Jay

    Tim- How much would it cost, and how long would it take to fabricate 500,000 Km2 of solar panels with current technology? Can you figure that out and post the answer?

    • jeffhre

      Solar costs are usually determined by capacity. The area used is variable, depending on type and efficiency. But yes, Tim could figure it out!

  • kenneth679

    Tim, if you’re going to talk about future energy sources, you need to look into liquid thorium fluoride reactors, which change the outlook on nuclear quite a lot. See Kirk Sorensen’s many videos on Youtube to start. Here’s a fairly short one:

    • James Heffernan

      THANK YOU! I was actually about to share this.

      • jeffhre

        Future? Rickover won, Weinberg lost in the 1960’s. That is a tough road to travel from here out.

  • jeffhre

    Tim, I’m not certain, but there may be something missing from your explanation of the chart with the big sun. The fuels listed generally show all known reserves whereas, the power from the sun is from one year of sunlight. My bad if you already covered it…You provide so much info in Wait But Why that I can get lost in it quickly!

    OK yes, now I see what the very large letters at the bottom of the chart with the huge arrows mean!

  • Steve O

    Hey, do you think you could come up with a better source for your your “how much area would the solar panels take” question? That map seems to be made by a Gizmodo writer who didn’t bother to explain his math at all.

  • ?

    Or we could keep working on harnessing the energy from the ionosphere? That eliminates the need for any sort of grid..or excessive infrastructure. Pretty sure that MIT is working on that now but they’ve only gotten to the step of transmitting the electricity wirelessly, not absorbing it from the ionosphere. I feel that solar, nuclear (of any sort), hydro, etc.. is simply a soft patch fix until we rediscover how to do what mr. Tesla discovered.

  • Titanius55

    Good post! I’m loving all of your stuff on Musk so far. Personally, I’m more for the idea of solar over any alternative power source. I live in Massachusetts in the U.S. and I remember a lot of complaining over the idea of putting wind turbines up all over the Cape. People didn’t want their views of the ocean and beaches and what-not messed with. Not the best reason to avoid a new technology but solar seems like it could avoid that problem. It seems like one of the simplest and safest solutions we have right now for providing a lot of clean energy.

  • Eleonora Rossi

    What do you think about Desertec project?

  • Harpreet Singh Sandhu

    Around 496,806 sq kilometers..thats almost the size of SPAIN !!!..this alone makes me sure that solar alone is not the answer and we’d have to accelerate Thorium and Fusion research too..sorry to say, but Tim is a little too impressed with Elon musk..sure we need Solar, and we need it all and now. But nuclear also needs a lot more attention and credit..

    • S. T.

      But think about how much roof-space there is in the world. You could easily put most if not all solar panels on top of buildings. Those that don’t go there could (and are) be placed on deserts, old landfills, or otherwise unusable land.

  • Jeff K

    After reading the Tesla post I immediately started looking into getting solar for my house. I’ve found that most ways people do solar is through a process called net metering where during the day you use your solar and sell any electricity extra back to the power company. Then at night you use the grid to power your home. The energy purchase from the grid at night is easily offset by your energy sale to the grid during
    the day. I’ve found that in order to qualify for most rebates on solar this relationship is a requirement. This means you will not be storing energy in a battery (i.e. powerwall).

    Another reason to not use the battery is storing energy in a battery and then using it incurs loses in the system. Net metering does not have any loses. Batteries have additional installation costs and need to be replaced every 5-10 years. Net metering has no replacements and no extra installation costs.

    Don’t get me wrong I love the idea of the battery and being off the grid but the concept of net metering does not incentivize customers to buy a powerwall.

    • neroden

      The utility companies are trying to get rid of net metering (it isn’t a very good deal for them). So there are some good reasons to try to go totally off grid, if you can pull it off. Most people can’t pull it off yet — need better, cheaper batteries.

Home Archive