On May 5, 1983, Eastern Airlines flight 855 departed Miami International Airport (MIA) bound for Nassau, Bahamas. The Lockheed L1011, registration number N334EA, was almost twenty minutes out of Miami when the #2 engine low oil pressure light illuminated. The Captain shut down the engine and made the decision to return to MIA.
As the aircraft turned back to Miami, it was climbing through 20,000 feet, when both the #1 and the #3 engines’ low oil pressure lights illuminated. Ten minutes later the #3 engine flamed out; the aircraft was now flying on one engine and still fifteen minutes from MIA; #3 engine’s bearings had turned to molten slag. The crew attempted to restart the #2 engine, but the #1 engine flamed out, its bearings seized. The aircraft descended from 13,000 feet to 4,000 feet without power, not unlike a glider. At 4,000 feet the crew restarted the #2 engine and the L1011 limped back to MIA on one engine.
Losing one engine was normal training for the crew; though not common, an engine-out was not outside the realm of possibility. Even a two engine-out scenario was a training scenario, but it was less likely. All three engines out, even with the same symptom – an engine low level light preceding a flameout – was statistically impossible. And yet, apparently, it was not.
The emergency was brought about by the absence of magnetic (MAG) plug chip detector O-rings, normally installed with the MAG plug in an engine’s oil system following a MAG plug check. This check required the MAG plug be removed from the engine, the magnetic probe checked for metal contamination and then reinstalled with a new O-ring to seal the plug to the engine port. A MAG plug check was (and still is) a normal maintenance item accomplished on the engine manufacturer’s dictated schedule. The three engines had a scheduled MAG plug check performed. The oil in each engine was depleted during the short flight time before the crew turned back for MIA.
All three engines were worked on by two mechanics; one worked on numbers one and three engines and the other worked the number two. The MAG plug kits the two mechanics used on N334EA contained the replacement MAG plugs, but without the O-rings. The mechanics, and other mechanics in Miami, usually received the kits, complete with the O-rings. This was a contributing cause of the accident.
The mechanics, accustomed to having complete MAG plug kits never verified that the O-rings were on the plugs before installing them. They then performed a leak check for the engines to confirm that the plug installations did not allow oil to bypass the MAG plugs. However, the time that the mechanics spent on ‘motoring’ the engines was inadequate: ten seconds; the oil did not have enough time to flow through the system to where the MAG plugs were. A second contributing factor to the accident.
There were these two contributors and one root cause that led to the accident flight. The first contributor: the mechanics’ failure to check the kits for the O-rings. The mechanics had been lulled into a false sense of security by having complete kits handed to them. Both mechanics were still responsible to verify the O-rings were included in the kits and installed with the plugs. The rut the mechanics fell into was complacency. When they performed a redundant task so many times, they assumed they knew the job by heart, e.g. replacing a tire or servicing oxygen. Due to modifications or even accidents, maintenance manuals are revised with different torques or an extra step to prevent a problem already identified. It is the mechanic’s responsibility to make sure the job is done right. No matter how often a job is performed, the manual must be checked for changes and a component for complete parts.
The second contributor: the inadequate amount of time spent ‘motoring’ the engine. The job card should have identified the motoring time needed for a proper engine leak check. From my experience running engines, you can motor an engine for at least two minutes without starting it; after that time the starter could be damaged. Two minutes was enough time for a proper leak check. Another issue was gate time, as in ‘the time before the aircraft is to be on the gate for the next flight’. Rushing through a job to meet an ETC – [pronounced ETIC] (Estimated Time – Completed) – to make a flight departure time, could push the safety limits by employing shortcuts.
The root cause, however, was culture. Had this plug problem happened at other air carriers? Delta, Pan Am, TWA and other carriers flew the L1011; did they have problems with incomplete kits or destroying an engine due to oil starvation? What part did Maintenance Planning play? Management? The Engine Manufacturer, Rolls Royce?
Or did this only happen at Eastern Airlines? Did the culture at Eastern or perhaps only at the Eastern Miami maintenance facility, contribute to the confusion about MAG plug kits? Did the National Transportation Safety Board (NTSB) go to the Eastern maintenance facilities in Kennedy airport, Atlanta or Boston to determine if the problem was limited to Miami? No? They did not? Why not?
Culture issues; they have been contributors to many accidents since 1983. Culture issues have plagued mergers over the last few decades. Culture problems have not been limited to Maintenance but have emerged in Operations (Asiana 214), with cargo contractors (Fine Air 101) and maintenance providers (ValuJet 592). Yet with these and many other accidents, culture issues were never mentioned in accident reports; indeed, culture has never even been looked at as a viable issue. Was it because culture issues never existed or was it because culture issues were never thought to exist?
With Eastern 855, universally accepted complacency was the Eastern Miami maintenance facility’s five hundred-pound gorilla, the elephant in the room. The mechanics’ failure to be diligent and the time-related practice of inadequate leak checks were traceable to the Eastern Miami maintenance facility’s culture. To list all the wrongs that led to Eastern 855’s near ditching at sea would take more than one article could handle; Eastern and the 172 souls aboard flight 855 dodged a major bullet. If other air operators had learned what there was to learn from Eastern 855, many more lives could have been saved.
Randy Pausch, professor of computer science at Carnegie Mellon University, once said about the ‘elephant’ metaphorical idiom, “When there’s an elephant in the room, introduce him.” The industry, e.g. airlines, repair stations, manufacturers, etc. know there are cultural problems in their ranks – they know this. Unlike human factor issues, these are indigenous to the certificate holder and to the groups within.
Safety conversations with Federal Aviation Administration inspectors could result in an unbiased view of an operation’s culture, undetectable from the inside. An outside auditor can identify recognizable patterns that lead to major mistakes. Relying on management to unearth any culture problems would be a dead end when they could be part of the problem. If a certificate holder expects the NTSB to help, they usually show up in connection with an accident. It is unlikely the NTSB would find the culture problem.
Culture issues are the Lessons Unlearned that have led to numerous accidents since May 5, 1983. Although rarely recognized, culture problems have been major contributors to accidents; the root cause overlooked in favor of probable cause. We should not have to wait another thirty-six years for someone to point it out.