© 2003 Mark E. J. Fay

What Really Brought Down Flight 592?

Within days of ValuJet Flight 592’s May 11, 1996, disaster, the Federal Aviation Administration (FAA) did some very unusual things. Those things, and others, lead me to believe that: The oxygen generators did not start the fire; An unapproved part started the fire, and; The FAA conspired to cover up this information.

ValuJet Flight 592, a McDonnell Douglas DC-9-32, crashed near Miami and claimed 110 lives. The National Transportation Safety Board (NTSB), in accident report #DCA96MA054, stated that fire was initiated by one or more oxygen generators loaded in the forward cargo compartment. The report implies that the loose-load oxygen generators, without safety caps installed, actuated from being jostled around. The oxygen generators undoubtedly intensified the fire, but consider the possibility that the oxygen generators did not start the fire.

There is evidence strongly suggesting the FAA, immediately after the accident, suspected a different ignition source and took steps to cover it up, or at least reduce the impact if that different source were to become known.

A day or two after the accident, the deputy director of Douglas Aircraft Company’s spares division gave an assignment to a staff specialist. The deputy director told the specialist that the FAA, on or about May 12, 1996, had asked American Airlines personnel at Dallas-Fort Worth Airport (DFW) how they procured primary longitudinal trim relays, part number 9207-10296, used on DC-9-32 airplanes.

American Airlines was a maintenance contractor to ValuJet at DFW. Under Federal Law, that particular aircraft replacement part, at that time, could be sold for installation on a DC-9-32 only by Douglas Aircraft Company. Records showed that Douglas Aircraft Company had never sold that part number. The deputy director instructed the specialist to help the relay manufacturer, Leach International, convince the FAA the relay sold to the airline was essentially identical to the relay Douglas Aircraft Company installed on new aircraft at production. In their discussion, Leach informed the specialist there was an FAA mandated maintenance action, an Airworthiness Directive (AD), applicable to the relay.

Airworthiness Directive AD 91-21-07 (and Douglas Aircraft Company Alert Service Bulletin A27-316, dated January 4, 1991, incorporated into law by reference in the AD) applied to the ill-fated DC-9-32. The AD states, “To eliminate overheating of the primary longitudinal trim relays and the possibility of fire in the forward cargo compartment, accomplish the following.”

The requirement was to replace the relays prior to the accumulation of 8,000 flight hours, and thereafter at intervals not to exceed 8,000 flight hours. Flight 592 had 68,400.7 total flight hours at the time of the accident according to the accident report. Were the relays replaced within the period mandated? Even if the relays were replaced as required, could one “overheat” in less than 8,000 hours?

The Alert Service Bulletin was even more damning. Its Reason paragraph states, “Four operators reported six instances of overheat failure of the primary longitudinal trim contactors (relays). Contactor failures have resulted in smoke and burning behind the right side cargo liner at the forward side of the forward cargo door. In some instances smoke was reported in the main passenger cabin.”

The Reason statement ends, “In one case an intercostal (metal aircraft structure) was damaged and required replacement.” Parentheses are mine.

Leach also told the specialist the FAA was pushing Leach, immediately after the accident, to obtain FAA Parts Manufacturer Approval (PMA) on the relay. With PMA Leach could legally sell the relay directly to airlines. Leach obtained PMA May 29, 1996 (Ref. PMA No. PQ2059NM, dated May 29, 1996). The FAA granted PMA based on a licensing agreement betweeen Leach and Douglas Aircraft Company dated May 22, 1996.

Obtaining PMA within 18 days of the accident was a minor miracle. It was especially noteworthy because Douglas provided the PMA assistance (license) within 10 working days of the accident, a process that ordinarily took months. And, the FAA granted PMA within three working days of Leach’s application.

