Bjorn’s Corner: The challenges of hydrogen. Part 23. Hydrogen fuel cells

By Bjorn Fehrm

February 5, 2021, ©. Leeham News: Last week, we started the discussion around fuel cells as a source of electric energy in airliners. We went through the principle and asked some vital questions.

Now we look at different types of fuel cells and for what applications these are suited.

Figure 1. The principle of a hydrogen fuel cell. Source: Airbus.

What fuel cell type is suitable for aircraft use?

Last week, we learned fuel cells use an electrochemical process to release the energy stored in hydrogen. The knowledge around fuel cells has been built over the last 90 years after Francis Bacon developed a stationary alkaline fuel cell in the 1930s. This type is still in use today, predominately as backup power in industrial applications (first column in Figure 2).

Figure 2. Summary table of fuel cell types with main characteristics (click to see better). Source: GENCELL and Leeham Co.

The fuel cell type for aircraft applications is the second type, Proton Exchange Membrane, or PEM fuel cells. These are in intense development for mobility applications such as forklift trucks (operating indoors), trucks, buses, cars, and now aircraft applications.

The PEM cell is the lightest type, and it has good volumetric density. Present developments aim for a system installed weight (fuel cells need control and cooling systems) of 2kW/kg and a volumetric density of 5kW/l.

The fuel cell emits electricity, heat (the 2H2 + O2 → 2H2O reaction is highly exothermic), and water from an input of hydrogen and oxygen. The heat from the cells is a cooling problem as long as it can’t be used somewhere on the aircraft. In static building applications, the heat is used for heating, and then the system efficiency reaches 80% or more.

The PME fuel cell efficiency of 50%-60% is on the cell level. If we include system losses due to the fuel cell’s cooling and control, the system-level efficiency drops around 10%-15%.

The PEM fuel cell has an anode, the PEM, and a cathode. Each cell delivers around 0.7V, and the current dependens on the cell surface. To get practical power levels, several cells are combined into stacks.

Figure 3 shows a stack from the PEM fuel cell company Ballard. This stack is designed for cars in cooperation with Audi and has a 140kW rating.

Figure 3. The Ballard/Audi FCgen-HPS fuel cell stack for cars and other mobility applications. Source: Ballard Power Systems Inc.

What we don’t see in Figure 3 are the control electronics and the cooling system. Fuel cells have no moving parts, which makes them reliable. But they are sensitive to the environment where they operate, see the table in Figure 2. There is, therefore, a lot of further development required before a fuel cell system can operate in the non-pressurized part of an aircraft.

Airbus has started this work with a German fuel cell supplier, ElringKlinger, and Universal Hydrogen, that targets the turboprop market with their ready-filled hydrogen gas tube stacks, cooperates with fuel cell company Plug Power.

We now have two ways to provide energy for propulsion and aircraft systems, hydrogen-burning gas turbines and fuel cells. Which type is suited for what purpose? It will be the subject of next week’s Corner.

34 Comments on “Bjorn’s Corner: The challenges of hydrogen. Part 23. Hydrogen fuel cells

  1. Is there any discussion of the use of a heat engine to recover and usefully employ heat dissipated from a fuel cell?
    Traditional tumble dryers in Europe used to have an energy label C, with a power outlet rating of 2.6kW, but modern models now use heat engines (heat exchangers) to recover “lost” heat, leading to an energy label A++ and a power outlet rating of just 900W — a stunning increase in efficiency.

    • That’s possible because the useful output of a dryer is heat, and the waste product is also heat. For a fuel cell, the useful output is electricity. As Bjorn mentioned, if there is a heating load on the aircraft that can utilize the heat, as for buildings (cogeneration), then the efficiency can be higher, but the energy input can’t be reduced.

      A Stirling cycle engine might be used to transform the waste heat into electricity, but the Carnot efficiency would be low (maybe 25%) due to the operating temperature of the fuel cell. Real-world efficient perhaps 10%. And the engine would increase overall weight. But it is possible.

        • The mechanism of moisture removal is evaporation, which requires the heat of vaporization to be supplied. All dryers produce heat in some form. Even a clothesline relies on the sun and ambient air to supply the needed heat.

          The process is accelerated by the gradient in partial pressure of water vapor, which is also a function of temperature, and thus of the heat supplied.

          • “Inner” process or “outer” process?

            In the efficient case you reclaim the heat of evaporation by condensation. … and you only have to provide for minor losses.
            ( “inner”, blackbox process )

            In the evaporation by heat followed by getting rid of the water vapor you see an “outer” process. about as inefficient as it can get.

