In this new feature, we take a deep dive into an NTSB Accident Report, Worn ignition switch and key leads to fatal unintended engine start, giving you the full narrative from the NTSB docket, with analysis of the accident and lessons that can be learned from Capt. Brian Schiff, a captain for a major U.S. airline with more than 20,000 hours who holds several flight instructor ratings.
From the NTSB Investigation:
On July 26, 2018, about 5 p.m eastern daylight time, a Cessna 182P pilot died when he was struck by the propeller during a preflight inspection of his airplane at the Cleveland Regional Jetport (KRZR) in Tennessee.
The pilot’s wife reported they had flown to KRZR earlier that day, and her husband performed a normal shutdown of the engine using the mixture control. The couple ran some errands, then returned to the airport.
While her husband performed his preflight inspection with the airplane’s key in his pocket, his wife was standing outside the airplane behind the passenger door facing her seat. She heard the “propeller move,” which she described as unusual, and said the engine sounded like it was starting or trying to start. She looked up and saw her husband fall to the ground. She went to him, then went inside the FBO to get help.
According to FAA and airport personnel, airport security video did not capture the accident.
The 1976 Cessna 182, which was not damaged during the accident, was powered by a 230-hp Continental O-470-S engine equipped with two independent engine-driven magnetos electrically controlled by an on-condition, key actuated, five position rotary ignition switch mounted on the lower left portion of the instrument panel.
An examination of the cockpit by FAA investigators revealed that the mixture control was in the idle-cutoff position, the throttle was full out, and the battery and alternator rocker switches were in the off position.
From the pilot’s perspective, the ignition switch appeared to be in the off position. However, on closer inspection, the ignition switch was more toward the right magneto position.
When investigators inserted the key, the switch was able to be turned counterclockwise to the OFF position.
The investigator also noted that the ignition key was easily removable from the ignition switch in the right and left positions.
Subsequent testing of the ignition system was performed by a mechanic. The ignition switch was marked where it was found after the accident (between the off and right positions), and in that position, the right magneto was hot, or not grounded, and the key could be fully removed from the ignition switch that was tightly secured to the instrument panel.
When in the off position, the switch did not align with the instrument panel placard marking.
The pilot bought the airplane in April 2007. According to his wife, there was no previous instance in which the key could be removed from the ignition switch in an intermediate position or any position other than the off position. She also indicated that since owning the airplane, there was no previous issue of the key coming out, falling out, or being loose.
Cessna Service Bulletin SEB91-5 Revision 1, dated June 14, 1991, specified an inspection and lubrication of the ignition switch, however there was no requirement to determine if the key could be removed from any position other than off.
Review of the airplane service manual that describes inspection procedures at 50 hours, 100 hours, 200 hours, and special inspection items, and review of 14 CFR Part 43 Appendix D, which documents the scope and detail of items to be inspected in the annual or 100-hour inspections, revealed no specific mention to check if the key could be removed from any position other than off.
Airworthiness Directive (AD) 93-05-06, applicable to ACS Products Company and Gerdes Products Company ignition switches and the pilot’s Cessna 182 by serial number, specified an internal inspection for corrosion, lubrication of the ignition switch, and to determine if a diode or other surge suppresser was installed on the starter solenoid. The AD was required to be completed within 100 hours after the effective date of the AD (April 29, 1993), or at the next annual inspection, whichever occurred first.
The AD did not specify a test to determine the integrity of the key to the switch, or to determine if the key could be removed from any position other than off.
The maintenance records reflect that the AD was originally signed off as being completed on July 27, 1993, at airframe total time of 2,880.7 hours. The next compliance with the AD as part of an annual inspection was on Jan. 10, 2018, at airframe total time 4,895 hours.
Based on pilot records, excluding the flight earlier that day, the airplane had been operated about 31 hours since the inspection was performed. There was no entry in the airframe maintenance records indicating replacement of the ignition switch or key.
The mechanic who performed the annual inspection and complied with the AD said he also installed new internal ignition switch contacts and terminal board supplied in kit A3650-2. He also reportedly checked the ignition key to switch integrity in all switch positions other than off, reporting no discrepancies.
