Statistically Speaking: Will you have an engine failure this year?

The good news for many T&T readers is the FAA says turbine engines have a failure rate of one per 375,000 flight hours compared to one every 3,200 flight hours for piston engines. This means jet engines are 117 times less likely to quit than reciprocating ones. According to the NTSB, there are somewhere between 150 and 200 accidents per year that are caused by power loss. For piston twins and experimental aircraft, the accident rate is higher. 

There were over 4,000 accidents attributable to engine failure during a recent five-year period; that’s about two per day. But the actual engine failure number is likely double or triple that rate when you consider the number of engine failures that resulted in a successful landing with no damage i.e., not an “accident” – like mine and my buddies mentioned below. It’s a new year; will this be the year that you have your first engine failure? If you have more than 5,000 hours and have not yet had one, statistically speaking, the odds say yes. 

Nobody who gets too damned relaxed builds up much flying time.

  – Ernest K. Gann

Fate is the Hunter
Fate is a fixed course of events. It may be conceived as a predetermined future, whether in general or of an individual. The word fate traces back to the Latin word fatum, and something that’s your fate is a done deal, not open to modification. The quintessential fate endorsing works of Ernie Gann notwithstanding, the words random and statistical probability or odds are less philosophical when describing the likelihood of an engine failure than is fate. The word “random” generally means “with a uniform distribution.” In statistics, the “odds” of an event reflect the likelihood that the event will take place. And “eventually” is defined as occurring at an undetermined time in the future. If we exclude higher risk flying such as crop dusting, aerobatics, air racing, combat, other edge-of-the-envelope maneuvering and chronically poor maintenance, one would think that “eventually” the “odds” of having an engine failure would occur “randomly” across all pilots and that over time, we should all experience at least one. If only that were so.

Not-So-Evenly-Distributed
In 45-some thousand hours of flying time, my Northwest/Delta/Duke instructor friend has experienced 20 engine failures. Some of them were in GA and some were at the airlines. Some were in single-engine airplanes and some were in twins. Some were in his own, some in that of an employer or a client. Some piston, some jet. And of all the failures, the one this past summer was the first that ended in an off-airport landing. His engine seized during day-VMC, and fortunately, the airplane had enough energy to make it to a soybean field just short of the airport. His 20 engine failures in 45,000 hours caused me to make a quick comparison of his flying time to engine failure ratio and my own. 

In 25,000 hours, I’ve had four engine failures and all of them ended with on-airport landings with no damage. One was a recip and three were jets. The First Officers that I fly with have 10,000 to 15,000 hours, and most have had one or two engine failures, but many have had none. While I am a proponent of anthropomorphism and intuition, I’m not so much inclined to endorse unicorns, four-leaf clovers or fate. So, what the heck is going on with the engine failures? From the above piston engine vs. jet statistics, we know that recips, like the board game Mouse Trap, are a Rube Goldberg jumble of moving parts much more vulnerable to failure than a jet. So, let’s take a look at some recip components from a “Mike Busch” (a well-known engine guru) perspective to learn more. The turbine folks can skip to the end of the story, but you may gain some empathy (and sympathy) for us recip schmucks by reading on. 

Bottom End
The bottom end components of our piston aircraft engines – crankcase, crankshaft, camshaft, bearings, gears, oil pump, etc. – are very robust. They normally have a useful life that is many multiples of the TBO. 

Top End
The top end components – pistons, cylinders, valves, etc. – are considerably less robust than the bottom end. It is not unusual for top end components to fail prior to TBO. However, most of these failures can be prevented by regular inspections and use of a digital engine monitor. Most top end failures are random (there’s that word again) and do not correlate with TSMOH (time since major overhaul).

Crankshafts
Lycoming did a study that showed their crankshafts often remain in service for more than 14,000 hours (that’s seven-plus TBOs) and 50 years – no problem here.

Ruined cam and lifters from my Duke replaced with carbide tipped lifters and a new cam.

Camshafts and Lifters
Cam and lifter spalling is the number one reason that engines fail to make TBO, and it’s common in the owner-operator fleet (i.e., T &T readers) where aircraft tend to fly irregularly and sometimes sit for weeks at a time. Tiny corrosion pits can lead to rapid destruction (spalling) of the surfaces, sending metal flakes, or even chunks, into the oil filter and beyond. The good news is that this problem has been mostly negated by a friend of mine (Gary Bongard) who has a patent on carbide tipped lifters. He recently completed the arduous process of getting approvals from The Man and has begun manufacturing. I have his lifters in both motors on the Duke.

Bearings
Bearing failure is responsible for a significant number of catastrophic engine failures. Bearings fail prematurely for three reasons: They become contaminated with metal from some other failure (i.e., lifters), or they become oil-starved when oil pressure is lost; or main bearings become oil-starved because they shift in their crankcase supports to the point where their oil supply holes become misaligned. Contamination failures can be prevented by using a full-flow oil filter and inspecting the filter for metal at every oil change.

Connecting Rods
Connecting rods usually have a long useful life and are not normally replaced at overhaul. Many rod failures are caused by improper tightening of the rod cap bolts during engine assembly. Failures can also be caused by the rod bearings, usually due to oil starvation. 

Valves

It is quite common for exhaust valves and valve guides to develop problems well short of TBO. Failures are less common nowadays because problems can usually be detected by monitoring EGT’s on digital engine monitors. Even if a valve fails completely, the result is usually only partial power loss. I will add two caveats: There is an increasing failure rate of valves in some engines and one of our readers (good job and thanks for the story Pete) recently had a valve fail, which cascaded through multiple components and eventually caused both the turbo and the engine to fail. His was yet another engine failure example that didn’t create a data point because he landed on-airport with no damage. 

My on-airport landing after an engine failure occurred in poor weather conditions.

Suck Squeeze Bang Blow
(Turbine folks should rejoin us here).

Comparatively speaking, turbine motors are moving-parts-limited and have few failure modes. As long as none of the processes in the paragraph title – intake, compression, ignition/expansion and exhaust – are interrupted, turbines will serve us well. Just use good clean fuel, good clean oil, mind the limitations, don’t suck in anything but air, and it will run (almost) forever. The most frequent cause of jet engine failure is the ingestion of objects. This can cause both damage and rotor imbalance. On the ground is the most likely place this happens but birds, chunks of ice and airframe pieces are common causes while in flight. Also, any mechanical problem that causes a rotor imbalance can cause microscopic cracks to form on the turbine blades, leading to their failure. Other than that, it takes human intervention to kill a jet.

Never wait for trouble.
-Chuck Yeager

I recommend that you not practice engine failures during takeoff in the airplane. Save it for the simulator because any realistic takeoff-failure scenario in the airplane would be dangerous. Using a zero-thrust power setting once above 3,000 or 4,000 feet, and with an instructor, however, is valuable training. Make sure the surprise factor is there. Practice failures during a turn on the SID, at some point halfway to altitude during a distraction, and one while at cruise. These maneuvers should not be considered complete until the engine is (simulated) secured, the airport of intended landing has been selected, and the route to that airport and the approach to be flown have been loaded. Practice flying the airplane at zero-thrust while talking to ATC (your instructor) and loading/programming your GPS/FMS/FMC. Then in the sim, practice the approach, landing and single-engine go-around.

Motors Don’t Abide by The Statistics
An FO I flew with described the astonishing sight of a dissipating hurricane working its way up the Eastern seaboard as they overflew the system. I never gave much thought to the weather below me while traveling from A to B until my precautionary shutdown in the Duke when I had to land in crap weather on one motor. Make sure you consider an engine failure when flight planning – and not just for your departure and destination airports. Poor engineering or maintenance, metal fatigue, fuel contamination, bad luck, probabilities, four-leaf clovers or fate. Many variables are in play as we consider if and when we will have a loss of power, a precautionary shutdown or a failure. These events are supposed to occur randomly and only every 3,000 or 4,000 hours. But apparently, motors don’t abide by the statistics. Ernie Gann may have been right.