« Back to Full Story

Post-Takeoff Engine Failure: Not So Rare

  • E-Mail this Article
  • View Printable Article
  • Text size:

    • A
    • A
    • A

If you haven't seen this video, take a few moments to watch it. In a nutshell, it's a video summary from several angles of a post takeoff engine failure in a Bonanza and the subsequent forced landing just off the airport. The pilot, who along with his passenger, emerged unhurt from the accident said he posted it as an educational exercise. Frankly, I'm at a loss to say what you might learn from the video, other than the importance of bank control and airspeed management. I'd surmise that most of us already know that. The airplane, by the way, was equipped with four external cameras to catch all this and you're allowed to wonder, as I did, how odd that it is that an airplane so equipped would crash. But that's not the main point, really.

The real educational moment is the underlying cause: an engine failure or power loss just after takeoff followed by an investigation that reveals nothing wrong with the engine. If this seems odd, it shouldn't. It happens a lot. In fact, it happened to a Mooney we once owned, leading to the loss of the airplane. For our Used Aircraft Guides, we routine review about 100 or more accidents each month for a specific type. The engine failure on takeoff scenario isn't a leading cause of accidents for most airplanes, but for some it ranks high. Mooneys and Arrows equipped with the 200-HP Lycoming IO-360 are one, but you see this across all types.

If the airplane or engine is recovered, the investigators set it up and try to run it and it many cases it starts and runs fine, as did our Mooney. The probable cause is usually given as engine failure for unknown reasons. It would be convenient to blame the pilots for all of these, but they're just too similar for that to be the case, in my view. No one seems to particularly care about this identifiable trend, either, not the NTSB, not the FAA nor the insurance companies. Just before they wrote our partnership a $110,000 check for a hull loss, I asked our insurer why they don't investigate these things more thoroughly. The answer was right out of the MBA playbook. Not many result in injuries or deaths, there aren't enough of them to represent an addressable loss pattern and digging into them further is expensive and likely to be inconclusive. As long as the insurance business unit is hitting its numbers, no one seems to care. And that's how system "safety" works in GA.

In its accident report on the Fairbanks crash, the NTSB had more data to work with than it usually does, given camera roll from four angles. But the footage doesn't reveal much. In the body of the report, the NTSB notes that there was no evidence of high RPM operation on the ground, implying, I take it, that no run-up was done. Well, big deal. Who can say if a run-up would have made any difference. I tend to think it wouldn't, because a lot of these post-rotation stoppages occur after a normal run-up. While a run-up might reveal fuel flow or pressure anomalies, they're mostly focused on magnetos and my theory is that it isn't ignition that causes these unexplained stoppages but some kind of fuel issue, perhaps a vapor lock that clears after the engine has cooled. As far as I know, no one has investigated these incidents in any systematic way, so who can really say?

As for the run-up, there are two schools of thought here. I'd guess the vast majority of pilots perform a run-up before every takeoff, while a minority do the run-up only on the first flight of the day. I'm in the latter group. When I used to fly multiple legs a day for charter and freight, I'd do one run-up in the morning and that sufficed for the entire day. A mag check will catch a bad mag or a fouled plug, but neither of those is likely to put you into the trees. If you accidently switched the fuel off, a run-up may or may not reveal that. So it's a calculated risk to skip the run-up. Just because some people don't doesn't make it wrong.

Maybe the real lesson in the Fairbanks accident, then, has nothing to do with run-ups, airspeed management or bank control. It's the larger consideration of realizing this could happen to anyone at anytime and you'd be wise to expect it on every takeoff.

Comments (25)

I'm at a loss to say what you might learn from the video... I am not. I fly a twin when my family flies with me. In a twin at that point it would be a routine maneuver: Identify-Verify-Feather. Then pitch for blue line, get yourself some speed & positive climb, turn around & land.

Posted by: ROBERT ZIEGLER | June 27, 2013 1:51 AM    Report this comment

This explanation by the Insurance company sounds exactly like what is said about UFO investigations...we know they are there..they haven't killed many people, can't explain them, so we will just pretend they are not there. Maybe these failures on takeoff are due to lack of proper warm up of engine since the internal parts of the engine are not yet done expanding due to a quick taxi and takeoff. Whatever the reason, this "max power" scenario during takeoff and initial climb is the highest stress environment for any engine, be it turbine or piston....

