We read accident reports involving sudden, catastrophic engine failures and propellers that fling themselves from crankshafts or landing gear that won’t extend and sometimes wake in the wee morning hours wondering if we’re completely foolish to be operating machines that rise up off the ground. Are we completely at the mercy of maintenance technicians who happen to be human and make mistakes?
Um, well, to some extent, yes. There are maintenance errors that do not get caught and lead to mechanical failures at the worst possible time. However, in looking at accident reports, we’ve observed that in a surprising number of situations, a pilot/owner was in a position to catch the mistake before flying, initiated the problem by not having needed maintenance performed, or, uncomfortably, approved maintenance that wasn’t needed and increased the risk of a failure.
Numbers
We’ll take a look at the numbers on maintenance-related accidents and how a thoughtful owner can reduce the risk of having one.
The good news is that there aren’t a lot of accidents due to mechanical failures. The not so good news is that they are split about 30/70, according to our research, between maintenance that should have been done and wasn’t and mistakes made in the process of maintenance. Of the maintenance-induced accidents, a frightening number happen on the first flight out of maintenance, and many could have been prevented, or the seriousness reduced, had the pilot been focused on the risks of first flights after maintenance.
McSpadden Report
Every year we review the newest McSpadden Report and look at its analysis of a year’s worth of aircraft accident data. The report is prepared and published by the Air Safety Institute of AOPA and was known for years as the Nall Report before being renamed for Richard McSpadden, longtime director of the Institute.
Over the last several years the McSpadden Report has reported that “mechanical” was the cause of on the order of 16 percent of nonfatal accidents and 7 percent of fatal. It also has shown “unknown” as the cause of about 14 percent of nonfatal and 27 percent of fatal. For reference, the rest of the accidents were placed at the pilot’s feet, usually somewhere in the 70 percent range overall.
We’re uncomfortable with the high percentage of accidents listed as unknown. Over the last few years, we’ve had conversations with some of upper management at the NTSB that grew out of problems we experienced with the new CAROL NTSB accident database search engine. Our experience was excellent, changes were made and the search engine was tweaked to where we think it’s one of the best in the business for drilling down deeply into aircraft accident data.
A side benefit was that our conversations expanded well beyond the original topic into our concern that about 15 percent of engine power loss accidents were classified as unknown by the NTSB.
To our surprise, NTSB management said that they agreed with us and were already working on digging deeper into investigations of engine stoppage events. We’ve seen some recent accident reports that indicate to us that more effort is being spent on engine power loss accidents. Frankly, when almost half of engine power loss accidents put the cause as unknown, we feel like we’re playing Russian Roulette with our engines.
Our Data
That is background to our own data on accident causation that we develop every month after reading the 100 most recent accident reports on an airplane type for the Used Aircraft Guide. When we see an engine power loss accident where the NTSB could not find the cause, we generally place it in the “engine/mechanical” column. An engine needs spark, fuel and oxygen to run. A pilot can mess up and deprive it of one item, but we consider that unlikely in cases where it was reported that the engine was running happily until it lost some or all power and it couldn’t be restored prior to landing. Accordingly, when we find an engine power loss event that was not related to running out of fuel, we stick it in the engine/mechanical column and consider it to be a maintenance-related event unless we can show otherwise.
That also means our numbers for maintenance-related accidents may be high. If so, that’s good news for us as consumers of aircraft maintenance and probably not so good news as pilots who may have screwed up, but we didn’t catch the pilot’s error and erroneously blamed maintenance.
The average we found after a review of 210 aircraft accident reports, fatal and nonfatal, was 16.9 percent were due to “engine/mechanical” causes. To that we added accidents involving mechanical problems with retractable landing gear—it would not extend in whole or in part, or it collapsed when on the ground. They averaged, for retractable-gear airplanes, 4.4 percent—we consider those to be maintenance-related, usually due to maintenance not performed, although for some airplanes such as 300-series Cessnas, it was often due to maintenance improperly performed.
We also noted, to our surprise, that for the Piper Seneca, there were eight crashes involving flight control failures, four because the stabilator was not installed correctly and one because the electric trim was hooked up backward.
Our bottom line is to consider that a maintenance issue was a direct cause of on the order of 18 percent of all general aviation aircraft accidents.
Praise for Techs
The first thing we’ll say after that conclusion is that this article is not to dump on aircraft maintenance technicians. They are some of the most highly trained maintenance personnel in any industry and we’ve found that only a tiny percentage don’t give a damn about their work (they tend not to stick around) or have mental issues that make them actively hostile to flight safety and willing to sabotage an aircraft. The majority are doing their level best, day in and day out, frequently in unpleasant conditions, amid toxic substances and often cramped, poorly lit, dirty workspaces.
Mistakes
We feel that is essential to recognize that maintenance mistakes are just that, mistakes. Mistakes are made by every single human being, no matter how intelligent. The causes are numerous and include distraction, poorly lighted workplaces (under the panel of an airplane), cramped workplaces (tailcone of an airplane), confusing instructions, lack of experience with a task, boredom, poorly labeled components/equipment/supplies, anger, indifference, hostile workplace environment—you name it.
What can owners do?
As owners who have aircraft maintained there are things that we can do to reduce the risk of maintenance-related errors and accidents by recognizing the causes of mistakes. There are little things that add up; notably, we can minimize distractions. Don’t make it a regular practice to walk into the shop and start talking to a tech who is engrossed in a task, especially if it’s related to your airplane. Your distraction may just be the event that gets an error chain started that ends with your off-airport landing.
