4-engine single-aisle airplanes could make comeback, says technology VP

Dec. 15, 2015, © Leeham Co: LNC’s Bjorn Fehrm started a firestorm of discussion last Friday with his Corner about twins-vs-quads column. His focus was on the Very

Could a four-engine, single-aisle airliner make a comeback? It’s something that might be possible. Photo via Google images.

Large Aircraft sector. Overlooked in all of the discussion was a piece of information LNC wrote April 6 from an interview with Alan Epstein, VP of technology and environment of Pratt & Whitney, in which he mused that quads could make a comeback—on smaller airplanes.

The original article was behind our Paywall, but the Summary with this reference was in the freewall portion. We’ve now opened the article to full freewall and it may be found here.

“Getting more efficiency from engines means the use of more advance materials, evolving and new technology but larger fans have traditionally been a key to gains. The 81 inch fan on PW’s Geared Turbo Fan on the A320neo is pushing the engine size on the single-aisle airplane in traditional configuration, Epstein says,” we wrote at the time.

“’It looks like you can get one more iteration at…taxi weight limits and longer landing gear. You could go for four engines,’ Epstein says—an idea that is counter-intuitive to trends since the introduction of the Boeing 757 for high-performance, 3,500nm-4,000nm aircraft, intended to replace the three-engine, mid-range Boeing 727-200A,” LNC wrote.

Maintenance costs of large engines on aircraft the size of the Boeing 777-300ER and the forthcoming 777-8/9 are huge and work against the secondary market of these aircraft. The smaller Airbus A350 and A330 have lower maintenance costs but the MRO costs of the 787s, and particularly their composite, intricately-designed nacelles, are more expensive than early customers figured into their cost equations, according to Market Intelligence.

“The idea of a four-engine single-aisle aircraft raises immediate questions about fuel burn and maintenance costs of four vs two. While our own analysis, with respect to the Airbus A380, concludes the four-engine maintenance cost perception is overblown, Epstein points out that if an airplane needs 80,000 lbs of thrust (for example), this may be achieved with two 40k or four 20k engines. Maintenance costs for four engines can be achieved for the price of two, he says,” LNC wrote in April.

Based on LNC’s discussions with engine experts, there is no likelihood of a four-engine single-aisle airplane in the next 10-15 years. But looking beyond 2030 could be another matter. The choice between conventional jet engines and the prospect of Open Rotors is not dead, either, despite an apparent cooling of the latter idea in recent times.

While many readers overlooked the fine nuances of Fehrm’s analysis last week—notably the per-seat application of costs and the capital costs of used airplanes—the future is likely to bring some designs that defy today’s conventional wisdom.


49 Comments on “4-engine single-aisle airplanes could make comeback, says technology VP

  1. Hello Scott
    Let’s think Split propulsion and E-fan ?
    A disconnection between engine count and thrusters

    Interesting the nacelle facts !

    • It can be an outboard shaft driven shrouded fan driven by the inboard Engine. A RR Bristol F-35 Lift-Fan type of drive from a high Power angle gearbox over to the next Engine with a fairing, making a double decker between the fans mounted on the same wing. Hence you get 4 fans but only 2 Engine hot sections on the Aircraft.

      • You have to keep it simple. Well for military purpose you might just go for risky designs, but in commercial aviation safety is priority.
        The other problem is that every gear loses power, especially when the direction changes, plus you have rather high loads, not a good idea if you want to build light. Next you might run into vibration problems,…

        • It will not be easy, but gears today very have high efficiency requiring small oil coolers for high Power like transferring
          20 000-30 000 shp and the direction of rotation stays the same as in any gas turbine.
          Just check the PWA PW1100G engine oil cooler. The reverse is done on the Cessna Caravan.
          Soloy is offering a dual-engine conversion of the 208B, named Pathfinder 21. This version is powered by a 991kW (1329shp) Pratt & Whitney Canada/Soloy Dual Pac powerplant, consisting of two PT6D-114A engines driving a single propeller.

