Flying an Airliner After an Engine Failure on Takeoff
Here’s the “yes” part of that: every multi-engine airliner in service today is designed and certified to continue a takeoff after an engine failure and fly on one engine, provided that the performance limitations are not exceeded and the correct single engine procedures are followed exactly.
Which brings us to the “no” part: if performance and control limitations are exceeded, or incorrect remedial procedures applied, chances of a successful single-engine takeoff and climb are slim at best.
Here’s a close look at the variables. First, the performance limits. Can an airliner execute a normal passenger flight with just one engine? From brake release? Of course not. What it can do is continue a takeoff if an engine fails with one inflexible limit: you must have achieved the correct minimum speed prior to the engine failure in order to successfully continue the take-off with only the remaining engine(s).
That speed is called Critical Engine Failure Speed (CEFS). To be exact, CEFS is the minimum speed you must have attained with all engines in order to successfully accelerate to takeoff speed after an engine failure, and then within the runway remaining, lift off and and cross the departure end of the runway at an height of at least 35 feet.
Stopping with a failed engine is a whole different discussion, to be addressed in a future blog. For now, consider the engine failure and the takeoff being continued. If we have met or exceeded the CEFS, we will continue the takeoff which is critical to down-line obstacle clearance.
The go-no go speed is called “V-1,” which is simply “Velocity 1,” the decision speed on takeoff roll: if you’ve attained V-1, you’re able to fly. If you’re at V-1, unless you’ve started braking, you’re committed to flight because you may not be able to stop within the remaining runway.
For me, life becomes easier at V-1: we can, and will, fly. That’s what the jet (and I) was intended to do–the thought of bringing tons of hurtling metal and fuel to a stop in the remaining runway is not appealing to me. In fact, I need less aircraft systems to fly than I do to stop, including no blown tires, operative anti-skid and spoilers. In that split second abort decision, how can I be sure I haven’t lost an electrical system that would inactivate the anti-skid, or a hydraulic system that could affect the spoilers, or a blown tire that would take out 25% of my braking–and maybe cause a wheel well fire?
The answer is, I can’t be sure, but I can fly with every one of those components inoperative, and to a pilot, flying a sick jet is preferable to wrestling a sick multi-ton high speed tricycle to a stop. So we fly, if we can do that safely.
My discussion from here pertains to the Boeing 737-890 aircraft I fly, but I would add that all airliners are certified to this same performance standard. Procedures vary, but the single engine performance standards are similar.
So in the event of an engine failure beyond CEFS, at rotate speed we will rotate normally and begin our obstacle clearance climb. This is where crew action is critical.
The first indication of an engine failure in the cockpit will typically be a yawing motion due to the imbalance of thrust between engines. Whether that occurs on the runway or, more likely, in the air, the response is the same: add as much rudder as is required to slew the nose back to normal flight. That’s critical for two reasons. First, the runway clear zone (the area over which you must fly) extends forward from the runway centerline. If you curve laterally away from the centerline, you lose the obstacle clearance protection of the runway clear zone.
Second, the correct amount of rudder eliminates the need for aileron use, which comes at a price: if enough aileron is input, wing spoilers will deploy, inducing drag. This is crucial because drag limits the climb capability which is a defined gradient required to attain obstacle clearance altitude.
So here’s the “yes” part again: if the aircraft weight is within prescribed limits, if the correct speed is maintained and the specified climb gradient is flown, and the lateral ground track of protected airspace is tracked, then yes, the takeoff and climb-out is certified to be successful.
Do we, in the event of an engine failure, add power on the remaining engine? Generally, no. Why not? First, because the calculated takeoff power setting is designed to be sufficient to allow a single engine takeoff and climb after an engine failure. Yes, more thrust is available and if you need it, you use it. Our CFM-56 engines are electronically controlled to protect against over-boost damage, but here’s a pilot thought: if the climb is proceeding correctly, why introduce more adverse yaw, and why strain the remaining engine?
Now, crew response. The person noticing the engine failure is normally (but not always) the pilot flying who feels and counters the yaw. That person, or often both pilots, call out what they see: “Engine failure, number __,” or “engine fire, number ____.”
Then, this and only this: maintain climb speed (and thereby climb gradient) and ground track. Let’s backtrack a bit. Before each takeoff, on taxi out I verbally review three altitudes with my First Officer: the field elevation, the engine out altitude, and the minimum safe altitude for that airport. And that’s our focus in the event of an engine failure: climb at the correct speed on the clear zone path to the single engine climb altitude.