Don’t try to hurry repairs. Yes, we’ve run into techs who take months to do an annual. However, they are usually well known, so don’t have them do work for you. When your aircraft is in the shop, communicate regularly, but at times when you are not interrupting work, and don’t push for a fast turnaround. The industrial engineers who study aircraft maintenance errors put time pressure on techs right up there among the top causes. As an aside, pressuring someone to hurry can also generate resentment that leads to errors. Treat techs as respected professionals.
One frequent example we’ve seen of accidents caused by hurrying techs is landing gear malfunctions in Cessna twins with the electro-mechanical gear. It’s reliable, but only if maintained per the Service Manual. Gear rigging, done right, takes two people who know what they are doing about eight hours and has to be done from the center working out on all three gear legs. Attempting to shortcut the process is a near guarantee that the gear will collapse or fail to operate.
Humans make more mistakes when doing new tasks. Accordingly, if you’re having an STC on your airplane, paying for the learning curve for a shop to do it the first time involves paying extra for the time it takes as well as the increased level of risk of a mistake.
The same goes for your type of airplane—have work done by a shop that regularly works on your type. Got an oddball or an antique? Part of the cost of ownership may be flying a thousand miles for your annuals.
Your Expertise
Even if your sole mechanical gene was surgically removed when you were eight, join the type club for your airplane and immerse yourself in information regarding its care and feeding. Having some specialized knowledge or knowing where to find specialized information on your airplane that you can communicate with your tech can save you money and reduce the risk of maintenance error.
Maintenance not needed
Once a tech has troubleshot a problem or given you a squawk list on an annual inspection, go through each item on the list with the tech. The discussion should start with whether it’s an airworthiness item. If it is, it needs to be fixed—there’s no further discussion unless it is a big-ticket item that you are reasonably certain the shop is foisting on you—at which point you get a second opinion. Fortunately, that’s rare.
If it’s not airworthiness, do some homework. For example, if the maintenance guidelines call for cleaning fuel injector nozzles, that’s one you probably shouldn’t have done. From what we’ve observed, those nozzles are self-cleaning and if they get plugged, it’s likely that it was due to maintenance performed on them.
The idea is to have repairs done on condition, avoiding work that is not needed and can result in a maintenance-induced accident. That means having as much knowledge as you can about your airplane and communicating well with your maintenance tech so you can make decisions on what work is needed.
Cylinders
We are especially concerned about cylinder replacement. Our recommendation is that it should only be done when absolutely necessary because we see so many accidents caused by improper cylinder installation—usually improper torque of the retaining hardware.
Low compression, by itself, is never a satisfactory reason to replace a cylinder. A borescope examination is essential. Lycoming and Continental have good guidance on when it is appropriate to replace a cylinder, and the homework that should be done in making the decision.
A cylinder replacement too often is a time bomb waiting to cause a catastrophic engine failure. Because of that, we opine that it is almost never appropriate to do a “top overhaul.” The cylinders are little engines operating in their own environment and they don’t make each other sick. The more cylinders you replace in the field, the more potential for having a time bomb turning your propeller.
As we call for on-condition maintenance, we also recognize that there are life-limited components that truly do need repair or replacement at certain times in service. The big one is magnetos. They have plastic parts that wear out predictably. We think that the widely made recommendation to pull them for inspection and repair or replacement every 500 hours time in service is most appropriate.
Maintenance Not Done
That leads to the next issue—we see a large proportion of accidents that are due to maintenance not performed. Too often an owner will defer too much needed maintenance. In a fatal accident in a Beech 36 Bonanza that resulted from a leaking O-ring in the fuel control assembly we noted that the area had not been touched in 26 years.
We see prop failure accidents where the propeller had never been serviced in 20 years. Things wear out, and pilots have a reputation for being tightwads. Owning an airplane is expensive—there’s a difference between making intelligent decisions as to what work should be done and cutting every corner possible.
Some of the most glaring examples of deadly accidents due to maintenance that is not performed that we’ve observed over the years have been in piston twins. They can be staggeringly expensive to operate, so some are astonishingly cheap to buy and attract owners who don’t realize what they are getting into.
Notable examples are the Beech Duke, Cessna 320 and Cessna 411. They are high-performance twins that were cutting-edge in terms of design when they came out. However, those cutting-edge systems that allowed high performance were usually not terribly robust and needed regular inspection and expensive replacements—notably the exhaust systems on the 320 and 411.
Bottom line—if you own an airplane where you are cutting corners on maintenance problems that can kill you, don’t mess around. Sell it. Now.
We have looked at accidents where fuel systems were so full of debris, rust particles and general gunk (at Aviation Consumer we only use the most precise technical terms) that filters or fuel injection units plugged up and stopped fuel flow to the engine. Those conditions didn’t develop quickly.
Some single-engine Cessnas left the factory with a plug rather than a quick drain in the low point of the fuel system—under the belly of the airplane. Those were to be removed for inspection at each annual. Of course there have been accidents where that wasn’t done over a few years and fuel flow was blocked. In our opinion, a better practice is to replace the plug with a quick drain and make draining it a part of the preflight inspection.
Conclusion
We don’t think that it’s possible to eliminate human error in aircraft maintenance. However, with a degree of thoughtfulness, good communication with your shop and a very careful first flight preflight, the risks can be reduced.
