Bjorn’s Corner: Electric aircraft, Part 8

By Bjorn Fehrm

August 18, 2017, ©. Leeham Co: In this Corner, we will finish the design of the hybrid propulsion system for our 50-seat regional turboprop. We use the ATR42-600 as a reference, as before (Figure 1).

We found an acceptable redundancy concept in the previous Corner, with an APU+generator+battery as backup power source. Now we will finish the design of the hybrid propulsion chain and compare with the original turboprop propulsion.

Figure 1. ATR42-600 serves as a template for our 50-seat regional turboprop. Source: ATR.

Hybrid power chain

The power chain that we envisage for our aircraft is shown in the lower part of Figure 2.  We want a large gas turbine to drive our main generator (a large gas turbine is more efficient than a small).

Figure 2. Hybrid chain with APU backup compared with Turboprop chain. Source: Leeham Co.

As backup power source, we have a smaller, less efficient gas turbine (an APU) driving our backup generator. During peak power periods, with one engine missing (take-off, climb), the battery chips in. The APU is also our source for ground power and cabin air conditioning during passenger turn-around. We also use it to start our main gas turbine after the stop.

The battery system is now a 46kWh battery. With a specific energy of 0.3kWh/kg, we get the mass as 155kg. To this, we need an inverter to convert the DC current to AC for the propulsion motors. The mass of the inverter is negligible compared to the other components. We will discuss the lower mass components we need to add in a complete system in the next Corner.

The dual propulsion motors, of 1,800kW each, drive the propellers. For normal use, we need maximum 1,600kW (take-off, climb) but at One Engine Inoperative (OEI) situations, we can use 1,800kW as described before. At cruise the motors run at 1,500kW each.

Weight of total system

We will now calculate the masses of the different components. From last Corner:

  • The APU style gas turbine weighs 335kg.
  • Its generator weighs 160kg (1,300kW with 8kg/kW).
  • The system battery weighs 155kg (46kWh and 0.3kW/kg specific energy).

To this we add:

  • Main gas turbine of 3,450kW shaft power weighing 600kg.
  • Main generator 420kg (3355kW with 8kg/kW).
  • Propulsor motors 2*360kg (1800kW with 5kg/kW).
  • We don’t need to count the propellers, as we assume these to have the same mass as for the normal turboprop installation.

In total, we now have a system mass of 2400kg. This shall be compared with the ATR42’s turboprop engines (PW127M) of 480kg. We shall also count the ATR’s batteries of 40kg. In total, 1,000kg for the turboprop propulsion chain. Had the reference aircraft had an APU, we should also add its weight.

But the ATR doesn’t have an APU as standard. It uses the right-hand turboprop engine in “hotel mode” (prop brake engaged) for ground power and cabin cooling.

Summary

The hybrid chain for our 50-seat turboprop weighs 2.4 times as much as the original turboprop propulsion.

In the next Corner, we will discuss what implications this has on aircraft performance. We will also discuss different optimizations that could improve the picture.

35 Comments on “Bjorn’s Corner: Electric aircraft, Part 8

  1. In your “system mass” equation, how much difference you consider for fuel? I assume the APU/Core arrangement will need significantly less fuel per trip than the “classic” layout?
    How much difference in performance you consider due to the missing “easing” of wing loads through missing fuel in the wings?

    • The hybrid chain lose 6% efficiency compared to the classical turboprop as described in previous Corners. We can gain back that loss if we can get a main larger turboshaft which is 6% more efficient than the PW127M. This is realistic, but not much more. Therefore we can’t beat the classical turboprop efficiency at first sight.

      Then we have 1,400kg more propulsion system to carry around. Therefore the empty weight goes up and we will consume more fuel per sector with the same passenger number carried. So no letup on fuel carried.

      We will cover this all more in the next Corner.

      By the way, if you take fuel out of the wing the wing bending moment increases. You want the fuel as close to the lift as possible.

      • Sorry, I didn’t get that…An ATR can carry a max. 4,500kg of fuel for its purpose of transporting people and/or goods. This, I assume, is part of the “system mass” of the classic layout. As the target of electrical propulsion system, as I understood, is getting rid of as much of this “carbon mass” as possible. Isn’t it?
        Therefore, my expectation would be, that the electrical design, even with a carbon eating core/APU arrangement, would need less than that 4.5t for the same mission? Otherwise, why should I do electrification in the first place?