      • Cabin air in/out could be improved with heat from the APU. Temperature at cruising altitude is such that using a heat engine is definitely possible but it is always the question if the added weight makes makes the plane in total more efficient. And then there is the cost and complexity issue.
        A H2 plane would be IMHO have a simpler turbine without bleed air or electricity generator and an APU who does those tasks. The main advantage is that this would allow the turbine to be much easier be improved

    • Tumble dryers remove moisture from clothes. The old tech relied on joule effect heaters (same as hair dryers, cheap & compact). New tech relies on heat pump cycles:
      – heat air in the condenser;
      – dry the clothes;
      – remove the moisture & recover the heat in the expander;
      – start the cycle again with the dried air, now cold but ready to be heated.
      Big gain against joule effect heaters.
      Stirling engines are unfortunately low specific energy devices, NASA tinkered long time with them to enhance power production of RTG with dismail results.

    • There is a big discussion regarding heat recovery of a different type of fuel cell called an SOFC Solid Oxide Fuel Cell. These operate at around 1000C. Hear recovery is via a gas turbine and they are refered to as “turbo compounded”. They are pressurised.

      • @ William
        My use of the term “heat engine” was intended to be construed broadly. Perhaps a phrase such as “heat energy conversion mechanism” might have been more applicable. A turbo mechanism coupled to an internal combustion engine is, of course, a pertinent example of heat recovery.

        Yes, I read that the operating temperatures of SOFC fuel cells are very high. For PEM FCs, I’m reading operating temperatures of about 80C. An adjusted configuration could easily get this above 100C, which would enable steam production for use in a turbine to generate electricity. Any (lower-enthalpy) heat remaining after that could be diverted to things such as fuselage heating / de-icing, etc.

        One way or another, it should be possible to put any heat produced to good use.

        • In the turbo compounded SOFC the fuel cell replaces the combustion chamber in a gas turbine. You need a blower/compressor to move in the air and maybe increase its density.
          anyway. The turbine recovers some of the pressure and heat in the exhaust. I imagine even a PEM cell would have some pressure in the exhaust to recover. It’s a fascinating thought that a turbo compounded SOFC could be 85% efficient.

        • There are heat recovery engines known as ORC organic ranking cycle that work of around 60-70C. They’re in widespread use but way to big for an aircraft by a factor of 10. Not worth it

    • You could buy “condensing” dryers for decades.
      BUT: more up front investment for a smaller electricity bill.
      guess what people bought 🙂
      increased energy pricing and a (tiny) bit of green thinking has moved that balance.
      you see a similar process in living quarter heating. improved insulation low circulation temp systems going towards Zero Energy building codes brought more interest to (electric) heat pump installations.

      on fuel cells:
      fuel cell power generation is not bound to Carnot process limitations. 95% are intrinsically possible.
      Limited by losses but not by process limitations like in the Carnot process case.
      i.e. there is a lot of room for improvements for fuel cells.

      • You may be thinking of a heat pump air dryer rather than a condensing dryer. Condensing dryers use a counter flow heat exchanger to use ambient air to condense the hot moist air before its vented. They are not energy efficient but they do reduce condensation in the room from damaging your house which is a kind of energy efficiency itself. The more advanced and expensive Heat pump dryers are extremely efficient and also very expensive though do a better job. My advice is foremost to have a washing machine with high RPM to get the clothes dry. The tumble dryer is secondary. If you are living in a sunny climate you may usually sun dry and use your dryer only on the occasions it rains 3 days in a row. In general a lot of so called “efficient” technology isn’t because the cost represents lost opportunity in more important energy efficiency investments. Quadruple glazing for instance.

    • In theory if you have a hybrid solution with fuel cell APU and LH2 combusition jet engines, the fuel cell heat can be dumped into the fuel going to the jet engine. Having cold fuel and heat you can run a Sterling engine helping the electrical engine driving the cabin compressors still I doubt it is worth the cost, mass, plumbing and risk.
      There is a possibility to have some of the LH2 fuel in streamlined pods ahead/below the wing to help flutter damping on very long and slender wings.

      • Good idea but I suspect there are better things to do with that cryogenic cooling effect. The Allan Bond developed Reaction Engines SABRE engine used the cooling capability of the hydrogen to reduced the volume of the air by a factor of about 4 (just above the pinch point where the air becomes liquid). The resulting turbo machinery (compressor, turbine) were much smaller and much more efficient. You need a counter flow heat exchanger that doesn’t ice with water or CO2 but that has been developed. also the super conducting generator/motors can get rid of gearboxes to match turbines to compressors and fans. In addition contact free super conducting bearings will be wear and friction free.