  2. With two smaller optimised GTF or Leap types plus two E-fans, the latter reversible for energy recovery during descent/approach, plus a pile-à-combustible E-plant and latest battery technology, the CO2 signature of tomorrow’s feeder aircraft could be alleviated vs A32X or 73X, if COP21 shall set the guidelines to aircraft design ? Choose the H21QR cabin configuration to put the cherry-on-the-cake in terms of pax-appeal and 24h fleet yield ?

  3. The geared fan will certainly change the whole game. I would love to have a chat with the R&D guys at Pratt, RR and GE to see what they are working on long term. I would guess something with a bypass ratio of maybe 20:1. Tiny core, big fan, all carbon and ceramics,…
    The real limit for such engines would be the outer diameter, and that is probably why Alan Epstein talks about 4 engines even for single aisles planes. So maybe 6 for twin aisles and 8 for double deckers?

  4. I think not all materials, dynamics, containment requirements scale up endlessly. With a given state of technology and optimal dynamics you end up somewhere. When new technology is introduced things can change again.


    There seems to be a kind a natural gap between the biggest twins and the A380. Theoretically 3 x 70-80klbs would be best to cover 400-500 seats, bridging ETOPS and using ” standard engines.

    The question I asked myself long ago (Ecoliner) was if 2 x 110-120klbs “standard’ engines + a small one for take-off (=160 klbs with one engine out) isn’t more practical, because it is easier to install / service something small somewhere in the cg line. Possibly combined with the APU like Boeing considered in the past in their “twin and a half” engined 777 development studies.


    • hi keesje,
      there is some serious trouble with three engine concepts (DC-10/727). Besides the installation of lots of very long cables and fuel pipes, which increases weight ist is especially the size of modern high-bypass engines that are really difficult to fit to a classic tube body.
      The E-thrust concept looks really nice, but to me looks like very far away. Second half of this century maybe?
      2, 5 engines? Too little thrust for too much effort and cost.

      • “especially the size of modern high-bypass engines that are really difficult to fit to a classic tube body.”

        Gundolf it’s also an concept counter-intuitive to trends / conventional wisdom. A 40klbs only weighs 8-9klbs and doesn’t have to very efficient / large. Its only operated 10 minutes a flight, it’s inlets would be retractable.


        • Maybe an electric engine in the tail for take off.
          Seems like the logical place, since you don’t want a non operating engine out in the slipstream causing drag if they aren’t producing thrust.

          Or two retractable engines coming out of the top of the body near the tail.

    • There has to be a better way of mounting the third engine than tail installation. But it would require a major rethink of the whole airframe… Blended wing/fuselage with above-mounted pods?

      With a conventional tube/wings, there is the possiblity of an asymetrical design, with 2 engines on one side, and one on the other. Asymetrical has been done before:


  5. Hmmm…

    I spent some time at the MIT Gas Turbine Lab back in the early 1970’s and Epstein was still a graduate student then. He worked on the MIT micro turbine before moving to P&W.

    IMHO, 4 engines are the solution if the fan size is so large that it forces the landing gear to be too long, with snowball effects on the rest of the aircraft structure (mounting huge engines on the rear fuselage is not too attractive from many points of view, structure, vibration , cabin noise …). So …

    Why not use 4 smaller engines with the same high BPR but smaller fans? The wing will be lighter due to load relief from the outboard engines. Added complications are fewer and manageable. May even help with OEI scenarios. Don’t need huge single engine excess thrust during OEI TO and don’t need to descend too much for OEI cruise. The only issue is inventory costs of spare engines and of course maintenance costs. If the maintenance costs increase exponentially with engine thrust (Bjorn can pipe in here), then 4 instead of 2 makes sense.

    I wonder if the new clean sheet from Boeing for the middle of the market (MOM) might very well be a 4-engine single aisle, with very high BPR GTFs! A twin-aisle just does not make sense. Time will tell!

    • Trouble is there is a real airline who did run comparable generation widebodies 777-200, 340-300, side by side in their fleet, Air France.
      Guess which one they reduced their numbers flying and which one they increased their fleet with.

      A340 , operated 30 currently flying 13 ( x2 writeoff), none of them the bigger 500/600. While they operated 68 777 with 67 flying.
      Yes the twins won over the quads, and this from an airline whose the home of Airbus.

