I think over all for the 21st-century the most important terrestrial problem that humanity faces is sustainable production and consumption of energy. It's getting harder and harder to find hydrocarbons, and it's getting much more expensive to extract them. The cost of extraction has doubled or in some cases tripled. Really we are just arguing about the when hydrocarbons run out, or become prohibitively expensive, not if.
If we continue to rely on producing, mining and burning billions of tons of hydrocarbons every year which effectively, and permanently from a human standpoint, affects the carbon content of the oceans and atmosphere, the future is going to be quite bad. The vast majority of the scientific establishment believes that. Anyone with a scientific background is unequivocal. And anyone who thinks that it is 100% certain that global warming is fake, and that massively changing the chemical composition of the oceans and atmosphere is fine, is a bloody fool, obviously.
That’s what this is all about, this is about trying to accelerate where we know we need to get to anyway. Not on like a small potatoes way, but in a very big macro scale way.
At the risk of being repetitive, there's going to be no choice in the long term to move to sustainable energy. It's tautological, either it is renewable or is not renewable. Non-renewables is like being stuck in a room where the oxygen is gradually depleting, and outside it is not.
New technology and innovation can have a downside, and one of the downsides is the ability to extract far more hydrocarbons than we thought were possible. But it’s still finite, we must still find a solution or we will face economic collapse when the resources become scarce. There are time extensions on the game, but the game is going to come to an end. That should be absolutely certain, obviously frankly. The only thing we gain by slowing down the transition is just slowing it down, it doesn't make it not occur it just slows it down.
I do think we should watch our consumption. We shouldn't be wasteful, but even if we are really conservative in our use of energy, and are very effective with recycling and all sorts of things, that delays the need to move to a sustainable future, but it doesn't eliminate it. If we don't have sustainable energy generation, there's no way that we can conserve our way to a good future. We have to fundamentally make sustainable energy available.
The important thing to appreciate is that it is inevitable. The question is really when do we exit the fossil fuels era, not if. The goal is to exit the era as quickly as possible. That means we need to move from the old goal. The old pre-industrial goal was to move from chopping wood and killing lots whales to fossil fuels, which actually in that context was a good thing. The new goal is to exit the fossil fuel era and move to a sustainable energy future.
I think in the future people will look back at the gasoline era the way we look back at the steam era today. We are going to look back on gasoline cars like we look back on steam engines, it was like a phase, it was a bit weird. We’re going to look back on fossil fuel power generation the same way. It was a weird phase. I think you will be telling your grandchildren: “Yes, you won’t believe what we used to do. We used to take up liquidized remains of dinosaurs and old plants, and put them in cars and burn them to move. And we did the same thing with the power plants and the like.” Gasoline, it's quaint, it's interesting, but it's basically a phase. In the future we will look back, and with the future I don’t mean super-far into the future, I'm talking about towards the end of the century, we will look back on gasoline the same way we look back on coal, as sort of a quaint anachronism that's in a museum. That sounds crazy, but that’s what it’s going to be in the future.
So yes we’re trying to have the non-weird future get here as fast as possible. We want to get out of that weird phase as soon as we can. If you believe that that’s the future we’re headed towards, we must find alternatives. That is a known difficult thing that we are going to have to solve.
What actions can we take that will accelerate the transition out of this idiosyncratic moment in history, where we’re digging up Cambrian-level fossils and burning them? Methane is the lowest cost source fuel on the planet by a good margin. Methane is a naturally occurring gas. Once you start getting into deep methane or deep natural gas you're actually tapping into things that are not related to dinosaur fossils. There are places in the Solar System where the atmosphere is primarily methane, so it does not require an organic origin. If we dig too deep for methane, we're actually going to a level that has never been seen before, not even in the very earliest history of Earth. So that's very dangerous I think.
