Bjorn’s Corner: Sustainable Air Transport. Part 22. Fuel Cell system efficiency and mass

By Bjorn Fehrm

June 3, 2022, ©. Leeham News: Last week, we looked at the powers and thermals of a Fuel Cell system for aircraft propulsion. Our example was a cryogenically cooled system with a superconducting Motor, Inverter, and Cabling.

Now we analyze the differences should we not use the liquid hydrogen (LH2) to help with cooling the system to a superconduction state. What are the thermal and mass consequences of using conventional electronics and motors? The detailed discussion is in the sister article Part 22P. Here we summarize the findings.

Figure 1. The parts of a fuel cell propulsion system. We discuss cryocooled (graph) and non-cryocooled variants. Source: NTNU.

Conventional versus Superconducting Fuel Cell systems

The present in-development fuel cell systems for airliners (Project Fresson, ZeroAvia, Universal Hydrogen) are all non-superconducting. It has to do with these using gaseous hydrogen as fuel and, therefore, the cooling capacity of liquid hydrogen (LH2) is missing. To use superconduction to bring down losses and masses for the system components, these projects must add a cryocooler, which means additional complexity and weight.

With liquid hydrogen as fuel, we open up several architectural choices. We can bring the Cabling/Inverter/Motor to superconduction like last week’s NTNU system or use it to help with the size and mass of the Thermal management system.

So which way is the best? Is liquid hydrogen necessary?

The sister article Part 22P analyzes all three cases and compares these to a baseline case, the present gas turbine system for a 70 seater turboprop like the ATR72. Here is a summary of the findings:

The baseline

The 70 seater turboprop has an installation mass of the gas turbines, their fuel system, and Jet-A1 fuel for the maximum range mission of 6.1t. It gives the aircraft an 800nm range when transporting 70 passengers with bags.

The typical efficiency of the gas turbine alternative, defined as shaft power out to fuel power in, is 30%.

The Superconducting Fuel Cell System

The fuel cell system we analyzed in Part 21 uses the cool sink from the liquid hydrogen (LH2) to maximum effect. It runs the LH2 past the Motor and Cable system to bring these to superconduction, then past the Inverter, Converter, and Fuel Cell, cooling these (blue path in Figure 1). The result of the superconduction is a system efficiency of 37%.

When doing the cooling, the H2 heats up from -253°C to the 80°C it needs before use in the Fuel Cell. The total system mass is 10.1t for this alternative, where the LH2 Tank at 3.1t and the Thermal management system at 1.9t are the heaviest parts. The 1.7t LH2 fuel has the same energy content as the 5t Jet-A1 of the baseline system.

The Liquid Hydrogen Fuel Cell System

Suppose we do not use the LH2 to bring the components to superconduction; instead, we use it to assist the thermal management system. In that case, the mass increases to 12t, and system efficiency decreases to 30%. Superconduction increases complexity, but it really helps with efficiency and system weight. The question is if there is a way past it for airliner fuel cell systems?

The Gaseous Hydrogen Fuel Cell alternative

As a check, we also included a gaseous hydrogen alternative in 22P. The system now weighs 26t, 3t! more than the Maximum TakeOff Weight (MTOW) of the ATR72. System efficiency is at 30%.

It’s clear by these numbers that the current gaseous hydrogen projects are R&D projects to learn to master the multitude of challenges of airliner fuel cell systems. These are crawl before walk projects and have little to do with operational propulsion systems for airliners.

25 Comments on “Bjorn’s Corner: Sustainable Air Transport. Part 22. Fuel Cell system efficiency and mass

  1. I assume the heat from the fuel cell is used to heat the LH2 to +80°C besides normal air heating added with cabin exhaust air. Sounds like a LH2 gas turbine driving a fan and letting its exhaust heat the LH2 for the fuel cell can be a way forward if the fuel cell have a higher efficiency than the gas turbine.

    • Hi Claes,

      you see the LH2 blue path from the tank at -253°C going to the motor to turn it superconducting, then to the HTS (Cable system) doing the same, and finally to the DC/AC Inverter, DC/DC converter, and Fuel Cell cooling these. The final cooling of the fuel cell is regulated so the H2 has 80°C at the entrance to the full cell. Not only does the LH2 help with superconduction and cooling, but it also increases the heating value of the hydrogen in the process.

      • Thanks, and the LH2 flow that gets heated to 80C is enough to produce the required power I assume (around 2700shp).

        • It’s the power requirement from the Fuel Cell that set the LH2/H2 flow. That’s why the motor and cable system is prioritized for the routing of the lowest temp H2, then the rest of the cooling is regulated so the Fuel Cell gets its H2 flow at the temp it needs. The Thermal mgmt system must cover the heat management that the H2 path doesn’t do.

  2. The ATR72 — as an example of a 70-seater turboprop — has a metal fuselage, and its wings are only 30% composite.
    Could a re-design with a composite fuselage provide enough weight savings to compensate for the 4 ton increase in propulsion system weight of the superconducting LH2 propulsion system (10.1t) versus the Jet-A1 system (6.1t)? The existing ATR72 has an empty weight of 14t.