Additionally, as part of his procedures, while the engine was running, he slightly pulled aft on the key to see if it could be removed from any position and also in between positions. The key did not come out in any position other than off.
According to the NTSB Materials Laboratory factual report, examination of the cut surfaces of the key notches revealed relatively smooth and reflective surface features consistent with a worn surface on the flank of the notch adjacent to the key retention ridge. A smooth and reflective surface consistent with wear was also observed on the tip end of the shank opposite the notched side.
Examination of the key cylinder revealed a quadrilateral-shaped area with a smooth and reflective surface consistent with wear on the lower side of the key slot.
The location and shape of the worn area was consistent with wear contact with the tip of the key as it was inserted and removed.
The ignition and key were also sent to Gerdes Product Company for examination.
Testing of the switch, which was manufactured in April 1976, revealed that the key was removable in every position (left, right, both, start, and off). In the as-found position, testing verified the key could be removed and the right magneto was not grounded.
The key tumbler was removed but company personnel were unable to determine why the key could be removed from the left, right, both, and start positions.
A review of the NTSB’s database for previous accidents and incidents in which a loss of ignition switch-to-key integrity led to unintended engine startup revealed three previous investigations. In all three, the key was not in the ignition switch, which was believed to be in the off position, but the engine started, which resulted in injury or uncommanded movement of the airplane.
Probable cause: The undetected wear of the ignition switch and key, which allowed removal of the key from an intermediate position and subsequently led to an unintended engine start-up. Contributing to the undetected wear of the 42-year-old ignition switch was the lack of guidance by the switch manufacturer and airframe manufacturer for procedures to detect lack of integrity between the ignition key and switch.
NTSB Identification: ERA18LA199
Lessons Learned
“This accident is one that could happen to anybody. I think that’s what makes it so attention grabbing — and it’s one that we can learn from,” says Schiff.
According to Schiff, the accident is a wake up call that there is “one more thing we need to look at” while flying.
When the pilot moved the propeller, the engine fired and the propeller hit him, Schiff says, noting NTSB investigators found that the key wasn’t in the off position.
“It was aligned so that it was kind of between off and the right magneto. In that position, the P-lead was not grounded, so the magneto was hot. So if you move the propeller, there is ignition, and fuel and the cylinders would fire at that point. And they did, and it caught the pilot and sadly killed him.”
“Also, they found that the key hole in the ignition was worn to the point where the key could be pulled out, even though the key wasn’t turned all the way to the off position,” he continues. “I don’t know who hasn’t flown a Cessna where you can pull the key out with the engine running in just about any position because the latch in there has worn so much. And Cessnas get a lot of wear on those key holes.”
Drilling down, Schiff says there are two things in play in this accident that are mechanical and one that is pilot related.
The mechanical issues: The first is that the key could come out with the key not actually turned all the way to the off position. And secondly, the ignition switch was installed in the panel with the off position not aligned with the off placard on the panel.
The pilot issues: The pilot moved the propeller blade in a direction of firing and it fired.
“Which, incidentally could happen, even if the mags are grounded,” he says. “If you just shut down and have a hot engine and you go out there to move the propeller, you could fire a cylinder if there’s enough fuel and enough heat in that engine, which is actually why we shut an engine down with a mixture.”
How Can Pilots Protect Themselves?
There are several steps pilots can take to protect themselves in this kind of situation.
“I’m an advocate of checking the mags before shutdown. I think there’s nothing wrong with that. I’ve talked to my mechanic and I’ve got an O-360 engine. Maybe it depends on your engine, so I would talk to your individual mechanic. I would hate to say that it’s okay for any engine to do that, but it’s important to do it right.”
It’s more than just doing a magneto check, he cautions.
“You need to make sure that you’re grounding the leads. And really all it involves is prior to shutdown, at idle rpm, just position the ignition switch to off and as soon as you hear that the rpm is clearly decreasing, return the switch to both. If you do it fast enough, you’re not going to have enough fuel built up in there. The problem is if you do it too slowly, you’re going to have fuel buildup in there. And then you go back and turn the ignition on, you’re going to have a pretty good backfire. That’s not good. So if it’s done quickly, it’s okay.”