        Speaking of wing bending; yes exactly this is what I mean – for some time of the flight cycle, the “fueled wing” has an efficiency advantage. How big is that?

        • It does not need less fuel. Read the previous Corners. The amount of energy we can take with us except for fuel is in the battery (which can be filled from power station power) and that is 46kWh. It’s needed for backup, and would only last for a couple of minutes of flight.

          It’s a hybrid, and like a hybrid car, once on that long trip it need to start the combustion engine like a normal aircraft. The problem for the aircraft is that normal aircraft fly straight and level and only descent once. A hybrid car is compared to a normal car which stops at every stoplight and burn all the kinetic energy while doing that. The hybrid recovers the brake energy and is therefore more efficient than a normal car, despite higher weight and lower efficiency in the combustion engine-generator-inverter-motor chain than the normal combustion engine chain.

          The aircraft has a tough task recovering descent energy due to the propeller and fan aerodynamics (the blade profile is optimized for one way function). It might come.

          All the Pink stories about electrical aircraft are from people who hasn’t even bothered to do the most basic math.

          • Yes, understood. That brings me back to the point: why?
            If I have an aircraft designed for a range “X” with “W” number of passengers needing “Y”kg engine system and “Z”kg fuel to consume. Why should I design an aircraft that also need “Z”kg of fuel to make the “X” range with “W” passengers but needs “Y+n”kg of engine system…?

          • The orthogonality of systems capabilities may ( or not ) provide advantages not available in the customary setup.

            example:
            Helo with hybrid drive train.

            Boost power added from the battery allows less peak engine power for (short time) hover. The fall back capabilities allow a single IC engine ( then backed up the electric drivetrain and not by a second IC engine).
            Less peak power allow for swapping out the less efficient shaft turbine for a much more efficient diesel engine ( diesel goes up in efficiency with lower power takeoff while the turbine is the reverse.)
            Overall you get a cruise efficient heavier than air vehicle that allows VTOL ( but not extended hover.)

          • I’m not the one saying the electric aircraft makes sense with the technology available for the next 15 years. I run this series to check how sensible it is, it seems to have major problems. Others jump on the idea because Tesla has proven it works for cars. But the situation is quite different, the normal car is very inefficient as it waste a LOT of brake and coast energy. The normal aircraft is not wasting energy and that is why a hybrid can’t compete with present and near future technology.

            What we are reveling in the series is how large is the problem and where are the culprits.

          • Bjorn, I’m disappointed to find that you aren’t going to provide some magic to make electric flight work. I apparently misunderstood your first segment when you said that battery power would be limited to local special use (which I thought was very generous) but that hybrid was more promising. I didn’t read that as more promising but still a failure.

            In any case, the series has been interesting. Thanks for writing it, and your other articles as well.

  2. Is the conversion from DC from the batteries to AC for the propulsion motors really necessary ? Its all energy loss.

    The propulsor motors seem a bit low powered ( 1.8MW). Airbus E aircraft road map is talking ‘Ten megawatts up to 20 megawatts for regional passenger aircraft.’
    http://company.airbus.com/responsibility/airbus-e-fan-the-future-of-electric-aircraft/e-aircraft-roadmap.html

    They seem to think 0.5- 2 MW will just carry a handful of business travelers. of course your framework has 2x 1.8MW
    Their test engine with 260kW continous is 5 kW per kg, that involves more than just the magnets and windings
    https://www.siemens.com/innovation/en/home/pictures-of-the-future/mobility-and-motors/electromobility-electrically-powered-flight.html

    • The inverter in the block schematic has many roles. I kept it as one to make things simple. It does not only convert battery DC to motor AC but reshapes the AC waveform to the propulsor motors to control their speed and torque. In effect, generators produce AC efficiently but with wild frequency. Power distribution is more efficient with DC. Motors need variable waveform AC. Batteries need DC. So we have conversions, more than in one place.

      But to see the main lines and main problem areas with electric aircraft we keep it simple. Otherwise you don’t see the wood for all the threes.