Posted by: Blaine Banks | June 27, 2013 2:17 AM    Report this comment

Being a CFI who has done a fair amount of flying in student's planes likely puts me in a group at high risk of this particular scenario. Almost test pilot status you might say and being involved in 3 of these incidents seems to add credence to this observation. No cause was determined in the first, a 260 commanche at night. A long runway enabled us to get it back on the remaining runway and gear up enabled a short field landing. The second in an RV4 was caused by a piece of the alternator belt shedding and blocking the air intake with the carb heat on. Yes there was a reason for takeoff with it on. The third in a Pacer was caused from a piece of RTV breaking loose in the carbuerator and plugging the high speed jet. Never did thank that A&P for using RTV to fix that pesky leak. No harm done , there was a familiar ag strip 3 or 4 miles off the end of the 5000' runway we were using for tailwheel touch & goes and were able to glide there and land with no damage. Sorry I have no insightful revelations from these experiences, just had to add my 2 cens worth to the conversation and say yes I would have to agree this is a high risk phase of flight for various reasons.

Posted by: Chuck Lemasters | June 27, 2013 5:31 AM    Report this comment

Paul, the most important paragraph of your blog is the last one, and it should be highlighted and underlined. The pilot who fails to brief themselves as well their passengers, on the appropriate actions in the chance of an engine failure at any point during the takeoff and initial climb is only courting disaster and the potential to do the wrong thing when the engine fails. Just as the good instrument pilot expects to end every approach with a missed approach and takes the opportunity to see the airport and land as an act of grace, should also incorporate the mentality that when the engine keeps running and the climb out is uneventful it is an act of wonderful grace. As one who has "been there - done that" I, like you, know this. Thanks for the good article and keep up the good work.

Posted by: doug stewart | June 27, 2013 6:34 AM    Report this comment

Well, here's my take on a "Run-up" requirement. I was trained to fly Cessnas, by a CH47D helicopter pilot, military of course. My CFI was ALL ABOUT the CHECKLIST! He taught me two IMPORTANT ITEMS, as many excellent CFIs do. "ALWAYS!! Use the Checklist, and NEVER skip the Run-up! As an A&P, I also "Use a Checklist" of sorts, to perform my maintenance. A Checklist gives me "Hope" that things will go as planned. That's my thoughts.
Thanks, BILLS

Posted by: Bill Story | June 27, 2013 6:39 AM    Report this comment

I have often wondered in these scenarios if the fuel was turned off, and the pilot turned it back on after landing and never admitted it.

Posted by: Michael Mcnamara | June 27, 2013 7:00 AM    Report this comment

I've nearly had this happen to me flying an unfamiliar airplane that had an alternate air source right next to the fuel shutoff valve. A long runup identified my mistake, which otherwise would have been revealed somewhere down the runway.

The bigger issue I see is preparation for an engine failure. I teach every student of mine to brief each takeoff, going over what the plans are if the engine quits. Takeoff briefing is on all the checklists I use. But I've seen its one if the first thing to be discarded when they start flying on their own. With all the flight reviews and aircraft checkouts I've done, I can only remember one pilot who briefed the takeoff. That's an issue that needs some attention.

Posted by: dennis mahan | June 27, 2013 7:16 AM    Report this comment

Oh my. Is the incident in Two Four Fox, which happened 46 years ago, still relevant? It was at Fox Field in the Mojave Desert that next Saturday. And it was on my first landing for what was intended to be a series of touch-and-goes. Touchdown was on target. Rolling on the runway, I brought the flaps up. Then I shoved the throttle to the fire wall and listened to the engine come up to full RPM. Half-way down the runway and lifting off -- trouble!
The engine reached its peak RPM and started slowing down. The sound -- "woo-AH-oo-oh" -- has echoed in my memory ever since. Four slurred syllables, "woo-AH-oo-oh." Without waiting for the engine to slow down further, I pulled the throttle and aborted the take-off. There was runway remaining. Not enough.
Two-Four Fox touched down again on the main gear. I blew both tires braking. The plane rolled off the end of the runway between threshold lights and chased jackrabbits through the sagebrush. The nosewheel dropped into the sand. For an instant I was sure the plane would flip. Dust filled the air.
With the plane stopped, the engine continued to run, nice as you please.

Posted by: Paul Niquette | June 27, 2013 8:00 AM    Report this comment

"In a twin at that point it would be a routine maneuver: Identify-Verify-Feather. Then pitch for blue line, get yourself some speed & positive climb, turn around & land."

That's assuming a well-practiced multi-engine pilot, and that the plane can even get a positive climb at the given DA and loading. At many of the airports I fly out of, an engine loss on takeoff in a twin is nearly the same event as an engine takeoff in a single. But regardless, the answer is the same: be prepared and well-practiced.

"A long runup identified my mistake, which otherwise would have been revealed somewhere down the runway."

This is the reason I've modified my start-up and run-up procedure to include starting on one take (usually the right), and then as soon as the engine is running, switch to the other tank. I'll then do a run-up and the departure on that second tank, and I've now verified both tanks have fuel I can get. I've also provided sufficient time to allow the fuel lines to drain BEFORE takeoff, on the off chance that I accidentally switched the fuel selector to off.

I timed it once with the engine at ground idle and the fuel selector to "off", and it took almost 3 minutes before the engine quit. Since taxiing usually only requires a little more than ground idle, I can see why many pilots have gotten to the runway and through rotation, only to have the engine quit due to fuel starvation.

Posted by: Gary Baluha | June 27, 2013 8:19 AM    Report this comment

there was a very extensive discussion of this on Beechtalk that brought up several scenarios one of which was an older fuel line having the inner liner collapse under the high suction occurring at maximum fuel flow.

this is a problem that we as pilots should try to get to the bottom of as although it seems it does not kill many peopl it still is probably an avoidable problem if we knwe the cause.

After reading that discussion on Beechtalk, I resolved to make sure all the flexible fuel lines on my aircraft are changed frequently.

Posted by: BILL LAWSON | June 27, 2013 8:30 AM    Report this comment

It is probably the combination of several classes of “oops” that leads us to perceive takeoff as being a particularly likely time for engine failure.

As several previous comments noted, it is a time of abrupt transition mechanically speaking, with sudden vibration onset and radical heat change probing for any weak spot. Then, in warmer weather at least, there is always the possibility of vapor lock, particularly when the engine is already hot and/or there have been long ground operations.

I suspect though that statistically the cause is most often some type of operational error, and almost always fuel-related: A selector on the wrong tank or not fully in the detent, brief un-porting due to low fuel level, etc. Unfortunately a run-up probably won’t detect these problems and even the checklist’s ‘fuel on fullest tank’ admonition will fail if your brain says “yeah, I did that”.

Posted by: John Wilson | June 27, 2013 9:17 AM    Report this comment

One often-overlooked reason for these failures can be a lack of preheat in cold (below 20F) weather. Probably not relevant in this case, as it was during the Summer, however.

Engine cylinders are made of steel, and pistons are made of aluminum. Additionally, the tops of cylinders taper inwards somewhat as they tend to be hotter than the bottoms, so that when they get up to operating temperature, they are close to straight.

Since aluminum expands faster than steel, the right combination of temperatures, you can get to a point in a cold-soaked engine, that the piston actually expands to a radius that is greater than the top of the still-cold cylinder. This can cause the engine to seize or operate very roughly.

This isn't repeatable later, potentially, because the investigators will bring the engine indoors, where it will warm-soak and be above a temperature where the above would occur.

Posted by: Joshua Levinson | June 27, 2013 12:06 PM    Report this comment

I fly a C172XP [ with the Cont IO-360],and a C175 Both are placarded against take off with less than 1/4 tanks, with 10 gal unusable. This is due two things, the 52 gallon tanks have 5 gallons of fuel aft of the fuel port, and the high Vx pitch and ROC of 1,200 to 1,600 fpm. Both are fun to fly, but I stick the tanks before every flight to ensure that I have at least 1/2 tanks.

A recent loss of power on takeoff occurred at KMYV. The responders reported that the left tank was full and the rt tank had 3-4 gal.,and the fuel selector was on the rt tank. By the time the NTSB got to the a/c, a week later, the carb and gascolator had fuel in them. They didn't care about the fuel in the tanks.

Posted by: JAMES HUGHES | June 27, 2013 12:45 PM    Report this comment

I had an Cherokee lose all thrust on takeoff once, a hot July early eve at a mountain airfield, and the cheery NTSB investigator told me he was happy to talk to me because there's usually no one to talk too after one of those. The NTSB ended up agreeing with the first ATP/A&P IA on the scene, that it was vapor lock.

Low wing aircraft need functioning fuel pumps to deliver any fuel to the engine, and hot fuel can cause cavitation with a loss of efficiency, sometimes down to nothing, but all that really has to happen is for the delivered fuel rate to be less than the rate of fuel being delivered by the carb,

I think this one of a very short list of the advantages of the C172 over the Cherokee competition. It's hard to beat gravity feed with the fuel selector pointing to a tank with fuel.

Summer is here; my suggestions are, especially for a hot day, to take any lower than usual fuel pressure indication as a no-go even if it is in "the green", and, late in the day after a heavy sun soak, add some fuel if you're fairly low, even if your flight doesn't require it. Nice cool fuel from the pump will reduce the temperature of the fuel in your wings.

My original CFI/CFII, still flying despite being a WWII WASP, knew of several Bonanza and Cherokee accidents that were probably vapor lock.

Posted by: Greg Goodknight | June 27, 2013 6:11 PM    Report this comment

I have found it very helpful to ALWAYS self-brief the takeoff so that any shortcoming in performance is noticed by the expected rotation point, with plenty of time to abort. Picking the spot on the runway at which you are aborting if you are not airborne is preferable to the "gee I don't seem to be accelerating as quickly as normal" approach.

This does NOT help with a normally-performing engine that quits a couple hundred feet airborne, though. For that, have the soft, cheap spot picked out that you will turn to below the minimum safe altitude for your airplane at max gross to execute a 180--you have practiced that at altitude, right?

I was in the backseat of a CFI friend's 172/180 when the carb ingested a nut that came adrift from the intake system and ended up killing a cylinder. A slight burp on the takeoff roll that the 3 CFIs aboard all ignored (any one of us would have aborted, but each of us deferred to the silence of the other two) and we got to practice the "impossible turn" back to the runway.

All safe but several lessons learned, most of all is expect the engine to fail, and you won't be surprised. (And don't ignore that odd noise or skip -- your engine is trying to tell you something).

Posted by: Lawrence Brennan | June 27, 2013 7:16 PM    Report this comment

in regard to Mr Mcnamara's post, a Bonanza won't flow fuel if the selector isn't fully engaged in the detent for each tank. I got a call from an owner who lost power on the takeoff roll, had him ensure the fuel selector was properly engaged on appropriate tank, it wasn't quite in and it caused a loss of power Just a thought.

Posted by: Josh Johnson | June 28, 2013 10:41 PM    Report this comment

We have engine failures on takeoff with or without run-ups. Run-ups only catch gross failures and, as the story says, post investigations don't find anything wrong either. It may even cause problems if you are doing a quick turn around and do a run-up with a hot engine on a hot tarmac.

Posted by: Mark Fraser | June 30, 2013 9:12 AM    Report this comment

Earlier this afternoon, I posted the latest on this topic...and the news is tragic. The pilot Paul mentions who survived the Bonanza crash was killed this week in a subsequent crash in a Baron, also in
Alaska. For some reason, that post has not appeared, so, I am trying again to post.

Posted by: ED LIVERMORE | June 30, 2013 3:21 PM    Report this comment

The Anchorage Bonanza accident description does not say if this was the first flight of the day or after a flight and heat-soaked. The Peterson autofuel STC letter that comes with the mogas stickers goes into great detail on this. The cautions associated with mogas also apply to avgas at high DA and/or a heat-soaked engine, especially Lycs with fuel pump and carb bolted to the engine.

The location of fuel pumps for injected engines could be a factor. I don't recall a discussion of such an event with the IO-360 200 hp Cardinal RG. The 177RG has a small vented fuel reservoir and boost pump in the floor. Could that design reduce vapor- or hydro lock?

Posted by: Thomas Connor | June 30, 2013 3:59 PM    Report this comment

The restart IO-lyc powered Cessnas have a fuel vapor return line and stacked fuel valves to circulate fuel thru the pumps and back to the tank it came from. I've never had a problem with them.

As a side note, the carb is bolted to the hot oil sump on most Lycs, and a theory is that during heat soak the float bowl boils dry, the float valve opens wide and vapor from the fuel vaporizer - er - fuel pump floods the engine. Sawing on the mixture control to find a sweet spot - sometimes all the way back to ICO - solves the problem long enough to formulate a plan-B. Monitoring fuel pressure and fuel flow for anomalies gets you ahead of surprises on hot and high days.

Todd Peterson calls fuel vaporizing at the float valve 'hydro-lock.' That's not to be confused with vapor-lock, where fuel boils in the pumps. In the case of high-wing planes with lyc engines, the hot engine mounted pump becomes a fuel vaporizer, causing more problems than it solves.

Automakers solved vapor lock years ago by mounting pumps in the tank to 'push' fuel to throttle bodies with return lines to keep it cool. They also solved hydro lock by raising fuel pressure above the boiling point and added fuel injection.

Posted by: Thomas Connor | June 30, 2013 4:00 PM    Report this comment

100LL has a vapor pressure of 7. Peterson and EAA are familiar with the high vapor pressures of mogas ranging from 7-15. Experimentals I fly with the O-320/360 engine burning mogas vapor-and hydro- locked horribly at high DA airports. We installed valved vapor return lines with a clear segment in the cockpit running from the carb to a tank. Before start we run the boost pump and push fuel thru the pumps plumbed in series to cool them until there are no bubbles in the sight gauge. After start we monitor the sight gauge and set the valve to put fuel pressure in the green and go. If bubbles return or fuel pressure falters, we open the valve to get more cooling to the pump and carb. Problem solved.

Adding a return line to a Legacy Cessna would be a piece of cake if it weren't for bureaucrats more concerned with compliance than safety.

Posted by: Thomas Connor | June 30, 2013 4:01 PM    Report this comment

I ran a C-182 for thousands of hours on mogas, including frequent operations out of Bullhead City AZ with heat-soaked tanks and ambient temps above 120F, never had a problem.

I think vapor lock has more to do with the individual aircraft...fuel line routing, etc., than with the fuel itself.

Posted by: John Wilson | July 1, 2013 12:48 PM    Report this comment

John Wilson: If your C-182 had the O-470 engine it lacks the two features that cause Lycs to have problems: The O-470 relies on static pressure from the high wing, hence no boost pump to vaporize fuel or vapor lock. the O-470 carb is also not connected to the hot oil sump, avoiding hydro lock.

Posted by: Thomas Connor | July 1, 2013 2:56 PM    Report this comment

Sometimes there's nothing you can do. A friend of mine was departing in daytime VMC in his company's C208 when, at 7000' AGL, the turbine threw blades through the case, and covered his windshield in oil. With not enough altitude to make it all the way back to the departure airport, and not enough visibility to identify a nearby grass strip (they said his radio demeanor was calm and focused), he put it down gently on a grass field into the wind, only to roll into a grove of trees, where a large branch (which he surely could not see) penetrated the windshield and killed him. Neither he nor I would ever advocate not flying, but I would now suggest we begin planning for the unthinkable - what would you do if you suddenly lost forward visibility and the means to maintain altitude?

Posted by: Bruce Liddel | July 1, 2013 4:07 PM    Report this comment

I'm late to the party - Gary offered the following start/run-up procedure:

> This is the reason I've modified my start-up and run-up procedure to include starting on
> one tank (usually the right), and then as soon as the engine is running, switch to the
> other tank. I'll then do a run-up and the departure on that second tank, and I've now
> verified both tanks have fuel I can get. I've also provided sufficient time to allow the fuel
> lines to drain BEFORE takeoff, on the off chance that I accidentally switched the fuel
> selector to off.

I'm not sure your startup would drain all the fuel from the system in to prove that your first tank was working.

I use a similar startup procedure in my Cherokee though - Start, Taxi, Run-Up, and Takeoff on the fullest tank (or as-required to even out lateral trim/weight). My instructor's logic, which I agree with, is that if I got all the way through the runup and out onto the runway I know the tank I'm using is supplying fuel to the engine (and if it's not I probably only have a few seconds of full throttle fuel left in the lines, so I'd still be on the ground).

Once I have some altitude between me and the ground I immediately switch to my second tank to confirm it too is able to supply fuel to the engine - If it doesn't I can always go back to the first tank (and head back to the airport I just left).

This procedure means that if my fuel selector fails on that first change I'm pretty well out of luck - but I consider that pretty unlikely, and at least I've got altitude (time) on my side.

Posted by: Michael Graziano | July 15, 2013 4:16 PM    Report this comment

Add your comments

Log In

You must be logged in to comment

Forgot password?


Enter your information below to begin your FREE registration

« Back to Full Story