On the plus side, there's actually an enormous amount of sustainable energy. There's many forms of energy generation that can be sustaining into the very long term. We want to use things like hydro, solar, wind, and geothermal. Nuclear is also a good option in places which aren't subject to natural disasters. We want to use energy sources that will be good for a billion years.
You know, electricity’s sort of like cash. You can generate it in multiple ways; you can spend it in multiple ways. My personal view is that we'll generate more electricity from solar than any other single source. It may not be a majority, but I expect at least a plurality from solar power. That will be a combination of photovoltaics at the point of use, like the roofs of houses and businesses, which is also good from the standpoint of not requiring additional power lines. And then at the power plant level I think we'll see a lot of solar thermal power generation. Where essentially you're just using the Sun to heat a working fluid, and then generate steam and power a turbine. There are a bunch of those projects that are going to come online in California and other places in the United States.
I'm not the biggest fan of biofuels because I try to look at things and just calculate the basic physics of it. Really elementary stuff and say; what percentage of the incident sunlight is bound up in usable chemical energy, and then once you have that chemical energy how much of that is then translated into electricity? You have to compare that total efficiency with just having solar panels. Unless I've made some really dumb mistake, which is possible, you're about a hundred times off with biofuels, I mean, at least two orders of magnitude. What it boils down to is watts per square meter electricity generated. With the best case biofuel - take every assumption and maximize it, so don't say, don't worry, maybe somebody could invent something better, say what is the best - just envelope the whole thing. Say you had unbelievably efficient plants, I mean, you can't violate any laws of thermodynamics, but assuming that you're at the limits of the laws of thermodynamics in all those cases then biofuels - at least your land-based biofuels - there's no way this makes sense. You end up being around maybe 0.2% efficient in turning sunlight into electrical energy. Whereas commercial solar panels are 20% efficient. So why would you ever do biofuels? it’s not as though there are large swaths of arable land unused. You have to say, if you go with biofuels, it's going to either result in wilderness being cultivated or an increase in food prices. You can also say, is it possible if you stopped all food production in the world to generate enough energy to meet the world's needs? like, yeah, you could probably it's about right, actually, if you stopped all food production you could just about meet the world's energy needs. Now, there is a possibility of ocean-based, because Earth's surface is mostly ocean. So, if you could find maybe some sort of ocean algae-based solution where you're unconstrained by surface area, although I still think you'd have to compare that to a bunch of floating solar panels, and I think you still lose on floating solar panels. I don't see how it would make sense.
I think there's nothing wrong with nuclear power, whether fission or fusion. Fusion is when you take like let's say two hydrogen atoms, or two hydrogen isotopes technically, and slam them together and form helium. That's fusion. And fission is when you got like a heavy atom that is decaying at a relatively noticeable rate, like uranium or plutonium, and decays into smaller atoms. That's fission. I actually think nuclear is not a terrible option, so long as you're not located in a place that's susceptible to natural disasters. That also I think defies common sense. So long as there are not huge earthquakes, or weather systems that have names coming at you, then I think nuclear can be a sensible option. There are much safer and better ways of generating nuclear energy, I'm talking fission here, than existed in the past when nuclear reactors first came out. You have some meltdown risk, although there is some new technology on the fission front that makes meltdown risk extremely low.
At some point in the future it would be nice to make fusion work, of course. I think it's definitely possible to make fusion work. I used to be a big fan of like having that as a long time energy source. With fusion the difficulty is keeping it going. The great difficulty is to keeping the fire from going out. It's quite hard to sustain a fusion reaction unless you have something very big like the Sun. The Sun has gravitational confinement of the fusion reaction. Since you can't do gravitational confinement on Earth you have to do some sort of electromagnetic confinement in one form or another, or kinetic confinement by slamming things into each other. It’s quite tricky to prevent a fusion explosion from immediately extinguishing, but I think the fusion problem is probably easier than people think it is, and by this I'm thinking about magnetically confined fusion.That’s a problem that gets easier as you scale it up, because you get a service to volume advantage. It seems like a pretty obvious thing that if you could get it big enough you could have a real effective sort of magnetically confined fusion reactor. That’s probably not the easiest thing to solve. You could do like a thorium fission reactor. Or a better fission reactor, maybe it’s better to do better fission reactors, but fission does have a bit of a marketing problem, and fusion is the energy forever solution. It's exciting to see what's happening with the ITER Project, which is a fusion plant that's being built in France. I think we can definitely make fusion work, but it is a far off technology. To make fusion at the power plant level work is probably, I don't know, 30 years away and a lot of effort.
That's why at least for now and I think maybe even in the long-term I'm a proponent of using the big fusion power plant in the sky called the Sun. The Sun is a giant fusion explosion and it shows up every day. If we have photovoltaics, solar panels, we can capture that fusion energy.
It's worth noting I'm not sure if people are aware of this, but the world can be powered by solar many times over if you had enough battery capacity to pair with it. Many times, probably like times a thousand, that it’s literally true. We have this enormous fusion generator in the sky that is laving out vast amounts of energy, and I'm just talking about using land area, it’s really amazing, it's crazy. In fact here's a little tidbit, for a lot of nuclear power stations if you would take a nuclear power station, and the whole clearance area around it, it ends up being quite a bit of land. You got let's say 3 to 5 km radius of clear area where you can't have significant construction, like building houses and dense office and housing space, usually people don't want to do that near a nuclear power plant. You can't just put a nuclear power plant out in the suburbs with a bunch of people around it, so you have to have this big clear zone, they use a lot of area. If you took the land area including the stay-out zones and everything, and said what generates more power, the nuclear power plant or just carpet that area with solar panels? The solar panels on that area will typically generate more power than the nuclear power plant. Just the area used by the nuclear power plant covered in solar panels would generate more energy.
The amount of energy that reaches the Earth from the Sun is staggeringly high. Just to give you a sense of how much energy is hitting the Earth from the Sun. It's very easy to do if I may just do a tidbit of math. 1 km² is 1,000,000 m², and there is 1 kW per square meter of solar energy. So on one square kilometer there is a gigawatt of solar energy. Which is mind blowingly huge, that's a super giant amount of energy. You could power the entire United States with about 150 to 200 Square kilometers of solar panels. The entire United States with about a 100 mile by 100 mile grid of solar power. This is literally true, what I've just said. Take like a corner of Arizona and that would be all the energy that the United States needs. Take a corner of Utah there's not much going on there I have been there, there’s not even radio stations.
If you just took a small section of Spain you could power all of Europe. It's a very small amount of area that's actually needed to generate the electricity we need to power civilization or in the case of the U.S. a little corner of Nevada or Utah could power the entire United States.
China actually has an enormous land area, much of which is hardly occupied at all given that the Chinese population is so concentrated along the coast. Once you go inland the population in some cases is remarkably tiny. You could easily power all of China with solar. It's true that in dense cities rooftop solar is not going to solve the energy need. What you can do is have ground-mount solar power near Hong Kong tapping into the existing power lines that are coming in. So you can supply Hong Kong with solar power; it would just need to be coming from a land area that's not too far away.
If humanity had to get all of its energy from the Sun, it could do so. Currently in terms of total energy usage in the world, utilities provide about roughly a third of the energy consumed by civilization, electric utilities to be precise. Then another third is heating and another third is transportation roughly. There is enough energy coming to us from the Sun to actually support all three areas.
It also needs to be stored in a battery so we can use it at night. Then we want to have high power lines to transfer solar energy from one place to another.
I think the important thing to bear in mind is it’s a big world out there and there are places where the cost of energy is much higher than other places. For example, in Hawaii energy costs are very high because they have to ship in all of the fuel for their power plants, so it’s very expensive. The economics of solar plus battery make overwhelming sense for a place like Hawaii and a lot of actually island nations out there. And really anyplace that has got expensive energy costs or even moderately expensive energy costs. It’s going to make sense for many parts of Europe, many parts of the United States, and then, over time, it’s going to make sense for everywhere.
