2018-12-31

Why Hydrocarbon Fuels Aren't Going Anywhere

Elon Musk has really done the surprising with Tesla’s commercially viable and high-performance electric cars. The general buzz around alternative energy sources piles on and one begins to wonder whether the days of gasoline are numbered.

From an engineering perspective, the reason electric cars are surprising is that compared to gasoline, batteries are heavy. The technical term is “energy density“: how many kilojoules of energy storage do you have per kilogram of battery or fuel tank? The following table is instructive:

Uranium: 80,000,000,000 kJ/kg
Hydrogen: 140,000 kJ/kg
Hydrocarbons (gasoline, jet fuel, fat): 40,000-50,000 kJ/kg
TNT: 4,600 kJ/kg
Lithium-Ion: 400-800 kJ/kg
Lead-Acid Battery: 170 kJ/kg

Nuclear fission of uranium has astounding, off-the-charts performance, but has its own set of challenges for mobile applications. Beyond some point for most applications, energy density isn’t what you’re really interested in. Uranium will probably stay confined to submarines, bombs, stationary power, and insane schemes for high-thrust, high-isp rockets.

Hydrogen is the theoretical maximum for chemical energy storage (assuming ambient oxygen). To sweeten the pot, it has the added perk of fuel cells for direct conversion to electricity, and water as its only oxidation waste product. But it has so many practical difficulties that no one really uses it for anything significant besides exotic space rockets, inflated rhetoric about an imminent “hydrogen economy” aside. Even with rockets, its practical difficulties all but outweigh its narrow-sense theoretical superiority over methane and hydrocarbons, so most new rocket designs don’t use it. The practical difficulties include the fact that it leaks through solid metal, degrades the performance of many metals through embittlement, is very low density requiring enormous tanks, requires special high pressure and deep crygenic thermos containment, currently requires exotic and expensive catalysts for fuel cell conversion, and so on.

Compared to the nightmare of hydrogen, the fact that the comparatively benign liquid hydrocarbons are well within an order of magnitude in performance is a pleasant surprise. Hydrocarbons are cheap, require only a plastic tank, and internal combustion engines are a reliable and very proven technology. It’s no wonder we use the stuff for basically everything from rockets, to planes, to cars. Even birds and animals use hydrocarbon-like fuels as their primary energy storage. The only real downsides are the CO2 pollution issue, and dangerous flammability.

This brings us to the humble lithium-ion battery, nearly two orders of magnitude behind hydrocarbons in theoretical performance. So far, it’s hard to see how one could get away with lithium-electric cars. For a long time, people thought it was impossible, or that they had to be slow, awkwardly streamlined, and with all the aesthetic grace of a golf cart.

But on nearly every metric besides raw narrow-sense energy density, lithium-ion claws its way back to competetiveness:

An electric car doesn’t need a transmission, the motor has only one moving part, you can drop the need for an alternator, the battery can have arbitrary shape and placement, there’s more room available for storage, and the whole system approaches being twice to three times more efficient. And in cars, weight doesn’t matter so much; one of the main reasons we cared about weight in cars was that it takes more energy, and thus more gas money and CO2 pollution, to lug it all around. Electricity to recharge is comparatively cheap, and the higher efficiency means you need less of it, so weight is much less of a concern. Its telling here that a Tesla weighs nearly 5000 pounds.

And it turns out you don’t need to design electric cars to look like plastic golf carts. Why design has anything to do with powertrain technology is a mystery, but people are weird.

So for cars, and maybe even trucks, trains, and boats, electric lithium-ion powertrains are going to gain market share over hydrocarbons, which are mostly tapped out technologically, though they will probably never completely displace hydrocarbons.

But what about flight?

Flight comes in two varieties: large human-rated aircraft, and small robotic aircraft.

For large craft, the gas turbine, whether jet or turboshaft, is the powertrain of choice. Turbines are expensive and finicky, but lightweight, relatively efficient, and very powerful. They are in many ways the optimal case for hydrocarbons, turning fuel into heat into motion in the most direct possible way, with a mimumum of hardware.

Elon, in his characteristic style, has mused about building electric aircraft to directly compete in this niche, but hasn’t substantiated these musings with any design details. The best we got was the implication of a craft with 70-80% of its mass dedicated to batteries. Of course only a rocket man could come up with something with a fuel mass fraction above 50%, and as a rocket man to the core, Elon has since settled on the hydrocarbon-powered (methane) passenger ICBM as his answer to the jet plane.

For small robotic craft, electricity is extremely convenient. You need electricity for the avionics already, electric brushless motors are cheap, light, responsive, and controllable, and batteries are simple and reliable, and can be easily swapped out for refueling. So small drones are electric, especially multirotors.

Drones are gaining in importance as the technology gets worked out, companies make more and more compelling offerings, and more applications are found. They could become very important.