      • Air France
        flies 777-200ER and 777-300ER
        The 777 fleet is 6..7 years younger than their A340 fleet.
        not really comparable, are they?

        try again please.

        • Thats very glib. The 777 fleet is younger because the latest plane just joined the fleet 3 months ago while they havent had an additional 340 since 2001.
          Its a real world example of an airline flying both 340-300 and 777-200, yet they have preferred the twin. This was the specific versions that Bjorn said were comparable generation and age.

          “There can be made more modern comparisons. The first is the Boeing 777-200ER versus Airbus A340-300. These came into the market within four years of each other. ”

          AFs 777-200er were introduced from 99-02, while the 340-3oo were introduced from 93-2001. Those few years difference are nothing.
          If Air France cant love its Airbus quads , who can ?

          • Not that I want to defend Airbus, but as I also recall Bjorn said they elected to go with CFM engines that were not comparable on the A340-200 and 300?

            Not that familiar with A340.

            So while topical, poor choice for efficiency maybe due to expediency rather than best long term results?

            Interesting topic, intuitively it seems two are better than 4 economically as well as fewer pylons, fuel lines etc but am listening now rather than the previous closed mind.

          • And just to add some interesting aspect, the CFM weights around 5,000 lbs and the Trent 500 10,000 (round numbers)

            So a lot more thrust, a lot more weight and the trade off is (leave it to Bjorn of course!)

          • all data i found indicate that the
            CFM56C weighs in at 3990kg/8790 lb (dry)
            Trent500 (A340NG! ) dryweight : 4835 kg / 10660 lb
            Trent700 ( A330) dryweight : 4,785 kg / 10,549 lb
            ? banana vs orange data ?

            Thus available data does not fit the narrative.

          • The CFM56 /V2500 were the only engines available so the 340-200/300 can only be compared using the engine it was fitted with.
            Digging a bit deeper the 340-500/600 are ‘not comparable’ since they are ‘over stretched’ and the 200/300 models have a short range engine for a long haul job.
            What was left ? A slightly stretched version known as 340-400 with RR535 engines from the 757? Or as the 600 had poor sales build a 400/500 version with a smaller scale Trent 500.
            Surely Airbus would have had all the data and market knowledge to build a plane that was reasonably competitive- but didnt.

            AS much as I appreciate the analysis, and it pleases me to see something that is right sized and have a safety advantage, the right quad wasnt available to counter the 777-200ER.

          • Well all compaaies makes mistakes.

            Also keep in mind Airbus did not think the real extended ETOPs would occur, so they put their long distance eggs into what should have been a good basket.

            Rules changed but they were stuck with the engines on a plane that were sub optimal. RR probably did not make any money on the 500 (probably lost money due to low numbers).

    • One of the reasons why big twins have been more efficient is that good compressor efficiency is more easily achieved on a large engine than on a small one for high pressure ratios.

      On a small engine, targeting a high pressure ratio results in a lot of loses to boundary layer friction at the HP end. By the time the air has got to that end it’s been reduced in volume so much that there’s not much of a gap between the casing and disks, and it is hard to force air through it.

      On a big engine with the same pressure ratio those boundary layers still occur. However the boundary layer is about the same-ish thickness as in the small engine but represents a lower portion of the air volume in the large engine. Thus a larger proportion of the air in the large engine’s compressor is laminar flow outside of the boundary layers.

      Consequently the compressor loses in a large engine are fundamentally reduced, assuming that all other design parameters have been optimised and ignoring things like GTF being available on smaller engines. This translates into quite a large energy saving for the 2 large engines.

      So for 4 small engines operating at the same bypass ratio as 2 large ones the cores of the engines have to be smaller. This means that they will not be able to operate at the same pressure ratio as the large engines and have the same efficiency. Tricks like GTF are required to compete on efficiency.

      Of course what Bjorn and you are pointing out is that the true efficiency metric of an aircraft – operating cost per passenger per mile per annum – takes into account many, many more things than fuel efficiency.

      • Does that apply to a 3 shaft engine like the big RR?