A wise old CRM (Cockpit Resource Management) instructor used to tell all the pilots at my airline as we cycled through for our annual recurrent flight training and evaluations the same very shrewd piece of advice for this and any other flying emergency. He was a crusty, retired Air Force fighter jock who’d hammer this home: “Whatever happens, before you react, you take a deep breath and say to yourself, can you believe this sonofabitch is still flying?”
Even after that, we don’t react–we respond appropriately. That is, between the two of us, we agree on what we have, and that can only be three things: engine failure, engine fire/catastrophic damage, and engine overheat. Identifying the problem and the engine is important, because the corrective procedures differ.
So in the minute or so that it takes to climb to our pre-briefed engine out altitude, we’re both analyzing exactly what happened, and which checklist we will bring out to accomplish step be step.
What if the First Officer, rather than me, is flying when the failure occurs? From my point of view, and I’m coming up on 24 years as captain, I say so much the better: all of our F/Os know exactly what to do and moreover, they’re flying, they have the feel of the jet and the corrections in–why throw a control change into the mix and try to handle it cold?
As an added bonus, as the pilot monitoring the pilot flying, I’m downloaded of the physical stick and rudder challenges which are significant single engine. I can concentrate on analysis, procedures, radio calls and clearances because “Bubba,” as they referred to F/Os in flight engineer school, knows what he’s doing.
So here we go: what do we have? Simple flameout? Do we have RPM? If it’s not turning, there’s damage. Temperature range? Fire? Oil pressure? Only when we both concur will I, being the pilot not hands-on flying, pull out the checklist and read it step by step as I accomplish each with the F/Os concurrence at each step.
Here’s where discipline and crew coordination is key: NOBODY is going to start flipping switches on their own and whatever is done will be done only as I read the procedure. The best way to mangle any emergency is for anyone to go solo and start operating off script.
In every engine failure scenario, there comes a point in the corrective procedure where a throttle must be closed and a fuel lever shut off, possibly a fire switch pulled. The throttle of course reduces the thrust, the fuel lever cuts off the fuel supply to the engine (it’s going to flame out) and the fire switch shuts off fuel at the tank and the wing spar (in case the engine fuel shutoff valve is damaged by fire or explosion) as well as hydraulic fluid, pneumatic bleed and electrical power.
These actions are drastic and with only one engine operating, they must never be done independently, unilaterally or without a double-check and concurrence. They are also most advisedly done only after level at the single engine altitude with obstacle clearance assured.
Here’s how that plays out in the cockpit, verbally and physically:
Me, reading the critical steps: Fuel Lever, affected engine (confirm)
[pause] I touch the correct fuel lever, F/O concurs; F/O guards the good engine fuel lever with his hand.
Me: Cutoff. [I perform the action] It is cutoff.
Then we go to the next step in the checklist, me reading, pausing for concurrence and confirmation. Bubba is focused on aircraft control, altitude and airspeed, validating each checklist step I read before and as it’s taken. I’m focused on the procedures, plus backing up Bubba’s flying.
If I were flying when the failure occurred, same process, just reversed roles. Each and every step in each appropriate checklist will be accomplished with crew coordination till we are ready to return and land safely.
The easiest engine failure to handle is a simple failure or “flameout.” You may try a restart under some circumstances, or you might not take the time and instead, just get the jet ready to land. The most difficult failure is the fire and severe damage situation, but it’s handled the same regardless: carefully, step by step with collaboration and concurrence.
Never singlehandedly or without concurrence. Because the deadly reality of two engine aircraft is this: if you apply any of the required procedures to the wrong engine, the only engine sustaining your flight, the results will be disastrous.
I’ve had to fly four actual single engine landings in MD-80 jets for various reasons, none so far in the rugged, reliable 737. We practice engine fires and failures every nine months in our recurrent simulator training, handling multiple scenarios each four hour session. The key to a successful single engine incident is procedural integrity, crew integration and communication, controlled pacing, and standard operating procedures followed to the letter.
In the end, a successful engine failure landing comes down to coordination, discipline, adherence to standard procedures and as my old fighter pilot buddy used to say, taking that second or two to collect your wits and say, “Can you believe this sonofabitch is still flying?”
For those who don’t adhere to all of the above, it won’t be flying for long.