I can vouch for these things because I was the Douglas Aircraft Company spares specialist. If the primary longitudinal trim relay did not ignite the oxygen generators, why did the FAA focus on ensuring that that specific part number relay received FAA-PMA virtually instantaneously after the crash? There are other relay part numbers in the AD, but the only part number the FAA pushed Leach to apply for PMA was 9207-10296 – the part number of the relays installed in the doomed airplane.

A cursory understanding of the worldwide aircraft replacement parts business is needed to understand the import of the foregoing. There is a relatively constant annual turnover of $10B in that business. Unapproved, bogus, or illegal parts comprise several billion dollars, or more, of that annual trade. The FAA has been struggling for years to control it. The last thing the FAA needed was for 110 deaths to be attributed to an unapproved part. By ensuring Leach obtained PMA on the part, the FAA anticipated the question, if it were asked, “Is the primary longitudinal trim relay FAA approved?” The FAA could truthfully respond, “Yes it is.” Only a skeptic would rephrase the question, “Was the relay FAA approved at the time of the crash?”

There was a powerful champion in the crusade to eliminate unapproved aircraft parts. The FAA’s parent organization is the Cabinet level Department of Transportation (DOT). The Inspector General of DOT when 592 crashed was the young, ambitious, and zealous Mary Schiavo.

She was outspoken. She declared that virtually all airplanes were flying with bogus parts installed. She and her team prosecuted numerous cases involving unapproved parts. Millions of dollars in fines and years of incarceration were meted out to unapproved parts miscreants under her leadership.

One of her reports quotes a convicted illegal drug dealer as saying, “I quit dealing in illegal drugs and got into the illegal aircraft parts business because there’s more money to be made and you meet a better class of people,” or words to that effect. She cited problems at ValuJet months before the accident, and refused to fly them for safety reasons.

She spoke out about the FAA’s failings. Ms. Schiavo left government service and “resigned” her post just shy of two months after the crash of Flight 592. Even though she stated no one pressured her to resign, the FAA must have been greatly relieved she was no longer liable to investigate a certain unapproved part.

Additional reasons to suspect the relays started fire aboard Flight 592 include the following.

*The NTSB assumes the oxygen generators actuated as they rolled around the cargo compartment when the firing pins struck the percussion caps. Tests at Douglas Aircraft Company facilities in Southern California failed to prove the oxygen generators could self-actuate. With great difficulty, and under conditions unlikely to be encountered in flight, the oxygen generators eventually actuated, according to a colleague who participated. Those tests are not cited in the NTSB report.

*Did the primary longitudinal trim relay generate sufficient heat to actuate the oxygen generators? If not directly, did they ignite passenger cargo in the forward cargo compartment, which in turn ignited the oxygen generators?

*The report speculates that the flight crew attempted to reduce thrust on both engines for emergency descent. The right engine went to flight idle about six minutes into the flight. The left engine did not respond to inputs to reduce thrust, resulting in asymmetric thrust.

There were no right-roll indications in the Flight Data Recorder heading data. The airplane was in a sideslip condition, most likely with the left wing down. This would cause the cargo to shift to the left, supporting the possibility the fire started, not where heat damage was most pronounced, but to the right – the location of the longitudinal trim relays.

*In static tests the FAA pulled the pins and actuated the oxygen generators. Two of the five tests resulted in no fire. (Ref. NTSB report, pp. 54-55)

*The NTSB’s report cites “…heat damaged wire bundles were not routed near the breached area of the cargo compartment, whereas the boxes containing the oxygen generators were loaded into the area directly beneath the breached area of the cargo compartment.” (Ref. NTSB accident report, pp 100-101.)

This begs the question, “Were the heat damaged wire bundles those routed, near the relays, ‘…behind the right side cargo liner…’ as cited in Alert Service Bulletin A27-316?”

The FAA knew there was an unapproved part on Flight 592, believed it might have caused the fire, and hoped to cover up that information.

Is the FAA guilty of aiding and abetting the unapproved aircraft replacement parts industry?