“You also need to be careful not to go too aggressive back to the on and re-engage or crash engage the starter or back to the start position. You want to go back to the both position. But if you go to off and the engine continues to run, then at least one of the magnetos is still on. And that’s a dangerous situation because while you think your engine has shut down and the P-leads are grounded and the mags are off, one is really not grounded and it’s hot. Any movement of the propeller could fire the engine.”
Don’t Move the Propeller
Schiff advises pilots to not move the prop unless you have to.
“If you’re just doing it every time you park because you like to see it horizontal, and you’re a little bit retentive about things like that, maybe get rid of that habit. If there’s no need to move it, don’t move it.”
“If you need to get it out of the way for a tow bar or something like that, that’s different,” he says. “There are some times when we need to move the propeller.”
In those instances, move the propeller backwards, he advises.
“Move it opposite the direction that it rotates,” he explains. “If you’re sitting in the cockpit and viewing the propeller, it turns clockwise. So if you’re outside of the aircraft looking at the propeller as you’re facing aft, turn it clockwise as well.”
“And if you can’t remember, just take a look at the propellor blade and remember that it wants to take a bite out of the air,” he continues. “So you want to turn it in the direction where it cannot take a bite out of the air.”
Increase Your Safety
Schiff notes that pilots can increase their safety by creating some redundancies or implementing habits that serve as another safety barrier or a safety net when something goes wrong.
For example, he advises pilots to have a hand on the throttle during climb out. “In case it slides back, that’s just one more barrier,” he says.
Always put your keys on the glare shield so everybody around knows the key is not in the ignition.
And always assume the magnetos are hot.
“Yes, we can shut them down. And that’s one barrier, but let’s assume that that barrier is not there. Let’s say that we just assume they’re always hot. So treat the airplane like those magnetos are always hot. Minimize prop movement to only when necessary. And when you do have to move it, of course, move it backwards.”
Do the post flight magneto check to ensure you are actually grounding the P-leads when you switch your ignition to off.
“If it doesn’t do that, then get it fixed,” he says. “If your key can be pulled out in other than the off position, have it repaired.”
Learning From the Misfortunes of Others
Pilots can learn a lot from reading NTSB accident reports, according to Schiff.
But he warns that you shouldn’t go to the mainstream media or a bunch of pilots hangar flying at the airport. Go to the source: The NTSB website and search the dockets for the accident.
“All you really need to know is the date it happened. And you can search by all kinds of different search parameters and find the docket.”
The docket has a wealth of information, including the ATC transcript, radar ground track, reports from those involved and those investigating the accident, and photos.
“It might even have a deep dive study on where the sun was at this point in an accident. The sun was this many degrees above the horizon. And there’s so many really, really comprehensive NTSB studies that are completed to recreate an accident and to come up with a scenario and find what really happened. If you really want to study an accident, you need to get into that.”
He notes that in this accident, there are pictures in the NTSB investigation that clearly show where the ignition switch was positioned after the accident, as well as the position of the throttle and mixtures. You can also see a view of the plane surrounded by police tape.
“People often say, ‘well maybe they had the mixture in…’ Well, you can suppose all you want, or you can go right to the NTSB docket and see the actual scenario. You can learn exactly how the airplane was configured when the accident happened. And that tells you a lot more about what happened.
“And we all want to know, as pilots, what happened in an accident. Sadly, you want to learn from what they did. We need to at least not let them die in vain, to learn from what happened to them so that this will not happen to anybody else. That’s why pilots have the innate desire to learn from the accidents of others. And it’s incumbent upon them to learn how to do so.”
Want to know more? Schiff has a lot of resources on his website, CaptainSchiff.com, including information from seminars he’s given, “cheat sheets” for students that simplify explanations of complex aeronautical concepts and regulations, and much more.
Check out a video of the conversation between Captain Brian Schiff and General Aviation News Publisher Ben Sclair about this accident below:
Old parts wear out and fail. Some are more critical than others.
So, the AD requires lube on the contacts and a suppressor diode, but no check for key removal in any position other than ‘off’. ?
I do the mags-off test at every shut down, to make sure that P-leads are grounded.
I also do a ‘left’ mag, ‘right’ mag check at shutdown to be sure both mags are good.
Only then do I pull the mixture to ‘idle-cutoff’.
I also to a ‘left’ mag, ‘right’ mag check after engine start.
