Nobody really knows how often something like an powerplant failure in a GA single happens. If the pilot manages to land without any significant damage to the aircraft, nothing ends up in the NTSB accident database because, by definition (see 49 CFR 830.2), there was no accident.
I’ve experienced a few of those in my career. One was a clogged fuel filter in a Pitts. Believe me, anything out of the norm in a Pitts really gets your attention. I love the airplane, but it sports a glide ratio akin to that of a brick. Only worse. Another memorable engine failure was caused by a broken cylinder in a Cutlass. The ensuing vibration and smoke convinced me to shut down the engine, and I glided the 10 or so miles to Corona with my commercial student and made a uneventful, real-world power-off 180° approach to runway 25. I even remembered to put the landing gear down.
Incidents like these illustrate one of the core skills of a talented pilot: energy management. This is true even if the powerplant is operating normally. From aerobatics to IFR flying, those who have mastery over the machine have an innate feel for the aircraft’s energy reserve as well as the energy required to achieve whatever goal is currently before them and can adjust accordingly. If this sounds a lot like flying a glider, stay tuned.
Over the years, I’ve come to believe that our training for engine failures is inadequate. For one thing, they seldom happen the way we train for them. In a multi-engine aircraft, a dual-engine failure is — to the best of my knowledge — never discussed or trained for. I wrote about this after the “Miracle on the Hudson” ditching:
The most surprising thing about multi-engine training is that it doesn’t really consider the possibility of multi-engine failure. Think about it: most multi-engine aircraft don’t even have a Vg speed listed in the Approved Flight Manual. Most type rating programs, even those for airlines, don’t include all-engines-out scenarios. Thousands of Boeings and Airbuses are flying around with flight crews who don’t even know what the best glide speed for their aircraft is.
I understand this is starting to change, but I’m still surprised it isn’t a major part of initial and recurrent training on any multi-engine aircraft. I can think of quite a few incidents in recent years where an airliner lost all engines. Just off the top of my head:
- a British Airways 747 lost all 4 engines after encountering volcanic ash. Engines were restarted at lower altitude. Major engine damage.
- a KLM 747 lost all engines after encountering another ash cloud. Same result.
- an Air Canada 767 ran out of fuel after a conversion error while fueling. Landed on a closed runway.
- a Pinnacle CRJ lost both engines after the flight crew exceeded the aircraft’s limitations. Engines core-locked and plane crashed.
- an Air Transat Airbus A330 lost both engines after a fuel leak. Landed safely on an island.
- an Ethiopian Airlines 767 was hijacked and forced to an alternate destination without sufficient fuel to fly that far. Crashed in the water.
- this week’s US Airways Airbus landing in the Hudson River
A more complete list of unpowered jet airliner accidents is available here. Keep in mind, that list does not include the many turboprops, bizjets, military aircraft, and other planes which have lost all engines in flight. There are so many ways this can happen: fuel contamination, fuel leak , fuel mismanagement, mechanical failure, sabotage, pilot error, bird strikes, hijacking, and the list goes on. It’s baffles my mind that these scenarios aren’t considered during every multi-engine training program.
Gulfstream Aerospace is one of the only multi-engine aircraft manufacturers I’ve seen that includes power-out performance charts for their aircraft.
In a single-engine realm, the typical training scenario involves an instantaneous total loss of power. Sure, it happens, but perhaps even more frequent is a partial loss of power accompanied by abnormal sounds, vibration, and engine indications. That adds significant complication to the decision making process at a moment that has already been imbued with tremendous stress for our intrepid PIC.
If your engine is failing but some power remains, do you land straight ahead, fly the pattern, turn around, bail out, pull the chute, or…? There are so many variables to consider: How high are you? What sort of terrain and obstructions are in the vicinity? What’s the wind speed and direction? How sick is the engine, really? Will it continue to provide some power? Is there smoke or impending fire? Are plane and/or pilot equipped with a parachute? And what’s the glide ratio of your airplane?
It’s that last question I want to focus on. Pilots of powered aircraft are taught a single “best glide” speed to use in the event of a power loss. This strategy has the benefit of being simple. Just one number, one pitch attitude to memorize. The problem is that this single number will not maximize glide performance because it does not account for weight, CG location, and most of all, wind.
Pia Bergqvist recently tackled this and suggested that “minimum sink” airspeed might be useful to powered pilots.
Of the many numbers associated with flying, the best glide speed is one of the most important. The best glide speed allows you to glide the farthest, giving you time to investigate and fix an issue and find the best place for an emergency landing. Having that number at the forefront of your mind when trouble arises can mean the difference between making it to a good landing site or not.