I would say we do need to think about transport in general, about a third of all energy is used for transport. I believe in electric transport because it allows for energy to be produced in a wide range of sustainable means, and you just charge the car.
Again it's worth noting that even if the grid was fully powered by coal and natural gas, electric cars would still generate less CO2 even if you take it all the way to the power plant level. The reason for that is when you are not constrained by mass and volume, you can make the efficiency of energy extraction much better in a power plant then you can in a car. If you take say a natural gas power plant from GE it's over 60% efficient, so if you take the source energy it's 60% efficient in generating electricity, it’s really good. Whereas typically a gasoline powered car over the drive cycle will be less than 20% efficient. That is because the big natural gas turbine can be really heavy, it can be really bulky, and you can take the waste heat and run a steam turbine and get even more energy out. Your efficiency is just fundamentally better, even when you take into account the transmission losses and the charging losses you are still way far ahead with electric cars than you are with gasoline cars. Stationary power plants are so much more efficient than small gasoline engines in cars, an electric car ends up getting more range for a given amount of say, coal or oil that's burned than a gasoline car gets. In other words the CO2 per mile is actually less for an electric car even if you draw electricity from a high CO2 source like coal or natural gas, or even directly from oil.
Now, of course, long term we have to find sustainable power generation and sustainable transportation. Both sides of the equation need to be solved, and even if electric cars weren't there we still need to get sustainable power generation.
So the great thing about electric cars is you can generate the electricity from a wide range of renewable sources like hydro, geothermal, wind, solar and nuclear where it's save to do so.
The energy density, basically the amount of energy you can store in a given amount of mass or volume, was a fundamental constraint on electric cars for a while, and that's correlated to some degree with the cost per kWh, the cost of storing that energy in the car. The advent of lithium-ion technology I think is really what enabled a compelling car. I think if we had to we could turn the entire automotive world to pure electric. Lithium-ion batteries continue to improve roughly on average maybe 8% or 9% per year. Which when compounded over several years ends up being a meaningful improvement. Even if there was no fundamental improvement beyond lithium-ion batteries, I think we could still take all terrestrial, all ground transportation could go electric.
It gets harder for airplanes, we do need a further breakthrough for aircraft, where the energy density requirements are at least 2 to 3 times more significant, but even with current generation lithium-ion we could go to mass market with ground vehicles. Certainly for cars, boats, and trains lithium-ion could do it. For rockets, well, there's no way to make a rocket electric, that’s for sure. Unfortunately Newton's third law cannot be escaped - I think. Certainly, there'd have to be a few Nobel prizes awarded if there was a way to get around it, that’d be really convenient.
The movement towards sustainable transport I think that’s going to be good for many reasons, but again not something that happens immediately, that’s probably something that happens over 30 or 40 years, the transition to electric vehicles. Electric vehicles are something that is a long term sustainable option.
Yeah it’s really we're talking orders of magnitude of difference between fossil fuels and batteries, not even on the same scale. And given that we have to solve sustainable electricity generation, then it makes sense for us to have electric vehicles as the mode of transport.That's why I think it's important for electric cars to be able to compete without economics being a factor. We just need to fix the incentive structure of the world to make sure that companies are incented towards sustainable versus unsustainable technology. This is fundamentally the problem.
If the big car companies see that our sales are good, and if they see that people are buying these cars, then they will have no choice but to conclude that electric cars are the right way to go. And that will accelerate the transition to sustainable transport. There are lots of naysayers out there that say 'electric cars are never gonna happen, we should just be resigned to burning hydrocarbons forever' well not forever, until they run out of course. Then they'll say certain technologies like hydrogen fuel cells, and it's like, ah God, fuel cells are so bullshit, it's really rubbish. The only reason they do fuel cells is because they aren't really believers, it's like a marketing thing. I think part of it is that they felt for a long time that there was this need to be doing something. Since fuel cells were always 10 years in the future and always would be, they could say they were working on fuel cells and that would satisfy people.