I got to the run up area on one flight to find the left mag dead, and then had to taxi back to my hangar,
and got to work removing the left mag and send it off for rebuild [ to zero time ].
Sad that happened. Personally my shutdown procedure is to set COMM to 121.5 and listen for ELT, in case my landing was less than a “greaser”. Then with engine at idle (700 rpm or so) I shut off both mags and hear/see engine shut off, then switch both on and continue with normal shut down (I fly a 7KCAB, no keyed switch, just toggles). I discovered a broken P lead while doing this after a flight. This is in my shut down/post flight check list.
This has actually happened to me. The key was on top of the dash and the engine started.
Working for a small school the aircraft were parked outside. In cold weather it was my object to sit the student at the controls and pull the blades through to limber the oil and prime the cylinders. The key was visibly on top of the dash.
I pulled the first blade through, no problem. Pulled the second blade through I felt a small kick back. Didn’t think to much about it. Key was visible! Told the student to keep feet on brakes. Pulled the third blade through, and the engine roared into life. The key was on the dash.
Quite shaken I found the original pilot had not turned the key right off. Just pulled it out after putting the mixture to idle cut off. The magnetos were left hot.
A worn rotary ignition switch was the problem.
Could have been quite disastrous for me. I now give propellers a great degree of respect, and consider them hot at all times.
One of the biggest contributors to worn keys is the habit of putting heavy objects on the key ring. I tell any customer of mine to NOT do that. The extra weight in combo with the aircraft vibrations cause extra wear on the key tumblers. If you need to I.D. your key, Only add a single lightweight tag.
I have a new respect for my ignition switch and I will check it before my next flight. My prop always stops in a position that blocks my tow bar. I will continue to turn it “backwards” slowly and stand clear for good measure. Thanks for the info.
A lot of butt-covering in this article. The FAA, mechanics, manufacturer all scrambling to escape blame. The real problem though is the AGE of this aircraft. This is what you get when you try to cheap out in aviation flying antique machines.
It’s “maintenence” not antique machines. Most of us don’t have $250,000 to spend on a late model 172.
Gently!! My Wifes ‘76 Citabria, my ‘56 PA22/20 and most of the other aircraft I’ve been flying over the last 52 years are hardly “cheap”. As an A&PIA with about about 7,000 hrs of military and GA flt time it is my experience that it takes learned safety skills On behalf of both the pilot and the mechanic to break the safety chain of events that lead to an accident. MOST GA aircraft are safe and airworthy. You may find broken P leads or other deficiencies on ANY aircraft. The key is to find the flaws BEFORE they bite you . . When I do an annual inspection I am signing for the fact that that aircraft meets it type certificate as designed (plus any legally applied alterations) AND at the time of that inspection it is safe for flight. ..oh yeah. I’m also a USN post grad school trained safety officer… It CAN and WILL happen even in your 2021 Cirrus. IMHO the only machine worse than a tired ‘old’ one is a brand new one… NEW often abbreviates Never Ever Worked… lol.
Humor aside, treat every prop as HOT. The backwards turned vacuum pump doesn’t apply to what most of us fly. Turn it backwards but only if you have to turn it at all.
One more bit. I have had starters engage when I turned the master switch on!! Be careful there as well…. even if you are just preflighting the fuel gages….🙄
The article is about “learning from the misfortunes of others”, your “new” airplane is not immune to having maintenance issues.
When I first read this NTSB report some time back the first thing I did was check the ignition switch on my C-170B to make sure the key could only be removed in the off position. Also reviewed my procedure for moving the propeller, and even decided it was time to overhaul the mag switch on my J3 Cub.
Pulling mixture lean past peak RPM then shutdown using the ignition, same as cars, verifies mags off and would have prevented the incident.
Also, paying attention, worn tumblers in switches and locks are always obvious and not uncommon in aircraft.
Handle accordingly or replace.
If memory serves I believe at least vaccum pump manufacturers discourage rotating the engine backwards as it can damage the vanes.
A better header may be ….things happen to the unwary.
Not having heard of this being an issue with mechanics perhaps ratings and PIC hours have little to do with it.
Actually it seems more and more GA crashes lately had ATPs or CFIs at the controls…