There is also a number called minimum sink, which is commonly used by glider pilots. This is the speed that will allow you to stay in the air the longest. If you have a good landing site within easy gliding distance, you may want to consider using the minimum sink speed. It will maximize the time you have in the air to investigate the problem and attempt to fix it. Minimum sink is, however, generally not published for powered airplanes. You can figure yours out by going to a safe altitude and experimenting by pulling the power and noting the descent rate at different speeds. You will notice that, as you get closer to stall speed, the descent rate will increase rapidly for each knot you slow down.
If you really want to maximize performance based on something beyond the standard “best glide” speed most pilots are taught, glider pilots have an even better technique in their tool box. It’s called speed-to-fly and it’s used to maximize distance traveled (much like Vg) but goes the extra mile — no pun intended — by taking wind into account.
Let’s look at an example: the engine quits and you dutifully establish your Vg speed of, say, 60 knots, assuming that your glide performance will be maximized. But you’re flying into a 60 knot headwind, so the resultant glide ratio is zero. Even if the airport was 10 feet away, you’d never make it.
In this instance, any increase in airspeed over the book number of 60 knots will provide a mathematically infinite increase in glide ratio. If you were to add 1/2 the headwind component and fly 90 knots, your ground speed would increase to 30 knots. Not spectacular, but a heck of a lot better than zero.
I use this kind of computation when flying to places like Catalina Island. There’s a lot of water between the mainland and the island, and most of the Catalina coastline is comprised of steep cliffs without any usable beach. So if you fall short of reaching the airport, you’re stuck with extremely steep, undulating terrain. I take all that into account, along with the wind aloft, in determining my point-of-no return and speed-to-fly when making the overwater leg. Very rarely is the PNR located at the half-way point, and rarely is Vg the most efficient speed if your goal is to fly as far as possible with Sir Newton in the driver’s seat.
So why has speed-to-fly remained exclusively in the glider world when the principles could be critically important to emergency situations in powered aircraft? Because the a) the calculations are complex, and b) an accurate wind speed and direction is difficult to obtain in flight, especially if altitude is not constant.
But technology is changing all that. It seems to me that “speed-to-fly” is particularly useful today because of the proliferation of glass panel avionics. The number of aircraft equipped with an air data computer is rising exponentially. The ADC constantly computes and displays wind speed and direction. It’ll even give you your headwind component, making a speed-to-fly computation quick and easy.
When I was training for my commercial glider rating, our rule-of-thumb was to add adding 1/2 the headwind component to the published “best glide” airspeed. But that was a crude estimate for a soaring neophyte flying without the benefit of computers. Paul MacCready developed more sophisticated techniques (including a variometer add-on called a “MacCready Ring”) to approximate speed-to-fly based on sinking or rising air.
If you’re thinking that this is way too much for a human being to be worried about during an emergency situation, I agree with you. Nobody could make those computations quickly enough, especially as the characteristics of the air change with altitude and distance covered.
But a computer can! It knows actual headwind and crosswind components as well as current vertical speed. All this data could be programmed into a glass panel so that, in the event of a power loss, optimized speeds could be suggested based on the actual wind and, with the addition of an instant or netto variometer, any sinking or rising air currents.
It could also provide a range ring that would give you engine-out glide range based on current conditions. In the example above, a 60 knot headwind might put that field just a few tantalizing miles in front of you well out of range, while an airport behind you and three times as far away that you’d normally have never even considered might be easily reachable.
Companies like Xavion already have apps out for the iPhone and iPad that run engine-out simulations and will guide the pilot to an airport in the event of an emergency. Perhaps it’s time we started teaching the speed-to-fly concepts to power pilots for use in flight planning. Timing is not as critical at cruise altitudes, but at 1000′ AGL, knowing how the wind will affect one’s glide capability can make the difference between landing where you thought you could and having to execute a Plan B at the last minute because winds were not taken into consideration.
With the advent of computers in the cockpit, speed-to-fly needn’t be limited to a conceptual idea we talk about in the classroom. We can make it work for us today. All that’s required is a bit of programming.
Great post! I like the thought of using the power of technology to bridge this training gap. In the Falcon I fly we have dual engine out procedures, but as you said it’s a ‘blunt club’ approach; it doesn’t account for the all of the physics at play.
It’s good to hear Dassault has given some thought to a “worst case” scenario! Love your site, by the way. I’ve added it to my RSS reader.
Nice post. It appears that the apprentice has long become the master.
Haha… thanks, Tim! Hope you’re doing well and have been able to do some flying!
Hey I remember that power-off 180! Good times. 🙂
Yeah, on the plus side I got to do a bit of flying AND demonstrate that the skills can really pay off on rare occasion!
During an engine failiure you should usually maintain the best glide speed especially if you are over water or rough terrain and are trying to make it to a safe and level spot, some distance away. Best glide speed will be your most important value there. You should only really use minimum sink speed if distance is not an issue, but you are trying to buy some time in the air, perhaps to have a few more goes at starting the engine. But really that speed is nnot of much use.