  2. Does’t the old school HHO systems where a wet or dry cells produces hydro/oxy gas through electrolysis works to improve milage on vehicals.
    Once tuned well it can help reduce consumption by 50% wt almost zero emissions??

    • It’s possible, but the overall efficiency of the complete cycle is quite low. The energy is stored as electricity, then used to electrolyze water. Then the hydrogen is burned in internal combustion.

      A fuel cell is somewhat the reverse of that, hydrogen is consumed to produce electricity directly, which then can be used directly for motive power. Fewer steps so fewer losses, and higher overall efficiency.

      Since the hydrogen must be produced somewhere, that conversion efficiency must also be factored in for the full cycle. But the case is better if renewable energy is used for the conversion.

  3. I see Norway has this week exceeded its previous maximum electricity production. They have the largest proportion of electric cars in Europe and are high users of electricity in the home for heating…yes its been very cold this week as well. They are using just as much electricity this week as neighbouring Sweden, which has twice the population and is more industrialised.
    If Britain used as much electricity per capita as Norway for this week alone, they would need to grow generation from 5oGW to 300GW and thats not even considering many heat intensive heavy industries who are a long way from non fossil fuels. Having enough electricity for everyone and all the future hydrogen production is a massive ask.

    But back to the mainpoint of this post and fuel cells of which is the way forward ?
    using electricity from generation for hydrolysis of water to produce hydrogen directly as a fuel which is burned or to use hydrogen in fuel cell to create the electricity again ( which is what you started with).

    Thinking back to the coal powered steam age for transport we know that was replaced in locomotives by 2 other choices. Electric traction from over head wires ( or 3rd rail) or diesel as the fuel source.
    Ships and planes cant run from wires or rails so that leaves the other choice, a refined fossil fuel that was a useful liquid.
    Diesel was interesting because they didnt use it to replace solid coal and burn it in a boiler even though it was liquid and higher energy value. Instead it was used in a different combustion process to create electricity on demand in the locomotive and this was then used in electric motors to power the diesel electric locomotives.

    Aviation is the same point , do you replace an existing fossil fuel with another non fossil liquid which is burned in the same combustion process as before or do you go with the same hydrogen which by a round about way- fuel cells- is used to directly create electricity to create thrust .

    It maybe that LH2 burned directly may be the answer, as the turbine is far more efficient than the old coal burning locomotive could be be with a liquid fuel.
    Still the issue the much higher production of electricity comes from where, but in the time honoured way thats some ones elses problem.

    • Agree.
      Personally, I think LH2 propulsion of aircraft is a last resort: I think there’s far more potential in biofuels, such as algae-based fuels. Why use electricity to create synthetic energy carriers when microorganisms will do it for you much more efficiently using sunlight? Of course if we ever get nuclear fusion up and running, the balance will shift.
      But, seeing as hydrogen-based propulsion for ground transport is now being rolled out to an increasing extent, it’s no harm to consider what would be necessary to extend this to air transport. At the very least it’s some green PR for Airbus, and will let some air out of Greta’s balloon.

    • The hyperloop is almost like electrical flying planes without wings and electrical powered. The tunnel i.d.’s is approx the size of a 737/A320 fuselage but flying at 1″ altitude at 100 000′ atmophere pressure.
      Still it has many challanges to solve including the still unknown certifiation requirements of the pods and the tracks with its switches and stations.

    • Production is not Consumption.

      Norway is exporting an awful lot of electricity at the moment. And they can because all of their hydro-power

    • Björn covered PtL “Power to Liquids” which converts C02 extracted from air and Hydrogen split from water into a carbon neutral hydrocarbon. It’s components are mature and fairly efficient. It’s seems it’s only problem is that the opinionated chattering classes are obsessed with Tesla batteries as the only solution.

      • Just wait until governments start to add a “recycling tax” to the price of new EVs, to preemptively cover (part of) the costs of collecting and recycling the batteries in cars. That might help the penny to drop 😉

        Moreover, when Li can no longer be easily extracted from salt beds and has instead to be mined in much more effort-intensive manners, just watch what will happen to battery prices.

        But you’re right: the average Tesla owner doesn’t reflect upon such matters.

        • Bryce, if you haven’t watched the Tesla Battery Day presentation then I recommend you do. It answers a lot of what you mentioned. They are looking at 1 billion miles from batteries. That’s not unreasonable because Tesloop ran a Tesla shuttle service between LA and Las Vegas and had a vehicle with 300,000 miles on the original battery and still going strong. Telsa also talked of recycling batteries and in classical Musk fashion said that there will be a point when no more Li needs to be mined, it will all be recycled.