        I looked at the pressure rise for the new CFM Leap1a, its around 40 ( I could be wrong?) but thats just under the 42 of the GE90 type of big fan.
        Thats a jump for the CFM56 which was around 30?
        Has that problem been solved by CFM ?

        Of course you are right about smaller cores, but of course for a twin the total takeoff thrust of one engine has to be able to continue the takeoff, while for a quad they have to be able to continue takeoff on 3 engines.
        From this the twins core is oversized (and fan to go with it).

        • Didn’t the Long Range engines always have higher pressure rise than the High Cycle engines?

        • It still applies to three shaft engines. It kinda sets a fundamental limit on how efficient an engine can be no matter what you do. As with all combustion engines, the higher the pressure ratio the more efficient it can be (assuming that you can engineer everything from to exploit it). So everyone strives for the highest pressure ratio they can achieve.

          Three shafts simply means you have yet another option for tweaking, in this case the IP compressor speed. This is closer to the ideal, which is where each compressor stage would be spinning at its own optimum speed.

          Obviously that ideal cannot be achieved. RR’s three spool engines are as close as we get. They tend to have high compression ratios without lots of fiddly and complicated variable stator blades, and don’t need quite so many stages. Complicated bearings, simplified everything else.

          GTF is interesting because you can think of it as being like a three shaft engine which goes without the (mostly useless) LP compressor. Still way better than a conventional 2 shaft engine.

          The Trent 1000 and XWB have pressure ratios > 50 at “top of the climb’, which is quite an achievement. Of course that means that everything runs veeeerrry hot, which needs clever cooling and expensive materials to deal with the heat and give a reasonable service life. Interestingly the Trent 900 (39:1) does worse than the GP7000 (43:1). You can see why Emirates wants an A380neo with Trent XWB derived engines.

          As for the CFM Leap, I suspect that for this size engine (still quite big really!) the achievable pressure ratio is limited more by other practicalities, and less by boundary layer effects. Making a small three shaft engine sounds difficult. Making a two shaft engine longer (more compressor stages) is impractcal. Improving the aerodynamics in a two shaft engine is what they’ll have done, no doubt along with a host of other improvements all round.

          You’re right about engine loss on take-off. It’s another design optimisation route that a twin cannot exploit.

          • Checking the data from CFM about their 5C2, 5C3 and 5C4 engines for the A340, gives a pressure ratio at ‘a max climb’ of 37 to 38.

            Obviously an improvement over their short range cousins, and not so much a disadvantage to the 40 of the big fans

  6. Get realistic. Air transport has a new job on its hands: reducing global carbon emissions instead of increasing them. We need over 75% reduction by 2050 and much in the next 20 years. Current 20 year traffic growth forecasts overwhelm technology available in the latest aircraft — emissions could double in that period. That’s not going to be allowed by governments or the public. There is much that the industry can do in operations, infrastructure, route structures, derivatives of current models and new aircraft — not to mention commercialization of biofuels

    The next all-new transport should be in a class which would save the most fuel and carbon emissions in global airline fleets during its production lifetime — say before 2045. That is not going to be a 250 passenger, intercontinental 5000 nm MOM. It should be a large capacity Domestic Short Haul (DSH) aimed at domestic markets including N.America, Europe, and the developing markets of China and India. Those markets burn the most fuel today and will burn far more in the future than intercontinental.

    The industry should be considering a DSH for 2024-2025 service with say 275 passengers in comfortable all-economy with growth beyond. Range 2000-2400 nm, cruise speed .72-.74 mach, twin aisle, 7-8 abreast, minimal commercial cargo, almost elliptical fuselage cross section with increased lift. It will require new twin engines of around 22,000 thrust and 16-20 bypass ratio — mounted below wing (it will fit), or above wing and aft per Lockheed BWB wind tunnel tests or on aft pylons. The wing would be composite with folding tips: 118′ span on the ground an say 134′ in flight. Some laminar flow should be available. Risk will be acceptable, not high.