      Re Airbus/Siemens and motors. The 10-20 Megawatt motors are for regional turbofan aircraft in the 2030ies. Propulsion power=thrust*speed. Increase the speed and you increase the power need, as we saw with the regional jet. We backed out of that as the energy needs for backup became unrealistic (4-5 tonnes batteries or another turboshaft the size close to the main). Lets see what will come from these guys, theirs and Zunum areo projects for the 2020ies are more modest.

      More on this in the next corner.

    • The conversion from DC to AC is necessary if you want to use brushless motors. “Brushless” DC motors incorporate a DC/AC inverter…

      Brushless motors are kind of essential for low maintenance, reliable operation. On large machinery, brushes are a real problem. They wear, they’re vulnerable to spark erosion, they can break, overheat, catch fire, and they’re messy. And the commutator also wears out. And to minimise all this you have to very carefully adjust the brush angle too. If water gets in, it can be a disaster.

      And of course you still need speed control. With a pure DC motor that’s achieved either by i) crude electro-mechanical means (switching in windings / resistors) which would be like having a throttle with 4 positions, or ii) using power electronics, which is essentially close to being an inverter anyway.

      Far better to use an AC induction motor, even if this is not quite as efficient as a perfectly set up DC motor running at full power.

      • All motors for such applications are essentially BLDC designs.
        VFD energized permanent magnet synchron AC engines.

  3. There is another “interesting” option around.

    Remove the brakes and replace them with full size “BLDC” motors/generators of similar weight. ( these would degrade to simple eddy current brakes if the inverter fails.)
    Now we can recuperate landing energy into our battery storage _and_ we can also with higher efficiency
    than via impulse transfer bring our plane up to V1 speeds.

    • Good idea. Will be a tough job to get the principle certified though. But you also get e-taxi. It improves the situation marginally though, our problem is when we have rotated and before touch down. The all battery chain weighs tonnes and the hybrid is not much better.

    • Uwe, some time ago I ran the numbers on wheel motors as part of an electric taxi project. Even if you did nothing with the turbofans and just used the power generated from landing for taxi and acceleration for subsequent takeoff (which is all the power you could get out of a landing anyway) it would still be orders of magnitude the wrong thing to do.

      Batteries are so heavy that their only possible (foreseeable) use is when power cannot be had some other way, as before engines are started or all have failed, and only for some low power use such as engine start or flight instrument backup.

      In addition, the wheel motor/generators required to absorb landing energy and accelerate to takeoff would be much heavier than those needed for taxi. Jet A plus either APU or main engines is much more efficient.

  4. May I propose taking a look at a setup similar to the Lockheed L-1011 TriStar, replacing the wing engines with turboprops and using the rear-mounted S-duct engine as the APU and Core.

    This will allow for a larger battery pack which can be used in an emergency event to land the plane. At cruising mode, it can charge the battery.

  5. Hello Bjorn,
    Can we have a gain if the “core” is sized for takeoff (excluding OEI) and “APU” for cruise ?

    Best regards

    • Hi Crise,

      not really. The cruise power is 3,000kW and normal take-off 3,200kW. The ratings are to close to justify a split of these two over different turbines.

      • Ok

        And what about supercapacitors ?

        Going full electric at the wheel level can also produce some improvements for the take off roll ?

        Best regards

  6. Work is being done to create composite structures that not only take load but also serve as energy storage. I doubt this would overcome the weight issue with electric powered aircraft but would go some way to mitigate it. See;

    https://www.wired.com/2013/10/volvo-supercapacitors/

    One great reason to go to electric aircraft – if its possible – would be to free ourselves of greenhouse gas emissions. If renewable energy sources are used to make anhydrous ammonia then this substance can be used both as fuel for a fuel cell (much more efficient than even the best of todays gas turbines) and also burned in a gas turbine or APU. In either case, the bi-products are Nitrogen and Water.

    • Two potential challenges with this are the toxicity of ammonia in the event of a leak and NOx emmissions.

      • There are many more than just two challenges – a fuel cell that delivers the power required must be small enough, light enough, reliable enough and responsive enough for use on an aircraft and it doesn’t exist yet. And might never exist.

        The challenges with handling ammonia are far less than you might expect. Its not as flammable as fuel, and its already made, stored and transported in large quantities.

        NOx would be an issue too – but I’d ask you to balance that against the complete absence of CO2!