    • I don’t composites so far let to dramatic weight savings. I looked at similar sized A330-900 and 787-9, on which Boeing went all out composites, and the weight differences are not substantial.

      If Embraer comes with a substantial better TP, ATR might have to strike first with an aircraft upgrade, which route they seem to indicate with the EVO study. A study I see as conservative, risk aversive and putting the ball squarely with the engine manufacturers.

        • The 20% weight saving is a theoretical figure for parts that have predominately high tensile stress (where CFRP has its strength). For compression parts, CFRP is not good at all, so no mass advantage. A fuselage has few areas that are designed with tensile stress as the design parameter, most are damage tolerance (toughness), and there CFRP is not good. In essence, the mass advantage for a CFRP fuselage structure is considerably less than 10%.

          • A very interesting insight!
            Seeing the relative misery that composite fuselages seem to be causing (costs, tolerances, lightning protection, paint adhesion…), and the remarkable economy that can be achieved by a plane such as the A330 neo, one can ask oneself if “reversion” to metal fuselages might be a wise step…

          • The weight saving is soley from ‘metal’ to composite. As even a highly composite planes like B787 and A350 are still only around 50% composite of their overall empty weight.
            The real gains would also include new manufacturing processes and integrated with reduced maintenance schedules and costs.
            I think the manufacturers were hoping the carbon fibre products including resins would come down more in cost as usage rose , but it didnt really happen

          • @Dueofurl & Bryce, Aluminum Lithium and GLARE are both excellent alternatives for CFRP fuselages.

        • You normally start with wings in composite for low mass and stiffness, thus you can make them slimmer and longer. Composites right now only makes great benefit for aircraft flyging >11hr trips over Al-Li alloy wings. See the 777-9 design.
          As engines get more expensive and more fuel efficient more ultra long range routes will come like Qantas SYD-LHR A350-1000LR or very heavy cargo flights where you want to load lbs of payload instead of lbs fuel.

          • Composites are useful in wings across the size range. 777-9, A350, A220, some bus jets, gliders high end model air plane gliders.
            * they allow higher aspect ratios which increases wing efficiency
            * they can have more flex resulting in a “kinder” ride

            I doubt we will see a new design single aisle or larger commercial aircraft with metal wings again.

          • Yes, composite wings is the norm, whole composite fuselage is for +10 Hr Airlines flights

    • The Dornier Do 328, now Deutsche Aircraft D328 claims to have saved about 30% weight on the fuselage structure by an optimal blend of composites and metal.

  3. I’d guess your electrically powered compressor eats up a lot of your system efficiency. Going with a SOFC fuel cell would allow a turbine to capture exhaust energy to drive the compressor. Immature technology but promising.

      • Looks like RR is on a similar track to use the fuel cell heat to produce thrust. The simlest way is a compressor+ h/x + turbine and boost the electrical motor with additional shaft power with compressed air for anti-ice and cabin air cycle machine. It is Honeywell “territory” so you would expect a fuel cell package from them as well or if Liebherr can make a move to really compete with Honeywell on this technology.

        • Trouble is the ‘heat’ doesnt have enough energy to run a useful turbine. Remember a jet turbine is both a lot lot hotter and the pressure is high also, as very rough guide a turbo prop ( PW100 series ) has ‘overall pressure’ ratio between 12 and 17 ( bigger fan turbines can be 45).
          Cooling of electric components and to increase the LH2 to the temp for combustion is a better bet than extracting a small bit of energy from waste heat. Even in most turbines a lot of heat is lost out the exhaust as the weight and complexity isnt worth it. Remember GTF was designed for a variable area exhaust nozzle to increase efficiency but never had it installed for production versions.

          • Yes, you might need a heat pump before the h/x.

    • I think some pressure recovery is certainly part of it but I think there is also an significant amount of heat energy to recover as well, perhaps most of it.

  4. Slight of topic: Lilium Jet has achieved transition.
    The transition is at 1.30 into the video.

    If anything it demonstrates the power of electric ducted fans (EDF) or electric jets.

    • Correction, Lilium has achieved MAIN wing transition.

      The critical canard has not achieved a transition seven years into the program. The problem lies in the design.

      Civil canards are notorious for their CG problems (gives a limited CG range). The Lilium design worsens this problem. Before the canard aerodynamics change from stall to attached flow, the nose up keeping moment must come from the jet exhaust. But to get to an AoA that achieves attached flow, the canard, and with it the thrusters, must be angeled to around 10°-15° AoA. At that angle very little horizontal component of the jets remain = nose pitch authority. Even with an ideal CG, which is possible in test flights, Lilium hasn’t done it to date. For the final configuration the ratio of canard thrusters are increased to help with this problem.

  5. Bjorn: You mean “very little vertical component remains….” ?
    With this massive slow-response canard containing lots of mass, what about pitch stability in case of an upset ?

Leave a Reply

Your email address will not be published.