    I would expect 40-50% fuel and carbon emission savings per seat mile in initial service vs the smaller, older, mostly metal, 3600 nm, .8 mach A320 neos and 737-8 maxes. It will be in class by itself. It will be reengined and maybe rewinged during its production lifetime

    Expect both Boeing and Airbus to build a DSH, one closely following the other — each likely to offer a choice of engines. There could be more advanced concepts for 2030-2032 service in the single aisle 150-200 passenger size for domestic markets — emerging from Europe’s Clean Sky program or the MIT/NASA rear mounted BLI engines between Pi vertical tails.

    For intercontinental fuel savings consider derivativess of the A380 and 777-9 with say 800 all economy seating and 600 respectively, each rewinged and reengined for 4000 nm range at .78-.8 mach — both flying the Atlantic to the USA among other high density routes.

  7. There is no question that it will be necessary to use a lot more carbon to reduce the carbon footprint of flying. 😉
    There is also little question about the general shape of planes.
    But the question of the engines and which kind of fuel will be at the heart of R&D for many years to come.
    Mid term there is no alternative to efficient kerosene burners. Long term that will probably have to change completely. But first I think less weight restricted consumers (cars, trucks, ships,…) will have to move away from burning oil.
    What I expect the next race in development of mid-size turbines will be for PW to make their core as efficient as that of the Leap, and CFM will come up with their own gear for the fan. RR is somewhere in the middle between the two, but maybe in a couple of years we will have all three in the arena with comparable solutions. That should give us the next 15% savings in fuel consumption. Hopefully.
    The next task will be to try making those engines larger. I guess we will most see GE and RR fighting in that arena, maybe PW too. But if that doesn’t work, maybe we will see 4 engines on a 777-3V and the A350-2V.

  8. This reminds me of the old saw about why does the BAe-146 have 4 engines?

    Answer: Because they couldn’t fit six.

    Any discussion of four engines in the future is entirely academic. The Airbus ad for “four engines for the long haul” is now a punchline.

  9. Two engines, single pylon on each side, treated as a twin for certification, performance and control purposes would provide a platform to increase thrust as required. The engines do not need to be inter-connected, have the same thrust or the same fan size. Let the FADEC figure all that out. The nacelle , fan ring, fan reverser, exhaust nozzle and reverser can each be oval-shaped as one unit. Use the increased gear height to grow the engine thrust, eventually going to a larger composite wing and you have a 757 replacement that can be scaled to handle a larger diameter fuselage.

        • I can think of wanting linke3d throttle but each capable of separate control.

          Back to the days of the C123 and its jet pods.

          One engine quit and you could always fire up one or both pods.

          I suspect both A and B are looking at all of it really hard.

  10. The return of the sardine’s box? Scrap the comfort and let’s save money!!!!

    • Point taken; however, as evidenced by endless debate on this and other sites, interior configuration and comfort level is an airline choice. Leveraging what is already available with some new software and certification efforts should reduce acquisition cost and allow airlines a choice on the comfort versus seat-mile spectrum.

  11. It’s true that due to the trend of ever larger engines, which is caused by the trend of higher and higher bypass ratios, OEM’s will find it increasingly more difficult to ensure that for next generation twins, there’ll be a sufficient level of ground clearance for engines mounted in the conventional configuration underneath the wings

    The advantage of higher bypass ratios for turbofan means that the core of the engines is made smaller in order to aid thermal efficiency, while the fan that creates the forward thrust, can be made larger.

    However, ever higher bypass ratios doesn’t automatically mean that quads will win out in the end. One should note that the fan itself is wasting some power in accelerating the air, and this is manifested as swirl. By locating a second fan behind the first, rotating in the opposite direction at the same, or at a slightly higher speed, most of that swirl energy is recovered. This means that if the diameter of the fans is the same, and with all things being equal, a contra-rotating fan will always have a higher bypass ratio than a single fan- by a considerable margin.

    IMJ, therefore, we’ll see both small and large next generation, twin engine tube-and wing airliners – powered by geared contra-rotating turbofans – long before any quad powered configuration would be re-introduced. In fact, the industry IMO will only move away from “twins” when hybrid electric propulsion might have reached a sufficient level of technological readiness for a manufacturer to go ahead and launch a single aisle airliner powered by said propulsion system.

