Bjorn’s Corner: The challenges of hydrogen. Part 24. Propulsion choice

By Bjorn Fehrm

February 12, 2021, ©. Leeham News: After covering the basics of fuel cells last week in our hydrogen airliner series, we now look at what type of system to choose for aircraft propulsion and onboard systems power.

We analyze the propulsion side this week.

Figure 1. A SAFRAN concept for a low emission airliner from its Clean Sky 2 presentation. Source: SAFRAN.

Is fuel cells a good choice for propulsion?

We have already described how a gas turbine can be converted from carbon fuels to hydrogen. In a conversion to burn hydrogen instead of Jet-A1, the size, weight, and efficiency stay the same, with no CO2 and 20% less NOx emissions. Contrail generation seems manageable as the crystals formed from the increased H2O emissions have a changed characteristic that hurt the environment less.

The alternative is to use hydrogen fuel cells to generate electric power to drive a motor that drives the aircraft’s propellers, propfans, or fans.

Figure 2 describes the alternatives we have to drive the thrust generating unit. From the top, the gas turbine core, the series fuel cell with its motor, and at the bottom, the parallel hybrid fuel cell configuration:

Figure 2. The propulsion alternatives for a hydrogen airliner. Source: Leeham Co.

The 165 seat hydrogen airliner

We start by looking at the EU Hydrogen-powered aviation reports‘ 165 seat airliner, Figure 3.

Figure 3. The EU reports’ 165 seat hybrid hydrogen airliner. Source: EU.

The key data for their simulation, which we also use, is given in the footnotes below the figure. It represents the best knowledge of parameters for a 2035 entry into service airliner.

We compare with today’s gas turbine airliner to give the report’s suggestion the most favorable setting. For this, we use the Pratt & Whitney PW1100 GTF as used on the A320neo (CFM LEAP data would be similar). The GTF engine weighs 2,868kg, according to certification data. Modeled with Gasturb, the core, including the fan gearbox, is about 2000kg and has a volume of 2.4m3. It delivers 20MW shaft power to the fan at takeoff thrust setting.

The EU aircraft is a parallel hybrid with 5.5MW electric cruise motors and 14.5MW gas turbines in parallel for takeoff and climb (the Figure 2 bottom alternative). A cryogenically cooled 5.5MW motor weighs 425kg when assuming 13kW/kg as we’ve done before and we use the report’s 97% efficiency (Magnix’s air-cooled magni500 560kW motor is at 4.2kW/kg and 93% efficiency).

If we place the fuel cells in the fuselage (which is plausible as they occupy 8m3), we need inverters both at the motors and the fuel cell (otherwise, the distribution cords pump 75,000A to the motors) to take the distribution to 3,000V (gives us 1,800A to each motor). These weigh 290kg on each side and 580kg at the fuel cell (we omit the weight of the distribution system to give the fuel cell alternative the best shot).

To this, we add a 12MW fuel cell so we can deliver 11MW to the fans as we have lost 1MW in the 97% efficiencies of the inverters and motors. It weighs 6.0t with the given power density of 2.0kW/kg (see Figure 3). We also add two takeoff/climb gas turbines and the gearboxes, in total  3,5t and 3.5m3.

In summary:

  • We have today’s turbofan driven A320neo that has two cores weighing 4t and occupying 4.8m3. A hydrogen-fueled 165 seat airliner would have the same core data.
  • The report’s 165 seat airliner with the hybrid hydrogen fuel cell propulsion would have the equivalent weight of 11.5t with motors and inverters occupying 3m3, the fuel cell 8m3, and the takeoff/climb turbines+gearboxes 3.5m3, for a total of 14.5m3.

The values for the fuel cell airliner are almost three times those for the gas turbine variant. It could perhaps make sense if the efficiency is considerably higher. It isn’t.

The GTF core efficiency at cruise is 55% (GasTurb simulation data), and burning hydrogen can only improve that. The above fuel cell alternative has an efficiency of 55% with the data given in the report. This is without counting losses in the distribution system.

So the gas turbine alternative is the best for the single-aisle airliner. Is it also for a smaller regional airliner? It’s often argued the fuel cell alternative wins at smaller size aircraft.

Regional airliner

Figure 4 shows the data for the report’s 80 seat regional airliner. The propulsive power to the propfans uses the middle configuration in Figure 2, a serial fuel cell alternative.

Figure 4. The EU reports’ 80 seat fuel cell airliner. Source: EU.

Calculated with the data from the report once again, the shaft power to drive the fans costs us 3.2t (we then don’t count distribution weight or losses). At this power level, a Pratt and Whitney PW127 driving the 70 seat ATR 72-600 weighs 481kg and delivers a maximum of 2MW, and has a 40% cruise efficiency.

The fuel cell system is 3.4 times heavier than the 30 years older PW127 (45 years by 2035). It provides a 54% cruise efficiency on paper, but is it worth 3t given a regional airliner’s short stage lengths? I would say no.

A 2035 EIS 2 MW turboshaft engine would be at 50%-55% efficiency and weigh around 600kg. It’s clearly the better alternative.


