Airlines are looking to lower costs every single day.

I wonder if it makes economic sense (fuel costs) to operate a airliner using only one engine during cruise, while keeping the second engine at idle in case of engine failure.

Would it make sense, would it be the same, or would it be worse?

Note that this would only apply to the cruise phase. TO/GA and landing would use both engines as normal for obvious reasons.


2 Answers 2


A twin engine aircraft with one engine operating (OEI) suffers considerable drag penalties. First, there is the drag of the windmilling engine. Since the engine is likely out on the wing, it causes a significant moment that requires rudder and aileron deflection to offset.

Best range for a transport aircraft occurs at high speed -- approximated by best $M\,L/D$. We usually cruise a little faster than that because we value time. Consequently, we usually cruise at pretty high throttle and the TSFC is pretty good.

Your suggestion would make more sense if we were flying for best endurance -- where we fly slower, require less throttle, and would incur a greater penalty for low TSFC from operating at low part-power.

That said, there is this proposal for a two-engine aircraft (with different sized engines) that would turn on and off depending on the phase of flight.

I believe the four-engined P3 Orion would shut down one engine during cruise.

Also, there are three-engined helicopters that shut down one engine during cruise to get longer range.

Edit below:

A jet engine has a characteristic that we call a 'thrust hook'. It is how the TSFC (y-axis) varies with thrust (x-axis) as throttle is varied at a single flight condition (Mach, altitude).

In this case, the thrust axis has been normalized by the sea level static (SLS) rated thrust. The flight condition is typical high speed cruise.

At this flight condition, maximum throttle only delivers about 27% of SLS rated thrust. This is due to thrust lapse.

The shape of the thrust hooks is what is important. They usually have a relatively flat region near high throttle settings. This is observed from about alpha=0.17 to 0.25.

Then, there is a fuel flow penalty at very high throttle -- from 0.25 to 0.27 (some engines don't really have much here).

On the other hand, for fighter engines, this would be where afterburners kick in -- the thrust jump would be much larger, but the TSFC jump is crazy too!

More interesting to us is the gradual (but steep) TSFC penalty at low part power (alpha less than 0.17) -- getting really bad at flight idle.

Lets say an aircraft was at a flight condition where flying with both engines operating required alpha=0.09 (TSFC=0.72). Instead it could operate on a single engine at alpha=0.18 which would give TSFC=0.65. Which is about a 10% improvement.

So, you're looking for a situation where the 10% fuel flow improvement is larger than whatever drag penalty you would have.

Many loitering missions involve flying in circles. For example, a long endurance drone that wants to continuously monitor a certain region of the world.

enter image description here

  • 1
    $\begingroup$ If during part of a flight, a plane wasn't going to be traversing any useful distance but just circling to remain airborne, e.g. because one was near the destination but couldn't land immediately, how would endurance using one engine at relatively open throttle compare with endurance using both engines throttled back? If the plane is supposed to fly in a circle, one might not need much rudder to compensate for asymmetric thrust. $\endgroup$
    – supercat
    Apr 1 at 18:04
  • 4
    $\begingroup$ "If the plane is supposed to fly in a circle, one might not need much rudder to compensate for asymmetric thrust." That's not now a coordinated turn works. With asymmetric thrust the need for rudder doesn't go away because you have an angle of bank. $\endgroup$ Apr 1 at 18:55
  • $\begingroup$ @MichaelHall coordinated turn just means there is no slip. If you have asymmetric thrust and use the ailerons to eliminate the forward slip induced by it, then you're doing a turn and are in coordinated flight, so it's a bona fide coordinated turn. Whether it's as efficient as a conventional rudder-induced coordinated turn is another matter, but I suppose it would indeed be relatively more efficient than rudder-corrected forward flight with asymmetric thrust. $\endgroup$ Apr 1 at 21:56
  • 1
    $\begingroup$ @leftaroundabout, OK you got me... Yes, I suppose you could counter the asymmetrical thrust with adverse yaw from the ailerons, but only in one direction. The point is that you still have to counter it with a draggy solution. $\endgroup$ Apr 1 at 23:10
  • $\begingroup$ @leftaroundabout The implication of your proposal is that the adverse yaw produced by the ailerons is about the same magnitude as the yaw produced by having only one engine producing thrust, but I believe that, except perhaps for aircraft with engines near the centerline, asymmetric thrust calls for significantly more rudder than does adverse yaw. $\endgroup$
    – sdenham
    Apr 2 at 12:46

Using actual numbers, a modern A350 at fairly heavy weight of 250 tons will have a specific range (distance travelled over burnt fuel) of approximately 77NM/1000kg at optimum cruise altitude of 37000ft. If one engine is shutdown, the performance is not adequate to maintain that altitude and aircraft is forced to descent to an altitude around 21000ft where density of air and thus drag is higher. The specific range drops down to around 60NM/1000kg so it is not economical even fuel consumption wise. On top of that the ground speed is roughly halved which doubles the travel time between two destinations, basically halving the amount the number of profitable tickets an airline would be able to sell.

And that is assuming we ignore safety, risking loss of total thrust in case of failure of the remaining engine - among several other less obvious risks such as secondary system failures and contrallability of aircraft.

  • $\begingroup$ I have accepted your answer since it betters addresses my initial question. Thanks for that info. $\endgroup$
    – Gabe
    Apr 3 at 13:45

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .