[Previous] [Contents] [Next]
[Comments or questions]

Copyright 1996-2001 jsd

15   Emergency Procedures

Q: ``What should I do if the door comes open in flight?''

A: ``Fly the airplane''.

15.1   Engine Out Procedures

This section discusses what you should do if your engine quits while you are airborne.1 This mainly applies to single-engine airplanes; additional procedures for multi-engine airplanes are covered in chapter 17.

15.1.1   Emergency Checklist

It is important to have an emergency checklist You should commit it to memory, and review it right before each flight. Do not wait until you are confronted with a ``deafening silence'' to figure out what is on the emergency checklist, and why.

If your aircraft manuals do not provide a suitable emergency checklist, you might consider adopting something along the following lines:

Emergency Checklist This list has been designed to make it easy to memorize. You should make every effort to commit it to memory, so that if somebody wakes you up in the middle of the night and asks you ``what is the emergency checklist'' you should be able to shout, instantly, ``AVIATE, NAVIGATE, INVESTIGATE, COMMUNICATE, SECURE!''

The first item, Aviate, is clearly the first priority. No matter what happens next, you want to be in control of the aircraft when it happens. There are lots of scenarios where an engine failure results in a critically low airspeed (especially if somebody is dumb enough to try to maintain the pre-failure pitch attitude, or (worse) the pre-failure altitude while deciding whether or not there has been a failure). If the airspeed is low you must re-establish the proper glide speed2 immediately, even if it means cashing in some precious altitude.

The opposite extreme is possible, too; namely it is quite possible that at the time the emergency begins, the aircraft is going much faster than the best-glide speed. This is not so immediately dangerous, but the longer you take the establish the proper glide speed the more energy will be thrown away in the form of unnecessary parasite drag. In this case, gently zoom upward, converting airspeed to altitude. Retrim.

In addition to trimming for the correct airspeed, you should configure everything else appropriately, as discussed in section 15.1.2.

The second item, Navigate, is clearly next in importance. In section 15.1.4 there is a discussion of clever techniques for judging which fields are within gliding range — but you should not pick a field at the limits of this range if there is anything suitable that is close. In particular, start by looking down at a 45 degree angle, or even straight down. If it is right below you, it is probably within gliding range!

The next item is Investigate. Sometimes when the engine quits, you know immediately what the problem is. Ninety percent of engine stoppages are due to fuel problems, so you should reflexively switch tanks and turn on the boost pump as appropriate. Then turn on the carburetor heat, because it is only effective while the engine is still warm. Then go left-to-right across the panel, checking everything in turn. Make sure the primer is in and locked. See if the engine runs better on the left magneto, right magneto, or both. See if it is happier with a leaner or richer mixture. Give everything a once-over before spending too much time on one particular item, unless you are pretty sure you know what the problem is. And above all, don't forget to fly the airplane.

The next item is Communicate. If you are already in contact with a controller, it is almost certainly a good idea to stay on that frequency. If, on the other hand, you have any doubt about what frequency to use, go immediately to the international distress frequency, 121.5 MHz. That's what it's there for.

Similarly, if you have been assigned your own squawk code, don't bother to change it unless ATC asks you to. On the other hand if you are presently on the all-purpose code 1200, do not hesitate to switch to the emergency code, 7700. That rings alarm bells (literally) at ATC and highlights you on the controller's radar screen.

Some people argue you should Communicate even before you Investigate. Certainly if you are in instrument meteorological conditions you should tell the ATC of your predicament even before you Navigate, (1) so they can vector you to a landing field and (2) so they can clear out the airspace below you.

The fifth item on the list is Secure. It is amazing how easy it is to forget this item. Wouldn't you hate to make a beautiful power-off approach to an ideal field — and then forget to extend the landing gear?

At 100 feet AGL, make sure you pull the throttle and mixture to idle cut-off. The main reason is that you don't want the engine to roar back to life just after touchdown. This could easily happen if (for example) there had been a fuel shortage, and the flare freed up some fuel from a corner of the tank. The reason for doing it at 100 feet AGL is to give the engine a chance to cool down, reducing the risk of a post-crash fire. Closing the fuel-tank shutoff valve helps reduce the risk of fire — but in most planes it is not a sufficiently quick way of stopping the engine so be sure to pull the throttle and mixture also.

Shutting off the engine will be difficult; it will require overcoming a huge psychological barrier. After all, you've spent the last several minutes trying to restart the engine, and now you are supposed to shut it off. Make sure you have made this decision in advance: promise you will shut the engine off at 100 AGL.

Switching off the master also reduces (somewhat) the risk of fire, but in an aircraft with electric flaps and/or landing gear, you might want to save the master switch for last.

15.1.2   Configuring for Glide

In the ``clean'' configuration, the airplane will be able to glide much farther, perhaps twice as far as in the ``dirty'' configuration. If you start out at low altitude, twice nothing is nothing, so it may not be worth messing with the configuration; just configure for landing and be done with it.

On the other hand, if you start out at a reasonable altitude and are trying to glide a long ways, then you want flaps retracted, landing gear retracted, and propeller in the coarse pitch (low RPM) position.

Some other books say that once the flaps are down, you should leave them down; they point out that at a low airspeed (below the bottom of the green arc) retracting the flaps will cause an immediate stall.

I look at it somewhat differently. This situation actually arose on my private pilot checkride. I was at 1000 AGL, with two notches of flaps extended for landing. Then Tower asked me to extend my downwind. By the time I was able to turn final, I was nearly two miles from the airport. At this point the examiner caused a simulated engine failure.

I went through the following thought process: At that point I shoved forward on the yoke. Zero angle of attack. Zero G. The examiner started gently floating out of his seat. I retracted the flaps all at once. I continued the zero-G pushover until we approached the canonical best-glide airspeed. Then I raised the nose, trimmed for best glide, and quickly ran the rest of the emergency checklist. I even rolled in some left rudder trim.

The glide took us to a place in the weeds about 100 yards short of the runway. I flew right down into ground effect and then flared. While skimming in ground effect I extended the flaps. When we reached the runway the stall warning was already on. I plopped onto the runway. We were stopped before reaching the big painted number.

The point of this story is this: If you need to glide a long ways, retract the flaps. Just do it in such a way that you don't stall.3

You can, of course, glide with flaps and/or gear extended if you want to make a steep approach to a nearby field.

Also, when you are through gliding (i.e. when you are ready to flare), make sure the flaps are extended, so you can touch down at the lowest possible speed.

For landing on water, in most airplanes you want the gear up. For landing on most other surfaces, you want the gear down. Don't wait until the last moment to put them down; with the engine off it might take longer than usual to get them down. Make sure you know how to use the manual gear extension system on your airplane.

15.1.3   Return to Airport?

We now focus on the special case of engine failure shortly after departure, since that is a relatively common and very critical case. Many people are tempted to turn back to the airport, but this is not usually the right answer.

The right answer depends on many factors, including:

Every situation is different, so the following analysis can't possibly fit them all exactly. On the other hand, it is worth your while to plan in advance. Know what your options are. For each phase of flight, make sure you have a backup plan (``Plan B'') appropriate to the situation. Be ready to carry out Plan B at a moment's notice.

Here is a piece of simple but important advice: if you can land straight ahead, do so. As an extreme example, consider this: a small plane departing from runway 31L at JFK (length: 14,600 feet) could climb to 500 feet, lose the engine, and still land straight ahead on the same runway with plenty of room left over.

Here's another piece of simple advice: don't turn back unless you are sure you can make it — and there are lots of situations where you can't make it.

For example, consider a fully loaded Cessna 152. It has a power-off glide ratio of ten to one. Unfortunately, in no-wind conditions the climb gradient is less than ten to one. Therefore, even if the airplane could turn on a dime, at every point on the return trip the airplane would be below where it had been on the outbound trip. Then when you take into consideration the altitude lost while getting the airplane turned around, it is easy to see why the airplane cannot possibly return to the point where it left the ground.

Under such conditions, the farther you have flown on the departure leg, the more options you have for an off-airport landing, and the more impossible the turnback becomes.

An important factor to take into account is that a simple 180 degree turn does not suffice to return you to the departure runway. The airplane will travel an appreciable distance sideways during the turn. You won't need to do a full-blown procedure turn, but you will need to do some additional maneuvering that makes an already-bad situation worse.

Given a sufficiently long runway, the airplane may be able to return to a point on the runway closer to the departure end — which is just fine. Again, imagine departing runway 31L at JFK, and climbing straight ahead. Suppose the engine quits at a point 1/2 mile beyond the departure end of the runway. At that point you should have more than a thousand feet of altitude. You should be able to reverse course and make a downwind landing near the beginning of runway 13R4 even though you could not glide back to the point where you lifted off.

A modest headwind on departure will help keep the airplane near the airport during the outbound leg, and will help hurry it back to the airport during the return trip. This sounds wonderful, because it increases the possibility that you can glide back to the runway. The trouble is that (whether or not you make it back to the airport) you are faced with a downwind landing. Even a modest amount of tailwind (say 15 knots) can have a tremendous effect. Suppose your airplane is capable of touching down at 55 knots. If you land into the wind, you have a groundspeed of 40 knots, but if you land downwind you have a groundspeed of 70 knots. Runway usage depends on the square of the groundspeed, so the downwind landing will use three times as much runway: (70/40)2 = 3.06. Also, in a collision, the amount of damage and injury is typically proportional to the square of the groundspeed — so if you turn downwind and don't manage to land on the runway you are in very big trouble indeed.

Here's another option for you to consider: suppose that your airport has a second runway running crosswise to the active runway. If your engine fails somewhere over the cross runway, you might be able to turn 90 degrees and land on it.

Even in less ideal cases, it is quite likely that a crosswind landing on a different runway (or even a taxiway) is easier and safer than a downwind landing on the departure runway.

If you are really concerned about engine failure during the departure climb, and the airport is the only safe landing zone for miles around, you should begin a gentle turn almost immediately after liftoff. Then if the engine quits, you're closer to the airport and you've got a more convenient heading. I don't recommend this in general, because engine failure is not the only consideration. For starters, we need to worry about causing a mid-air collision in the pattern. A turning departure climb toward the traffic-pattern side of the runway would cause you to enter the downwind leg from below at just about the point where inbound traffic is entering from the 45 degree leg.

In many cases, engine trouble results in partial rather than total power loss. You have to decide whether continued operation of the damaged engine is safe, but if so, it gives you some more options. Even if the aircraft is not capable of climbing on the remaining power, the rate of descent may be dramatically reduced. Think of the aircraft as a noisy glider with a good glide ratio. It may be capable of ``gliding'' to places that a totally unpowered aircraft could not.

I reiterate: don't try to turn back to the airport unless you are sure you can make it, and in most typical cases you can't. You should find a nice road or field and put it down under control. It helps a lot if you have practiced forced landings, so you know the power-off performance of your airplane and how to land from non-ideal approaches.

Another very serious consideration is this: reversing course smoothly wastes valuable time and energy, whereas reversing course quickly requires radical maneuvering. Nobody wants to recommend that pilots perform radical maneuvering at low airspeeds close to the ground in an unplanned situation. It might help you return to the runway, but there is a much greater chance that it will provoke a stall/spin accident. An off-airport landing is not usually fatal, whereas a stall/spin accident usually is.

Another reason for not attempting to turn back is that most people are so surprised by an engine failure that it takes them a few seconds to regain their wits. During this interval, precious time, energy, and distance have been wasted, so even though it might originally have been possible to turn back, it no longer is.

Indeed, (unless you are very well trained) your first reaction will be completely wrong — not just late, but dead wrong. When the engine quits, the airplane's nose will tend to drop. The flight path has changed from, say, a 10-to-1 climb to a 10-to-1 descent, so to maintain constant angle of attack the nose must drop a huge amount — 12 degrees. You may think ``Gee, I don't want the nose to drop'' and may be tempted to pull back on the yoke. This is a sure way to kill yourself. Please, do not think of the yoke as the up/down control. When the engine quits, the airplane is going to descend. The only question is whether you will spin in, or glide to a controlled touchdown. Remember, you can survive an off-airport landing if you touch down under control.

The obvious reason why you want to maintain control is that the rate of descent you get in a normal glide is much less than what you get if you stall and let it ``drop in'' — not to mention what you get in a spin.

A less obvious but still very important consideration is this: If the airplane is not under proper control, it is likely that one wingtip will hit before anything else. This will cause the aircraft to cartwheel, causing tremendously more damage and injuries than if you had landed under control and just skidded to a stop.

It is important to have a plan. At the airport(s) you use regularly, scout out the territory near the departure paths, and formulate a plan for where you will land if the engine quits. (Further discussion of the power-off glide appears in the next section.)

In any case, you need a plan for what to do with the controls. Your first priority is to maintain a proper angle of attack. Do not attempt to hold the nose up; let the nose go down (or push briefly to help it go down). Fine tune the pitch attitude and trim to maintain the best-glide airspeed.

Land into the wind if possible.

15.1.4   Power-Off Glide Perception and Planning

In a forced-landing situation, your glide path will be rather steep. The lift-to-drag ratio of typical Skyhawk or Cherokee (in best-glide configuration) is about 10-to-1, which corresponds to an angle of six degrees. This is perhaps twice as steep as a typical power-on approach. It is even somewhat steeper than the typical ``power off'' approach, since that normally really means ``engine idle'' and an engine at idle produces noticeably more power (and less drag) than an engine that is really off.

We can use the rule of thumb: a thumb at arm's length subtends four degrees.5 As shown in figure 15.1, you can just barely glide to any point that is a thumb and a half below the horizon (assuming no-wind conditions). If your engine quits, you should run your thumb around the horizon to help you select possible landing sites.

Figure 15.1: What Landing Sites Are Reachable?

In the presence of wind, the circle of possible landing sites will be shifted downwind from the circle described in the previous paragraph. Suppose you are gliding at 60 knots (airspeed) into a 30-knot headwind. Your groundspeed has been cut in half, so your glide will be twice as steep as in the no-wind case. Your destination must now be at least three thumbs below the horizon.

It is well worth knowing how far you can glide. Suppose you are roughly one mile up, so that you can glide ten miles. There is more area in the ring between seven and ten miles away than there is in the entire disk between zero and seven. (On the other hand, if find a close field, go for it. Don't glide a long way just because you can.)

Now let's assume you have picked a field and are gliding toward it. Now your ability to perceive angles really pays off. Whereas in a normal approach you use engine power to maintain a pre-determined glide angle and destination, in a maximum-distance glide the glide angle is fixed and you want to perceive what the destination is.

Here is the key idea: there will be some landmark that remains some fixed angle below the horizon, and that is the point toward which you are gliding. Pick a point. If its angle down from the horizon is decreasing, you will land short of that point. If its angle is increasing, you will overfly that point — unless you do something.

There are two critical reasons why you always need to explicitly identify the point to which you are gliding. For one, you need this information for angle of attack control. Remember, the angle of attack is the angle between the wing and the direction of flight. If you can't perceive the direction of flight you won't be able to perceive the angle of attack (see chapter 2). Secondly, you need to know whether your present glidepath will cause you to land long or land short; the earlier you can perceive this the sooner you can make any necessary correction.

A small correction early is better than a big correction late.

You should not pick a field that is right at the limit of the airplane's gliding ability, because you'll wind up short if anything goes wrong. Pick a field that is substantially closer than the limit, since you can always lose altitude by circling, adjusting the length of the base leg, extending flaps, slipping, et cetera.

Strategic turns are appropriate early in the game; flaps and slips are more appropriate on short final.

S-turns on final are almost never the best way to eat up unwanted altitude. Small-angle turns have almost no effect, and large turns take too long to perform, and take your chosen field out of sight temporarily. Furthermore, after two turns (one to the left and one to the right) you will be back on your original heading, but offset laterally; you need to make two more turns to get back on course. If you have time and altitude to do all that, there are better things to be doing.

If you are on long final and can (using flaps and/or slips) keep the field a constant angle below the horizon, you are all set. Glide straight on in.

If you are on long final with excess energy, or if you are approaching the field from a substantial angle relative to the intended direction of landing, do not aim directly for the field. Aim for the so-called base key point (figure 15.2), i.e. the point where the base leg begins.

Figure 15.2: Forced Landing from Base Key Point

From the base key point, you have a lot of options. If you arrive with the ideal amount of energy, you can fly a nice base leg and then turn final. If you arrive with slightly more or less energy than that, you can angle the base leg away from or toward the field. Fly along the base leg until the desired destination is the appropriate angle below the horizon, then turn final.6

Another point that is made in the figure is that you can use the width of the landing field to your advantage. You may be unaccustomed to this, since at an airport the tradition is to land always as close as possible to the runway centerline.

Also, do not plan your final approach to take you to the threshold of your chosen field. No matter how long or short the field, aim for a point one third of the way along the field. Remember there are lots of things that could steal energy from your glide, and you really don't want to land short.

It is better to hit the trees at the far end at 20 knots
than to hit the trees at the near end at 60 knots.

The energy, the expected damage, and the expected injuries all depend on the square of the airspeed, so the 20-knot collision involves nine times less energy than the 60-knot collision.
The main reason why you might wind up landing short (despite a well-planned glide) is the infamous decreasing headwind on final, as mentioned in section 12.12. If you are gliding at 70 knots, and you lose just ten knots due to a windshear, you will have to descend 60 feet to regain your proper airspeed.7 That 60 vertical feet (at a ten-to-one glide ratio) corresponds to 600 feet of horizontal travel. If you are unprepared for it, finding yourself 10 knots slow, 60 feet low, and/or 600 feet short during a forced landing is no fun.

15.2   Preventing Emergencies

15.2.1   Safety Margins

You need to have margins of safety and layers of protection. You don't want to be in a situation where any one failure causes harm. You always want to have not just plan A, but also plan B, plan C, plan D, et cetera.

For example: You should pay attention to anything that peels away one or more layers of safety. Keep track of how many layers remain.

If one magneto fails, park the airplane until it is fixed!

If the fuel gauges cannot be trusted, park the airplane until it is fixed! Do not rely on clock-and-dipstick alone. By the same token, do not rely on the gauges alone.

15.2.2   Fuel Management

Far and away the most common reason for losing engine power is fuel mismanagement. This includes running out of fuel as well as contamination of the fuel. The good news is that such problems are relatively easy to prevent.

It pays to be careful. I am pretty methodical about checking the fuel sumps. I used to check for little droplets of water at the bottom of the sampler. Once, after completing the check, I was about to dump out the sampler and begin the flight, but I decided to take a closer look. Then I noticed tiny drops floating at the top of the sampler. It turns out the entire tube was full of water, with just a tiny bit of fuel mixed in. I went back to the plane and got another tube of water, and another, and another. It turns out that the fuel vendor had just switched to a new pump/tank system, and had sold me more than a quart of water (at AVGAS prices!) along with the fuel.

So, here are some suggestions:

15.3   Dealing with Emergencies

As suggested in the epigraph to this chapter, the first step in dealing with any in-flight emergency is to fly the airplane.

This sounds simple and obvious, but there have been far too many cases where the aircraft stalled or flew into the terrain because the pilots were too busy fussing with something that should have been only a minor distraction.

For instance, what should you do if the door comes open during takeoff? Answer: fly the airplane. No general aviation airplane I know of will crash because the door is ajar. The door will be held open about an inch or so. There will be enough suction to make it rather hard to close that last inch. There will be a bit of noise and a bit of a draft, but perhaps less than you might imagine. You might not even notice at first. In any case, the safest thing to do is to ignore all the details and return for landing. When you are safely stopped you can fiddle with the door as much as you like.

As another example, suppose you are just about to turn onto final approach when you notice that only two of the three landing gear are indicating ``down and locked''. What should you do? Go around! Do not try to debug the landing gear on final. For that matter, do not try to debug anything in the traffic pattern — it is too close to the ground and (usually) too congested. Get out of there. If there is a control tower, don't forget to tell ATC what's happening: ``Tower, Five Seven Tango has some uncertainty with the landing gear.8 We'd like to leave the pattern while we investigate''. Then climb to a reasonable altitude, away from the airport, and take your time fixing the problem.

Engine trouble (or other trouble) during the takeoff roll is discussed in section 13.8, especially section 13.8.4. Being lost, which might or might not be an emergency, is discussed in section 14.6.
As part of the pre-takeoff briefing, remind yourself of the best-glide airspeed for the airplane you are flying. It varies a lot from plane to plane. Find this speed on the airspeed indicator and point to it. (Don't just remember the numerical value, because in an actual emergency you will probably be so excited that you lose your ability to remember numbers. Your ability to remember geometrical relationships will be less impaired.)
In retrospect, I wish I hadn't pushed quite so hard. A half-G pushover would have been more than sufficient, and would have kept the examiner from floating out of his seat. But basically I had the right idea.
i.e. the reciprocal direction on the same piece of pavement.
See section 12.3 for more discussion of how to measure angles.
Note this method gives you control of your destination using at most two large-angle turns, whereas S-turns on final would require using four.
Remember the law of the roller-coaster: 9 feet per knot, per hundred knots.
On a training flight, you might want to let Tower know it is only a simulated problem, so they don't get unduly worried. When in doubt, ask the instructor, but usually you know it's a simulation because of the satanic grin on the instructor's face.

[Previous] [Contents] [Next]
[Comments or questions]

Copyright 1996-2001 jsd