Alaska 261: 'A Maintenance Accident'

Overhaul & Maintenance

March, 2003

NTSB board member John Goglia summed up the January 2000 crash of Alaska Airlines Flight 261 in five simple words: ''This,'' he wrote in a submission included in the board's report on the accident, ''is a maintenance accident.'' His fellow board members agreed, and in January they adopted the report, which cited ''Alaska Airlines' insufficient lubrication of the jackscrew assembly'' as the main factor that caused the part to fail on the MD-83, leading to the plane's fateful plunge into the Pacific Ocean.

The probable cause was no surprise to those following the probe, as problems with jackscrew maintenance and Alaska's maintenance program came out long before the report. The bigger news came in 16 recommendations to FAA adopted by NTSB as part of the final document. While many of them deal with the DC-9/MD-80 jackscrew specifically (and therefore have limited industry-wide ramifications), a few cast doubt on some widely adopted maintenance practices.

NTSB urged FAA to review ''all existing maintenance intervals'' and make sure ''adequate'' technical data exists for any intervals that have been extended beyond original manufacturer recommendations, and that each interval has an ''adequate safety margin'' to account for missed tasks. Going forward, NTSB wants FAA to ensure that adequate data is taken into account when intervals are changed. When such changes are made, NTSB wants operators to supply FAA with ''technical data and analysis'' supporting each change.

The recommendations stem from what investigators determined were the key links in the chain of events leading to the crash of Flight 261. Basically, FAA allowed Alaska to extend maintenance intervals on various tasks, including entire MD-83 C check packages. At the time of the accident, the C check interval was 15 months. During every other C check, the airline did an ''end-play check'' on the jackscrew assembly to ensure the assembly wasn't worn out. The airline didn't -- and wasn't required to -- record any data from these checks that would have helped track wear rates of the jackscrews. So Alaska had no way of knowing that Flight 261's jackscrew was wearing more quickly than normal as a result of inadequate lubrication. And although the airline submitted some data to FAA to justify the C check extension, it did not -- nor was it required to -- provide data for each of the hundreds of items included in the C check.

In other words, somebody made a mistake (inadequate lubrication of the jackscrew), and the systems in place (Alaska's and FAA's) weren't robust enough to catch the mistake before it led to a catastrophe. The airline knew when its MD-83 jackscrews were worn out, but not how worn any one of them was at any given time during the interval between end-play checks.

Alaska also extended jackscrew lubrication intervals, with FAA approval. The MD-83 interval went from every 500 flight hours in 1987 to every eight months (roughly 2,550 flight hours, but no flight-hour limit was specified) in 1996. Investigators never determined what data, if any, Alaska provided FAA to justify the extensions.

''The safety board is concerned that the absence of any significant maintenance history pertaining to the jackscrew assembly was apparently considered by Alaska Airlines and the FAA as sufficient justification to extend the end play check interval as part of the C-check interval extension,'' the board said in its report. ''In general, the absence of maintenance history should not be considered adequate justification to extend the interval for the performance of a critical maintenance task. Any significant maintenance change associated with a critical flight control system should be independently analyzed and supported by technical data demonstrating that the proposed change will not present a potential hazard.

''The safety board is further concerned that the MSG and MRB-based process by which manufacturers develop initial and revised recommended maintenance task intervals resulted in significant extensions of both the lubrication and end play check intervals without any such analysis or support,'' the report continued. NTSB also cited McDonnell Douglas for not designing a fail-safe mechanism that would have prevented ''the catastrophic effects of total acme nut thread loss.''

Goglia, a former airline mechanic, agreed that a better design wouldn't hurt, but insisted that the human element of accidents like Flight 261 must be stressed to prevent similar events from happening. ''Certainly we should not rely solely on maintenance where mechanical design can compensate for maintenance failure Yet it is a seductive error to think that we can engineer out failure permanently. We can't. Fail-safe components subject to neglect and abuse are themselves susceptible to failure.

''This,'' Goglia continued, ''was not a design accident. This was a maintenance accident, perhaps more purely so than any others we have seen.''

While some may see Goglia's concerns as an over-reaction, a look at recent accidents suggests a disturbing trend. There have been four fatal crashes involving U.S. Part 121 carriers in the past three years going back to the Alaska accident. Probes into two of them -- Alaska 261 and the February 2000 crash of an Emery DC-8 near Sacramento -- turned up problematic maintenance, while the probe of the Jan. 8 crash of the US Airways Express (Air Midwest) Flight 5481, a Beech 1900, at Charlotte was looking closely at that plane's maintenance history.