  7. There could be two other potential small advantages to this layout:
    1. solar cells could be incorporated into the wing and other upward facing surfaces of the wing, providing some power
    2. possible aerodynamic efficiency gains vs. current layout of combustion engines mounted on the wings

    • Solar cells generate minuscule amounts of power. The ATR42 in this example has a wing area of 54 sqm and maybe that much again on fuselage and tail. Total irradiance for solar is 1kw/sqm best case. PV cell efficiency is 15-20%, so you get about 16kw of power from completely covering the aircraft upper.

      For comparison, the required power from installed motors in this example is 3.6mw, 200+ times larger. The cells might not be able to carry their own weight even at noon.

      Some very slow experimental aircraft with big wings have been able to generate enough power to stay aloft, mostly because of the very low power requirements of their glider like designs.

  8. Hi Bjorn
    If that battery/inverter system sees a peak power of 550kW, that should generate a few kW of heat in the fuselage. That means some kind of cooling. Which adds weight. The motors may have to be cooled as well on the ground. More weight.
    All those equipment would be joined by really big cables, with EM shielding. More weight.
    Any equipment, ducts or cables must be attached to the fuselage by some brackets. Installation weight is about 20% of the installed component weight.
    Of course the total weight of these things will depend on where and how the core, Apu and associated generators are installed. But we could easily be looking at another few hundred kg, maybe 500kg, most of which would not be found on the classic design.
    So the final weight status is actually even worse for the electric design.

    • The plane still runs kerosene powered turbines which are even bigger heat generators.
      The airflow around the propeller and the hub should be sufficient with careful design to dissipate any excess heat in that area.
      Light planes mostly use air cooled piston engines still , so it wouldnt be complicated to allow for that

      My desktop all in one computer screen has quite a complicated heat transfer system and has now have a liquid to move the heat to the edges where the fan can dissipate through the heat exchanger. Planes should be no problem

      • That is only true for the motors which may or may not benefit from the prop wash, depending on installation. The prop wash may or may not be sufficient depending on the considered scenarios : OAT, failure cases, turn around time, taxi time…
        Even if there are only a few corner cases in which heating becomes an issue, you still need to design for those cases. Alternatively, the already relatively limited aircraft will have extra operational limitations. Which ultimately means more costs.

        Other components (batteries, inverters, high power wires) will need to be in areas where the temperature/humidity variations are not too extreme. But if they are inside the pressurised fuselage, the generated heat must be taken out. This is The major issue for the 787 and other “More Electrical” concepts, assuming we are talking about aircraft which will climb above 10000-15000ft

        Brayton cycle turbines are indeed heat generators, but that heat is part of their value-adding energy chain. So it is desired and controlled, and most of the heat is internal to the engine, which itself can be installed outside the pressure hull without a problem. However depending on where the components are installed there would have to be a heat & fire shield (is the case for APU compartments on airliners, which may be counted as installation weight.

        To summarize, on the ATR there are 2 discrete heat sources which are naturally well isolated from the rest of the aircraft and easily cooled. On the proposed hybrid, there would be multiple heat sources scattered around the airframe, with their installation constrained by other conditions, resulting in a need for careful thermal integration. Which means a/more weight to be counted in the global assessment and b/if the weight comparison was not already so bad as to basically stop the concept, then the next step would be to carefully design the overall physical architecture of the electrical system before further comparisons can be made with classic designs

        Sorry for the complex post – I just answered while I had a couple of spare minutes, and don’t have time to refine the wording

        • Thanks Airmagnac,

          this is part of the discussion in the next post. For simplicity I have left out all these added complexities, to deal with next. Your description is perfect and I can now reference it in the next Corner.

  9. Haha, so can I get a cookie for the copyrights ? Or a round of sake ? 😉

    • Excellent input – you have summarised all my concerns – in my opinion hybrid airplane is landing on the same together with like flying cars – at least for now and next 15-20 years.

  10. Am I cutting corners when stating stowing the energy for serious (capacity / range) operations is the main limitation.

    Turning around the discussion, removing all fuel, fuel tanks and related equipment from an ATR42 (5.5t ?) and replace it by batteries, how much energy/ range gives that for 5t payload? (ignoring all other constrains)

Leave a Reply

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