    • Wow, now that really looks like a great idea regarding the efficiency. It reminds me of the TU-95 “Bear” with its contra-rotating turboprops. The gearbox would certainly be more complex than the of the PW1000, adding cost and weight and they will also be louder. To much to kill the concept? I wonder if something like this was ever attemted, designed, built or even tested.

        • But that is still propellers, not jets. Maybe PW have looked into this before they started the GTF development?
          Counter rotating propellers are too noisy to by usable for non-military purpose. The question is how much the noise can it be reduced by complex fanblade shapes, uneven numbers of blades, variations in the dimensions resp. a difference in the speed of the two fans.
          The smaller nacelle and fans might even make up for the extra weight of the gears and second fan. Maybe it would not even be so much more expensive…
          Still – such a R&D project would be really expensive and has a chance probably only when fuel prices are on the rise again and/or a engine maker is quite desparate for new business. Just as the GTF came into this world during the last oil price stampede and PW was in a really difficult position in commercial aviation.

  12. Let’s say a typical feeder needs a TO power setting of 33 klbf. This is the one-engine-out requirement. The engine is optimised for one-engine-out cruise at 83 % or 27.4 klbf. When both engines are working correctly, the cruise setting is 27.4/2 = 13.7 klbf per each engine or 42 % of the TO setting. Consequently, the cruise performance is grossly sub-optimal.

    Now we rework the SAME feeder with quads. Assume the weight/drag situation is identical. We have the same safety requirement : we need 33 klbf with one engine out. We deliver this TO thrust with three engines. The TO setting per engine is 33/3 = 11 klbf. We optimise the engine for cruise at 83 % or 9.1 klbf. When all four engines work correctly, we can set the cruise thrust at 9.1 x 3 / 4 = 6.85 or 62 %. Consequently the cruise performance is sub-optimal although to a lesser extent.

    Now then we know the thrust situation and we consider the engine ‘weight per thrust’ graph, let it be a straight line w = aT + b : two engines @ 33 klbf weigh 2(a33+b) kg, whilst four engines @ 11 klbf weigh 4(a11+b) kg. The graph will tell us W1 is substantially heavier than W2. We integrate this factor into the equation and reiterate. We do this three times. We will see that the TO setting requirement for the quad may be adjusted down, wherefore the cruise perfo for the quad is readjusted closer to the optimum. The fuel consumption of the quad is therefore comparatively better vs the twin. Now then we need less fuel for the trip, so we load less fuel, so we burn less fuel to carry fuel … we integrate this aspect and we reiterate three times … etc

    The quad is the better solution !

    Where did I make a mistake ?

  13. Mind you, there are other graphs of relevance to be acconted for such as :

    engine maintenance cost EMC = F(engine hours, engine cycles) where F is a function of the power setting ; and

    engine price EP = F'(klbf), where (ditto)

    My appraisal of the situation is invariably that 66 klbf installed thrust cost more to purchase and maintain than 44 klbf. Again, I may be wrong ?

  14. So the Kusnezow NK-93 has already flown in 2008:
    Bypass ratio: 16,6:1
    Thrust: 180 kN
    Fan diameter 2,90 m
    Dry weight: 3630 kg
    So it is larger than a turbofan of similar thrust would be, at a similar weight. The bypass ratio looks like one might indeed head for something like 20:1 when the latest technology and materials are applied. It looks to me like the fan is made from steel, CFRP should safe a lot of weight here, plus they could be wider and so the diameter could be smaller, which saves more weight.
    In direct comparison: The strongest PW GTF is the 1135G with 160 kN, 2,1 m fan diameter, 12,5:1 bypass ratio, dry weight 2860 kg.
    So the double-propfan engine would be too large for a single aisles plane with only two engines. But for a A350-1000 or 777X8 with 4 such engines they would fit exactly.

  15. Might be a good time to bring back the TriStar or DC-10.

    One less engine (maintenance) than a four burner but big ass fan (efficiency).

    A happy medium!!

  16. I’ve been wracking my brain trying to think of something suitably wacky to put forward, but then you require realise that it’s all been done before. Check out the Skyfleet S570.

Leave a Reply

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