Some will say oh hydrogen is the most common element in the universe. Yes, but not on Earth, which is an important consideration. I don't want to turn this into a debate on hydrogen fuel cells, because I just think that they're extremely silly. My opinion pre-dates Tesla. Not just because I have an electric car company, I could've started a hydrogen company.
Hydrogen is an energy storage mechanism, it's not a source of energy. You have to get that hydrogen from somewhere, it does not naturally occur on Earth. You either got to electrolyze water or you got to crack hydrogen carbon. If you crack hydrocarbons then you're just basically a carbon burning car in disguise, so there is no possible win there. If you get that hydrogen from water, the only way to get the hydrogen is to electrolyze water and split the hydrogen and oxygen apart, so you’re splitting H2O. Electrolysis is extremely inefficient as an energy process, a very energy intensive and inefficient process. Then once you split the hydrogen off you got to compress it or render it to liquid form, which also takes a tremendous amount of energy. Hydrogen is very difficult to store and transport, it is a very light gas so any tank you store it in is enormous, because of low density. Hydrogen has very low density. It's a pernicious molecule that likes to get all over the place. Then you got to store it, and then even once it gets in the car it then has to go through a fuel-cell power plant to get turned in to energy, and only then can it be applied to electric motors.
You get metal embrittlement from hydrogen. If you get hydrogen leaks, it’s an invisible gas you can't even tell that it's leaking. But then it's extremely flammable when it does, and has an invisible flame. Hydrogen is quite a dangerous gas, if it does escape is highly volatile and can I have extremely explosive consequences. You know, it's suitable for the upper stage of rockets, but not for cars. Putting up a huge hydrogen distribution structure is also extremely difficult. It's just very difficult to make hydrogen and store it and use it in a car. It's just takes in enormous amount of energy to create hydrogen. If you’re going to pick an energy storage mechanism, hydrogen is an incredibly dumb one to pick. You should just pick methane that’s much, much easier, or propane.
If there was a readily available source of hydrogen that wasn't bound up in water, or bound up in hydrocarbons, then there would be a possibility of a fuel cell vehicle. But the reality is that if you take a fuel cell vehicle, and you take the best-case of the fuel cell vehicle in terms of the mass and volume required to go a particular range, as well as the cost of the fuel cell system, and if you took best-case of that it doesn't even equal the current state-of-the-art in lithium-ion batteries. So there's no way for it to be a workable technology.
If say you took a solar panel and used the energy from that solar panel to just charge a battery pack directly—compared to try to split water, take the hydrogen, dump the oxygen, compress the hydrogen to an extremely high pressure—or liquefy it—and then put it in a car and run a fuel cell...it is about half the efficiency, it’s terrible. If you look at the total cycle of efficiency of a fuel cell system, it's theoretical best case is twice as bad as electric in terms of the energy cycle. That is if the technology is perfect. But the technology is not perfect so it ends up maybe 200 to 300% worse than solar electric alternative. The solar electric alternative is dead simple and the energy transfers are very high efficiency. Essentially collect power from the Sun, and then you charge the battery pack with almost no energy losses, and that's it you're done. The best case hydrogen fuel cell doesn’t run against the current case batteries, so obviously it doesn’t make sense. That will become apparent in the next few years. There’s no reason for us to have this debate. I’ve said my piece on this. It will be super-obvious as time goes by.
If you take the best case scenario for fuel cells, let's say you could fully optimize it and compare that to current lithium ion batteries in production, it loses. Success is not one of the possible outcomes, so why embark on that. Why would you do that? It makes no sense. It's crazy. The math is so super obviously in favor of batteries, it's like staring facts in the face and saying it's not true. At Tesla we call them "fool cells". It's one of those things that sounds like it's the future and it always will be.