      • Electricity2gas needs conservative 6x as much electricity than a BEV. It is a solution that is technically possible but idiotic in real life.

        • From the point of view of aviation battery energy densities are so low that an aircraft that flies greater than 100 km than an electric one will be three times bigger and this negates any advantage easily. The problem is that the battery insisters like to pretend that the system of conversion, transmission, charging is effectively 100% efficient, that the cost of large batteries doesn’t matter, nor that transmission costs (about 2/3rds now) doesn’t matter. They live in a fantasy. If the conversion efficiency of the PtL is 60% (and it can go higher with certain high temperature processes) and the conversion from Liquid to Power is 50% (possible in a HCCI + Hybrid engine) the overall efficiency is 30%. This is acceptable because 1 much of the PtL will be direct solar to liquid bypassing electrical generation and appearing as about 100% efficient making it look like 100% efficient. Electrical power has significant transmission and charging losses, it can store energy over a diurnal cycle but not seasonal or even weekly cycle. PtL also provides a great way to store off peak energy. Pluggable hybrids work well with PtL. Can you imagine a busy holiday weekend, say thanksgiving or Easter where millions of cars are trying to charge at once? Finally from the point of view of aviation battery energy densities are so low that an aircraft that flies greater than 100 km than an electric one will be three times bigger and this negates any advtnage easily.

          • It is idiotic for wheeled transport like cars. Aircrafts it may be a solution.

            Your PtL numbers are highly, even unrealistic, optimistic and your LtP can only be achieved with very expensive, complicated and large engine systems. Burning PtL liquid in a electric power plant and charging the EV for Easter is probably a cheaper and easier solution. There is also the question if total demand for electricity would peak on Easter, thanksgiving etc. in an EV world. People would use more energy to drive their car to visit family but factories etc. would be largely closed and few people would drive to work so total demand could be unchanged.

          • You keep saying “it is idiotic” but haven’t provided any numerate or informed answers.

            While direct conversion of electricity to liquid fuel is around 55%-60% at present there are other routes. PtL can use direct thermochemical water splitting thus bypassing the expensive and inefficient intermediate conversion to electricity there is also thermocatalytic (cerium oxide) that directly produces syngas with water and CO2. Low temperature processes (under 70C) can be used to liberate CO2 from absorbers using otherwise unusable waste heat. There is also photochemical and photocatalytic production of hydrogen, syngas and methane simply using sunlight.

            Using nuclear HTR or molten salt also bypasses the electrical stage. This is what China is planning.

            From the point of view of providing motive energy all that matters is land area and cost to deliver certain number of kw at the axel or shaft not some narrowly conceived ‘efficiency’

            How do you propose to export energy from the middle of the Saudi, African or Australian dessert?

            PtL integrates well with concentrated sources of CO2 associated with say fermentation or biogas production or cement making. It stores energy for years, can work with excess energy during peaks. It can work with solar, nuclear, wind, remote ocean wind and transport vast distances.

            When experts like you disparage the industry you destroy it to ensure the subsidies go to your pet area.

            If used in automobiles it will work well with hybrid and plug in hybrid.

          • It is so obviously idiotic because the efficiency of an ICE is so incredible bad and they have as second problem that they are dirty. Burning the PtL in a power plant with easily double* the efficiency and recharging the vehicles makes much more sense.
            Power plants have as advantage that it can use the PtL directly, ICE needs reformulation to fuel. Liquid can be barged, railed or pipelined in, fuel station likely less efficient trucked in. CO2 capture possible unlike with cars. And can be used for other high energy demand things like a coldsnap.

            55% is highly optimistic numbers for PtL and getting the massive concentrated CO2 needed for PtL for road use besides the less easily substituted PtL for Plastic, Aviation, Ocean, Back up or Legacy seems to me only possible with energy intensive methods.

            Efficiency should be miles driven, not kW on the wheels. EV need less cooling and are easier to streamline, thuse have better kW per mile numbers.

            Moving water & CO2 into the middle of Australia/Saudi Arabia is much more difficult than building an electric cable.

            It is not a question if it will work. It could but going full electric is just a much easier & cheaper solution with liquid fuel having the same market share in road transport as LNG or coal now.

            *) ICE is theoretically 30% efficient, power plants are above 60% in practice.

          • The internal combustion engine on Toyota’s hybrid synergy drive averages 40% efficiency in typical use, more in some regimes. The efficiency of low speed diesels is 55% and HCCI/RCCI automotive engines potentially will achieve 50%.

    • Planned for H2 airplanes in 2035. In between we are going to use it to power vehicles at the airport that move max. 100 km a day and for which batteries are a much more logical solution.

      Blue washing

Leave a Reply

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