The world of aviation is changing. Rational analysis makes it seem obvious that short- and medium-haul flights are going to end up being battery electric, and long-haul flights are going to be running on upgraded biofuels. But there are both challenges and deep biases with that paradigm that prevent this from being a universal view.
Having done the work on this in various forms over the past several years, performing economic comparisons of different fuel pathways, looking at battery density vs cost vs weight curves, talking with people like Paul Martin who have worked professionally with hydrogen for decades and looking at the specific challenges of aviation-related warming, I don’t tend to spend much time on alternatives to this world view anymore. However, three things triggered a bit of a deep dive in the past few days.
The first was a reaction to my discussion with Heart Aerospace CEO Anders Forslund. He and his team are following a very pragmatic, incrementalist step to electric aviation. The second was Wright Electric’s announcement that it was going to have a 100-seater electric regional jet certified and in operation by 2026. The third was a pair of announcements by ZeroAvia that it would have 19- and 76-passenger planes fueled by hydrogen in service by 2024 in western Europe and the Pacific Northwest of the US, respectively.
But first, the challenges of aviation fuel replacements, in brief.
Jet Fuel Is Excellent If The Atmosphere Is A Sewer
Pushing an airplane up into the air, keeping it there for hundreds or thousands of miles, and landing it with safety reserves takes a lot of energy. Jet and turboprop fuels are in the kerosene and naptha-kerosene families. They are fairly cheap, about $2 per gallon or $0.55 per liter pre-pandemic, although they are up 10-20% right now, along with most fossil fuels. They have a high energy density per weight and volume. That means that they don’t take up a lot of a weight or space in a plane, allowing sufficient cargo or passengers to keep costs reasonable and prices relatively low.
And they are easy to manage. As liquids at most normal temperatures, they are easy to pipe, truck, and pump. They are easy to move around planes from fuel tanks to engines. They are both efficient and effective, as long as we are allowed to treat the atmosphere as an open sewer and don’t care about global warming or air pollution.
It’s really hard to replace something that’s cheap, light, effective, and convenient.
Certifying Commercial Aircraft Is Difficult & Expensive
One of the things many aerospace engineers have been delighting over in the past decade or so is the sheer freedom that fully electric drivetrains provide for aircraft. A whole bunch of constraints have been lifted, and as a result, many truly odd looking airframes are popping up. Tilt-rotor systems like Joby’s, scaled-up quad-, hexa-, and octacopters like Ehang, and tilting wing 36-turbofan designs like Lilium get lots of media attention and YouTube views. There’s even one that uses the same principals as Dyson fans for propulsion, which took me a bit to wrap my head around.
But none of these are remotely certifiable in any reasonable amount of time.
Heart Aerospace’s strategy is to build an easy-to-certify, bog standard, high-wing, 19-passenger plane. 19 passengers is key because regulatory approval for a 20-passenger plane is the same as for an 800-passenger plane. The high-wing, suspended aerodynamic shoebox design is just a small Dash-7 and will use standard avionics and other components. All the novelty is in the nacelles, with exactly zero novelty elsewhere. This gives the advantages of an airframe developed specifically for batteries and electric motors with a comparatively easy path to certification.
Wright and ZeroAvia are taking the alternative route, one trod by Harbour Air already, which is to buy an old airplane that has a certified frame, and shoehorn the new electric motors, fuel tanks, power plants, and the like into it. This is a lot like Tesla’s choice with the original Roadster, but is potentially a lot less of a pain than the Roadster turned out to be. This leaves them with a pre-certified airframe with a flight safety record, a maintenance record and the like, and the novelty limited to a replaced power storage, delivery, and motor system. Both have the intention of building net-new airframes at some point in the future, but both are more concerned with the drivetrain.
Wright is intending to incrementalize that further by replacing one engine at a time until all four engines are replaced with electric motors by 2026. There are pros and cons with this, of course. Just as electrifying vintage cars or even manufacturing new electric cars with batteries and motor shoehorned where gas tanks and engine were leads to significant compromises, so does trying to put new power systems into planes optimized for the aforementioned cheap, light, effective, and convenient fuels used today.
