On May 11, 1996, a ValuJet DC-9-32 aircraft, registration number N904VJ, crashed in the Florida Everglades near Miami, Florida; a tragic end to a standard flight. Oxygen generators in a Class D cargo compartment started a fire, which engulfed flammable items, including an aircraft tire. The crew tried to get in front of fire-caused electrical and mechanical failures while the passengers were stuck in a metal tube filled with smoke. It is unsure that anyone survived to impact in the Everglades. The National Transportation Safety Board (NTSB) accident report AAR-97/06 (DCA96MA054) lists several probable causes of the accident in a rare two-paragraph form – more than necessary to answer the question: Why did ValuJet 592 crash?
The probable cause (PC) first paragraph states: the accident “resulted from the airplane’s Class D cargo compartment that was initiated by the actuation of one or more oxygen generators being improperly carried as cargo…” Wait … What? The bullets to this statement are that Sabretech (ValuJet’s repair station) had improperly labeled, packaged and loaded the oxygen generators on the aircraft; the oxygen generators were not made safe and should have been considered hazardous material. ValuJet failed to properly oversee Sabretech and that the Federal Aviation Administration (FAA) failed to require smoke detection and fire suppression systems in Class D cargo compartments.
The second PC paragraph stated that “contributing to the accident” was the FAA’s failure to monitor ValuJet’s and Sabretech’s programs; the FAA’s breakdown of adequately responding to previous oxygen generator fires with programs to address these issues. Last, that ValuJet failed to ensure Sabretech was aware of the ‘no-carry’ policy and provide proper hazardous training. THESE are THE probable causes of the accident. To those experienced in commercial airline operations, NTSB report AAR-97/06 reversed the significance of the two paragraphs, placing Operational Error second. Why?
To understand the probable cause’s first paragraph, Class D cargo compartments, like that found on the DC-9-32, have been used in jet aircraft since the Boeing 707, i.e. 1957, almost forty years prior. They are not containerized; the freight is loaded by hand, separated by netting and other restraint devices from movement that causes center of gravity shifts. Cargo compartments don’t cause fires; if used as designed, they could not be the ValuJet 592 accident’s PC. Neither the Class D cargo hold, the tire or the oxygen generators led to the accident; they were symptoms of a bigger problem: the Root Cause.
The term: probable cause, is a misnomer; it is a band-aid. Probable, as defined by Oxford Dictionary, is: “likely to be the case or to happen”. With accident investigation involving months and years worth of testing and analysis, the travelling public deserves more than “this is likely to be the” cause. Discovering Root cause is much more effective; it is laser sighting. An analogy: Break a weed on the surface (probable cause), the weed disappears for a while, but eventually returns. To kill the weed, you dig the root out of the ground (root cause) completely; that’s how one stops the problem permanently.
The Class D cargo compartment, the oxygen generators and the tires were coincidences; three innocuous items brought together by Operational mistakes that led to disaster. The worst kind of mistakes, they come from ignoring rules and policies, i.e. Operational Error. THAT is why ValuJet 592 crashed.
Report AAR 97/06 supported its PC/Class D theory by referencing, first, a Saudi Arabian Airlines L-1011 which caught fire in 1980 in the C-3 cargo compartment during flight. The source of fire was not determined, which begs the question: Why was it referenced?
The second example occurred February 3, 1988; American Airlines Flight 132, a DC-9-32, had an in-flight fire on final approach; the fire was started by a hydrogen peroxide solution (an oxidizer) and sodium orthosilicate in close proximity. The combined chemicals didn’t burn, but, the report suggests, ignited combustibles in the luggage. The root cause: the chemicals were not labeled as Hazardous, which was strictly Operational Error. The accident report AAR-88/02, however, cited in four Recommendations about cargo compartments, focusing away from the Hazardous materials and Operational errors.
The decision to improve Class D compartment integrity was a good direction to take; any moves to improve safety are always welcome. Shipping deflated aircraft tires is not dangerous nor is the shipment of oxygen generators. Class D compartments should have been upgraded to make them safer, not because the Class D compartment ‘caused’ the accident. Citing the Class D as the cause, probable or otherwise, distracted from the real culprit: a total disrespect for procedure, e.g. shipping hazardous materials in a cargo bay not rated for hazardous materials. The issue became skewed.
Why the NTSB went in this direction was puzzling; logically, it made no sense. It is like blaming the iceberg for the Titanic disaster; the iceberg did not hunt down and jump in front of the Titanic; the ship sunk due to human error in design (unsealed watertight compartments), in manufacture (inferior rivet materials) and in safety (moving too fast for available visibility). A similar perpetrator that caused Titanic’s sinking was responsible for ValuJet 592: Culture.
A word search was conducted on report AAR-97/06, looking for the word, ‘culture’; it came up one time, in an FAA letter to ValuJet. But culture was the one concept that escaped this NTSB investigation. Why? Because investigators MUST be experienced in commercial aviation. How many investigators had/have worked for air operators or repair stations? How many investigators had the career experience to relate to what was happening at ValuJet in 1996? How many investigators have never: met a flight time; worked with competing cultures; dealt with union issues; experienced FAA inspector oversight; deferred a component; worked multiple flights; midnight shifts; got pressured from management; met a ‘hard down’ airplane; screaming flight planners; aircraft-on-ground (AOG) situations; weather; deicing; working with contractors and being solely responsible for an aircraft’s airworthiness? In NTSB accident meetings I attended, culture issues were a foreign concept. Then seven years after ValuJet 592, my NTSB maintenance investigatory group saw similar culture issues with the Air Midwest 5481 accident.
Why is culture so critical to ValuJet 592’s tragedy? Because the accident happened during a time the contract maintenance provider (CMP) became an industry norm. For years airlines accomplished their own maintenance or worked closely with other airlines with established facilities. However, with the rise of low-cost airlines, e.g. Peoples, Air Tran, Southwest, and international expansions into foreign markets, overhead costs associated with maintenance were first to be targeted, e.g. manning, employee benefits, pay and heavy maintenance equipment; business expenses that could be transferred to the CMP.
While the NTSB did not know CMPs existed, the FAA had limited CMP experience, grasped even less about outsourcing. This was further confused by the expanding Regional airline (RA) contracts. Industry gave more responsibility to RAs and CMPs with no vested interest in the air operator. The contractors learned expensive lessons about how far an airline would go to save money. ValuJet’s CMPs outsourcing practices created problems with training and authorizations to conduct work. ValuJet’s oversight of its contractors suffered, adding layers for the FAA to keep tabs on, all while ValuJet bought more aircraft and expanded. This is how shipping oxygen generators as company materials got through the net. Multiple cultures that repelled or clashed with each other, obscuring problematic issues.
Culture should have played a pivotal role in AAR-97-06’s probable cause because the FAA allowed ValuJet, a low-cost carrier, to get away from them. ValuJet’s place in the market grew … at an irresponsible rate. The airline exploited a good market, its popularity accelerated expansion. ValuJet’s failure to oversee its contractors masked the FAA’s ability to oversee, especially outsourced maintenance. If these red flags were mentioned in AAR-97/06, could the FAA have recognized similar problems with Air Midwest 5481 or Emery 17? We will never know.
It is encouraging to know that one accident can prevent another, but only if the issues are recognized the first time. The cycle should have been broken, repeat problems addressed and prevented. We have seen/are seeing accidents that repeat themselves. This just makes missed probable causes all the more frustrating.