As for the battery-based electric or hybrid airliners, we find the promotors of fuel cell propulsion are far off the reality, even when we use their most optimistic data and leave all spiral consequences of the inefficiencies we find by the side. Why is this?

I can only assume they are blinded by the fuel cells’ 60% efficiency and neglect that today’s gas turbine cores are at 55%. When we add up the losses and the weight/volume penalties, it’s clear the hydrogen gas turbine is the clear winner for delivering shaft power to propellers, propfans, or fans for hydrogen airliners, be these 165 seater jets or regional turboprops.

Is it also the choice for an auxiliary power source? Now we look at delivering electric power to the aircraft’s systems instead of shaft power, and the fuel cell alternative can skip the motors and their inverters. The distribution network for power to the aircraft systems is also at a manageable level.

We analyze the hydrogen APU versus fuel cell in next week’s Corner.

26 Comments on “Bjorn’s Corner: The challenges of hydrogen. Part 24. Propulsion choice

    • I saw that too Bryce, there are definitely moves to be ready to have Hydrogen available as the world shifts to lower carbon technologies.

      As I think has been mentioned in other responses, Paris airports operator Groupe ADP “announced it was working with Airbus, Air France-KLM and other French entities to explore the use of hydrogen at Parisian airports.”

      Of course for this to be workable you probably also need a pipeline (similar to the UK’s GPSS/CLH) connecting the Hydrogen terminals to the airports, rather than delivering Hydrogen by road tanker.

      So are we going to see the first Hydrogen powered regional airliner in France before the rest of the world?

      • @ JakDak
        The pipeline network to which you refer is already largely present in Europe (for natural gas), and is to be extended.

        Regardless of whether or not it ever takes off for aviation (no pun intended), hydrogen is going to be a big thing for ground vehicles in the EU. Several bus companies in NL already operate fleets of hydrogen buses, and there are hydrogen-powered regional trains in Germany. Much greater range than battery-powered vehicles, and re-fuelling is a matter of minutes rather than hours at a re-charging station [fast charging kills battery life] 🙂

        Hydrogen is a wonderful means of converting every single available kWh of time-dependent excess electrical energy into a useful form.

        • Thanks for the link Bryce,

          I see a target of 2040, and that a large amount would be conversion of existing natural gas pipelines? So what do they do with all the current users of natural gas, do they convert their devices to use Hydrogen? E.g. home gas powered heating boilers, gas hobs etc…

          GPSS/CLH that I am referring to supplies aviation fuel directly to airports, parts of this network was built during the Second World War as a means to protect supplies, and to avoid road tankers being targeted during raids.

          • An interim solution is to blend hydrogen and methane for pipeline transport. Studies have shown this can be done at up to 20% to 25% concentration with the existing infrastructure. Most users can burn the blend, but separators can be used to obtain pure hydrogen at the needed destinations, with the separated methane being reinjected into the pipeline.

            That may be one method for mass hydrogen distribution to get started. Obviously works only for gas pipelines, not liquid.

          • @ JakDak
            Building an electrolysis plant at or near an airport solves the problem to which you’re referring. Although that may be a problem for a small number of urban airports, it won’t be an issue for most.

            In NL, a plan is being considered to build large refueling stations at strategic points along busy motorways, with in-situ production of hydrogen at an attached facility. No transport problems. Excess grid power is flexibly allocated to the locations that have the highest production requirements at a given time. Somewhat of a “just in time” model.

          • 2040 (only 20 years from now) sounds extremely optimistic. We don’t have large scale battery electric cars, we don’t have economical technology for recycling of lithium batteries, we are starting to see commercial compressed hydrogen vehicles (Toyota Mirai) but at a very low scale and these are stridently opposed by the BEV lobby. The renewable energy industry is still massively underwritten by subsidies and based on lies by omission (claiming wind has the lowest load levellised (NPV) cost of production but not quoting that network costs are the bulk of costs now due to renewables surge and remote nature. When cars are mostly electric BEV or fuel cell FCEV and Iron is produced by reduction with hydrogen, tractors are pulled by PtL fuels and Cement is carbon neutral then then powering aircraft by LH starts making sense. In the meantime it’s best to keep up publicity and futuristic plans and dreams reasoning won’t work.

      • Hydrogen can be generated locally from electricity. That will get things started.

  1. Depends on how expensive gas turbines are. Electric propulsion should be cheaper.

    The 165 seat fuel cell delivers 11MW, but the power density is 2kW/kg, not 1.75.

    • I deliberately used the power density from the EU study which was 1.75kW/kg. Last week I gave 2kW/kg but this was, as stated in the article, without the cooling system included (which costs power and weight).

  2. Figure 3 shows 2kW/kg, figure 4 shows 1.75, both incl. cooling.
    EU data of 3 and 4 is different.

    • OK, thanks, I corrected this. It doesn’t change anything but the numbers shall be correct.