Hydrogen Decreases Revenue & Increases Costs A Lot
In the case of ZeroAvia, it is taking a 90-passenger Dash 8-400, ripping out 14 of the seats, about 16% of paying passengers, and presumably some of the luggage area, to jam the hydrogen fuel tanks and fuel cells in. Wright’s white paper on its fueling options makes the same point about hydrogen as a fuel, and notes that they would have to get rid of 20% of passenger seating. So that’s problem one with hydrogen. You basically take a 16-20% revenue hit on every flight for a comparable airframe. That’s a problem.
And hydrogen is a lot more expensive as well. Wright’s white paper on its fueling choices make some really interesting choices to cost justify hydrogen. The company uses a 2040 price projection prepared by McKinsey for a hydrogen-aviation consortium that claims that delivered liquified, chilled hydrogen at airports — assuming a massive hydrogen economy and all ground vehicles being run on hydrogen — will be at $2.60 to $3.50 per kilogram. In 2040 dollars at that, or about $1.6 to $2.20 in 2021 dollars.
$1.6 to $2.20 per kg H2. Retail. Delivered. Chilled. To say that this is as likely as pigs developing wings and becoming personal aircraft for farmers is to be unnecessarily harsh to the chances of the pigs.
As the Lazard LCOE for hydrogen makes clear, you need both incredibly cheap electrolyzers — cheaper than anything we build now — running at 90% of capacity 24/7/365 with $20 per MWh electricity to get to below $3 production costs. That’s not wholesale cost, that’s the cost of production. That’s not retail cost. That excludes storage, distribution, and chilling costs.
Black, brown, and gray hydrogen from fossil fuels with no carbon capture delivered today in sufficient volumes to power cars and trucks runs about $15 per KG retail in the US, and $8-10 in Europe where there are hydrogen valleys. Without chilling, just pressurization.
Wright, ZeroAvia, and others are betting on retail hydrogen costs in 2040 being about 10-15% of what they are today. Once again, flying pigs are more likely.
Carbon-neutral hydrogen will be more expensive than today’s retail, bulk hydrogen. Both carbon capture and sequestration at natural gas steam reformation plants and electrolysis with electricity are going to be more expensive than just throwing away the CO2 from natural gas processing. There is no way in which that economic circle is squared. What is more realistic is to assume hydrogen, chilled and delivered, would settle at around $8-$10 in 2021 dollars if a future hydrogen fuel economy ever came to pass.
Jet A fuel has an energy density of 42.8 MJ/kg. Hydrogen has an energy density by mass of 120 MJ/kg. That means you need, all else being equal, a third of the mass of hydrogen to go the same distance. A kilogram of Jet A fuel is about 1.2 liters, so it costs about $0.72 at today’s prices. At $8 per kg and at the higher energy density, hydrogen will cost about $2.70 for the same amount of energy, or about four times more.
Aviation fuel was about 19% of an airline’s expenses before the recent price increase. Low-carbon hydrogen would increase overall expenses by close to 60%, and 50% of total airline expenses would be for fuel.
So, 20% reduction in revenue, and 60% higher operating costs. That turns into double the cost per passenger or cargo mile. If hydrogen were the only choice, this would be potentially feasible. We’d end up flying a lot less as the market did its magic, but okay. And that’s the end state. Actually low-carbon hydrogen provided for aircraft in the next decade in Europe and North America is 20%-50% more expensive, so there’s a lot of additional fuel costs that will exist before the bottom end stabilizes.
However, a hydrogen-fueled aircraft would have to have hydrogen fueling infrastructure at all possible diversion airports as well, otherwise the first time it lands somewhere except at its start or beginning, it’s stuck there until masses of liquified hydrogen can be delivered to wherever it is at great expense. Yeah, double the cost before adding in all of the infrastructure and supply chain costs of making hydrogen available at every likely diversion airport.
Wright recognizes half of this problem, the 20% reduction in revenues, but bases its fuel cost projections on the flimsiest of tissue. Even then, it doesn’t like the numbers.
And so that’s the first part of this series on refueling aviation. The problems are not intractable, but they do substantially slow the pace of change. And hydrogen is a very expensive alternative, so other options that are less expensive and have equal or better characteristics are going to dominate. In part 2, Wright’s really odd idea about aluminum air batteries — something shared by some enthusiasts and a bunch of academics it appears — is addressed. In part 3, biofuels and the summary.