I think in terms of energy storage, lithium I think is definitely the future and will be for a long time. The nice thing about lithium is that it is extremely abundant on Earth. Lithium is number three on the periodic table. It's actually extremely common. Lithium is the third most common element in the universe. The first being hydrogen, and the second being helium. Now the hydrogen is all bound up in water on Earth, water or hydrocarbons. You don't find naturally occurring hydrogen on Earth. The helium being a noble gas just doesn't combine with anything and basically floats away. But lithium is a metal and does not float away.
There is an enormous amount of lithium, in almost any salty solution there is some amount of lithium. There is enormous amounts of lithium in the oceans. Any kind of dried lake bed where there was a salt lake bed dried long ago there will be enormous amount of lithium. So the nice thing about lithium is that it is very plentiful. Any salty water has lithium.
There definitely won't be a lithium constraint on energy storage for batteries. I feel pretty confident that one could make enough batteries to store all the energy that the world needs with the current resources that are available. There is lithium in salt form virtually everywhere so there is definitely no supply issues with lithium. The actual amount of lithium in the world is far in excess for what is needed for electrification of transport. Lithium constitutes only about 2% of the battery, so it is like called lithium ion and it's sort of the active ingredient but it constitutes only about 2% of the battery.
I think we're actually in a pretty good spot, and I am reasonably optimistic that there will be a breakthrough in high energy density capacitors. It's sort of interesting. If you do the sort of basic physics on the energy density potential of a capacitor, using naturally occurring materials, it's quite hard to beat lithium-ion batteries, but if you can figure out a way to make unnatural materials I suppose, that are accurate to the molecular level, then I think you can actually have some fairly significant breakthroughs. The ability to do that was developed in the photonics arena and applying those photonics breakthroughs to capacitor technology is what has the potential for a really big breakthrough there. I think we may see something on that level, but it isn't entirely required for cars.
Once you've built the battery, then at the end of life of the battery you can recycle those components. It's something that has no long-term or negligible long-term impact on the carbon cycle, because essentially you get the lithium, nickel, and cobalt, and you create the battery. Essentially, you get those materials once, and then you recycle them forever. I think it's really a negligible impact for batteries on the environment.
I'm quite confident that solar power will be the single largest source of electrical energy for humanity in the future. We can actually generate way more energy than we actually need to operate civilization just with solar panels. I think the primary means of energy generation is going to be solar, it's at least going to a plurality or a slight majority. So no problem to generate all the energy we need for electricity, for heating, for transport, from solar with some contribution from wind and geothermal and tidal. No problem at all, we just have to do it.
Ultimately all of that has to go electric. That means a tripling of the energy consumption by electric. Thinking about that in context the demand for electricity will increase dramatically. It’s going to be very important to think about how do you make so much more electricity. That is such a huge problem, there's so much that can be done in that arena, that it's really more than enough to absorb I think any number of start ups and companies, because it's such a though problem and such a big problem. There are so many uses of energy that need to be sustainable.
There's obviously this sort of threshold to when solar power will become cheaper than conventional electricity, that’s a massive inflection point. There’ll be some long tail before the final coal plant finally stops operating, the final natural gas plant stops operating. There will be some long tail because it’s going to look like an S-curve as is typical for new technology adoption. In the beginning of the S-curve people tend to under-predict what’s going to happen and then it goes to an exponential growth phase, and then an approximately linear growth phase. Usually people over-predict what’s going to happen in the steep linear portion of the growth phase and then it goes back into a logarithmic to complete the S-shape. That’s what happened with the Internet for example, and cell phones, same thing will happened here. If you look at the growth rate of solar that's where it's going to go. Compound growth is very powerful.
I think everything will be completely electric, it's just a question of when, 100%, it's just a question of the timeline. Anything that we can do to accelerate that growth is a good thing, because it means we will have power as long as the Sun shines. I actually think, as long as the Sun is shining, we'll be fine. And if the Sun doesn't shine we have larger issues.