Military drones in particular are a big question area. Almost certainly going to make an impact, but how and how much are still unclear. We’ll have to wait for the next big war to see for sure.

But as drones push closer and closer to the limits of possibility, the simplicity of slapping a battery on a frame with some avionics and motors is going to take a back seat to pushing the performance envelope. This will open the field up for more complex technologies that beat lithium ion at raw performance.

Nothing is going to beat lithium on power density any time soon: it already has way more than needed for anything but the most kinetic applications. To illustrate, a lithium-ion powered device that you can fit in your pocket can jump start a large car, with a peak current well over 300 amps at 12 volts.

But on energy density, hydrocarbons have a lot of slack in that nearly two orders of magnitude of performance gap over lithium to make up for the added complexity and inconvenience.

Lithium can usually only provide flight times of less than 30 minutes for multirotors, longer for fixed-wing. This is useful, but for many applications won’t be competetive with the much longer and more frequent trips possible with hydrocarbon fuels.

Hydrocarbon fuels for drones requires economical, compact, lightweight, reliable, and efficient generators that can be more or less just swapped in in place of batteries. There are multiple companies working in this space, and at least one such project will succeed in a big way.

In particular, I recently talked to a guy from Pegasus Aeronautics, a Canadian startup developing a compact gas-powered generator set for drones. Their edge is proprietary electronics for more efficiently turning the high-frequency output of the alternator into smooth DC that the drone’s electrical systems can use, and in-house construction and optimization of engine, electronics, and generator hardware. Promising, but we’ll see.

That conversation got me thinking about how solid the position of hydrocarbon fuels are in general, and what kind of permanent companies can be built in the space.

If hydrocarbon fuels are able to get a foothold in power plants for mobile robotics, which they almost certainly will, that foothold will very likely be permanent, and the overall market will expand with time as the use of robots expands.

The only real technical competitors to hydrocarbon internal combustion are:

  1. Lithium. As above, lithium will always have market share, but can’t dominate the whole market, leaving huge space for other technologies.

  2. Other battery technologies. There may be battery technologies superior to lithium, but they don’t exist yet. I used to work in battery tech, and I can’t come up with any realistic possibilities for beating lithium. The only dimension any other battery tech can beat lithium on is cost, especially at scale, and even there lithium is rapidly gaining ground on historically cheaper options. Still, something could come out of the woodwork here.

  3. Hydrogen. I’m not optimistic on hydrogen, for reasons described above. It’s most fashionable among greentech buzzword engineers, who sell equity and get grant money on the basis of high-minded but nebulous environmental PR, rather than selling product on the basis of superior technical merit. It’s notable again that hydrogen basically isn’t used for anything. When there’s more hype than tech, it’s a bad sign.

  4. Nuclear. Delayed critical fission isn’t going to flying any time soon, especially not in miniature. Radioactive decay does have a respectable record of powering spacecraft, but isn’t going to satisfy power density, cost, or legality constraints. Especially cost; we’re running out of Plutonium-239.

  5. Hydrocarbon fuel cell. The problem with non-combustive conversion of hydrocarbons is that you need to oxidize two separate elements with different chemical properties and activities (Carbon and Hydrogen). Edge cases like direct methanol fuel cells exist, but they have their own problems. Whereas for combustion, as long as it can heat the working fluid, its cool. Animals burn fat with advanced nanotechnology, but that’s way beyond anything that we can do right now. Solid oxide fuel cells do exist in this space, but basically don’t work yet so its hard to say how well they will actually do in practice.

  6. Other hydrocarbon internal combustion. Besides our familiar two stroke and four stroke piston engines, we may also see micro turbine and free-piston engines, especially as materials, manufacturing precision, and control systems continue to improve.

That leaves drones of the future to be powered by a combination of lithium and hydrocarbons, with gas prices, applications, and technology development determining where the boundary sits. But whatever happens, we can expect that companies making hard bets on hydrocarbon fuel in drones will not find any technological factor pulling the rug out from under them on the energy storage dimension. They will have at the very least a niche, if not an outright empire.

Again, the military question is central: the ability to fly payloads and observational capability around for hours, and to integrate into military jet-fuel based logistics, will be a key capability. Whatever company captures that market will do well. We’ll have to wait for war to see the full extent of this.

Overall, we see hydrocarbons losing market share in ground-based transport, gaining in robotics, and continuing to dominate flight in general. Even in ground transport, if demand falls, the price of oil will fall, opening up new demand. But more likely is that fracking and such eventually fail to continue staving off peak oil, and prices go up, driving alternatives to hydrocarbons.

But for sheer practical energy density, hydrocarbons can’t be beat, and for that reason, they aren’t going anywhere. This is important background to keep in mind for wondering what the future looks like.