  3. During the 20th century, people dreamt of crazy long-range aircraft. The 747 could fly across any ocean, then came the 777 and A350. The soon-to-come project sunrise kangaroo-hop A350 (presumably) will be the last paragraph of this particular story. It was a great journey, and kudos to all the great engineers!

    Now we need to reimagine the future dreams of aviation. I believe we need to reinvent the sub-1000 nm flight. When Wilbur and Orville Wright flew their first aircraft, they weren’t necessarily thinking of 7000nm flights. They were probably thinkning of 100 nm flights. Or perhaps 400 nm flights.

    I belive these future short flights need aircraft with a crew requirement of one, that are silent in operation, and can operate using relatively simple ground infrastructure. I believe electric propulsion could be a tremendous enabling technology.

    • You are looking at the Urban Air Mobility “UAM” market where electrical unmanned helo/aircrafts take 1-4 people from the airport to city helipads and regional flights all preprogrammed and connected to its ATC. That market will expand with its E-hang, Joby, Volocopter and Lilium with others trying as well many startups connected to car companies even Cadillac. Airbus and Boeing are doing their trails and watch closely.
      1-2hr 100-180 pax jumps is another business and clearly burning LH2 in gas turbines and using the immense possibility of cooling different engine parts with LH2 before it is combusted can improve the gas turbine alot or make it possible to be made of much cheaper materials like steel and aluminum alloys.

  4. Reducing the speed of a short haul aircraft from mach 0.74 to to mach 0.65 is going to lower the energy required by ~15%, and cruise power by ~25% with a negligible contribution to door to door travel time. Start the efficiency drive there, that lowers your fuel cell mass substantially.

    If you put your thinking cap on you can probably also put the power difference between cruise and takeoff into a battery system, and laminate all of the batteries and inverters into the leading edge of the wing. Combines inverter cooling with deicing, shortens wire length, and removes a huge amount of the structure needed both to stiffen the wing and reinforce the wing box.

    • Hi Chris,
      if you put in electric in our search box upright, you will get Corners where I check the idea of having batteries storing energy to use later. It doesn’t work, batteries are 70 times heavier than jet fuel and 200 times heavier than LH2 for the same energy content. It kills all battery ideas. Batteries works for cars where your problem is the tire rolling resistance, it doesn’t work for airliners where drag due to weight (induced drag) goes through the roof.

    • Hi Markus,
      no I’ve taken the typical voltage out from a fuel cell of 70V to 100V and the 5.8MW we have out from the fuel cell to each motor, then you need to pump around 75,000A to transport your power to the inverters at the motors.
      It’s to explain why we need inverters also at the fuel cell to increase the tension to avoid these high currents (that forces a heavy distribution system). One of the challenges for all promoters of electric hybrids that are not solved is the distribution of 10s of MW around an airliner that flies at 40,000ft.

  5. How would this math work out if you used Airbus’ concept with high wing and pods holding LH2 + fuel cell + motor & propeller? (No wires + only one inverter per pod) Just a long shot Airbus theory catering to popular beliefs, or a concept that could have some future? And if future, what would be required for this to work? Smaller/lighter inverters cooled to superconducting state by LH2? Necessary target energy density of Fuel Cells for this to work? I do of course understand that Airbus has a legitimate need to show that they are evaluating all options at this point, even solutions that they might not fully believe in.

    • I have just read that Airbus see the multi pod 100 passenger, 1000km range as the low hanging fruit for first development. I suspect larger jets (to 250 seats) will take 10 years longer. Long range Wide body may never change. Such a tiny part of the whole emissions picture

  6. Once in a research meeting we discussed fuel cells for commercial aviation, the conversation went like this: “Some people are looking into using fuel cells.” … “Ok, but hydrogen is combustible, why not just burn it?” and that was the end of that discussion.

  7. If the pressure (and perhaps temperature) of the gases going into the fuel cell are increased, what effect does that have on the power output, and consequently the specific power output?

    Also, opportunities exist for a hybrid power-cell/gas turbine combination to improve efficiency through intercooling, recuperation and prevention of diversion of costly high pressure compressor air for cooling purposes.

    I understand (from the internet, I admit) that fuel cells are more efficient than gas turbines – by 10-20%. Surely there must be a place for them, if only to replace the generator on a gas turbine gearbox?

  8. In this study which is focused on fairly short flights, is it relevant to only consider the GTF core efficiency at cruise? In this sort of mission profiles, which fraction of the total fuel burned is burned during cruise?

    • Very relevant question. For a regional aircraft on short routes, when you integrate shaft horsepower for the whole mission and do the math with block fuel, you will find that a PW127-class engine efficiency falls down to the low 20%’s (taxi and descent are killers). A hydrogen turbine will not escape that fate either. Fuel cells on the other hand have their peak efficiency at lower power settings (so you normally oversize them to takeoff power incidentally no longer needing batteries) and also have comparatively very low fuel flows for power settings near zero thrust, so the mission efficiency generally ends up near 50%. Or twice as good as turbines and zero NOx emissions.

Leave a Reply

Your email address will not be published. Required fields are marked *