Obscuring Safety Hazards

Since 1970, more than 240 incidents involving jet engine malfunctions — some of them catastrophic — have been recorded. The event search was based on the key words, “shutdown, vibrate/vibration, bangs, loud, seized, contained, uncontained” or “bearings.” Oil and fuel systems issues, as well as bird strikes to engines, were discarded. In other words, only the most serious reports involving internal mechanical failure were considered. Suffice to say, problems run from the inlet to the hot section to the exhaust ducting.

The actual total is probably higher, but the Federal Aviation Administration (FAA) does not enforce any discipline regarding reporting. What reports do trickle in are the heart of “identified safety risks.” The risks, such as represented in the FAA’s incomplete data base, are well short of the actual frequency of incidents.

And this is just for engines.

The term “identified safety risks” is part and parcel of the FAA’s bureaucratic language used to transmit corrective actions — in the form of airworthiness directives — to the airlines. These directives mandate corrective action, usually on a very generous timeline so as not to inconvenience flight schedules.

From various and sundry airworthiness directives, “identified safety risks” encompass virtually every aspect of the airplane:

— Uncontained engine explosions that compromise continued safe flight. Not only loss of thrust, but shrapnel damage to hydraulic and electrical systems, leading to a cascading series of failures and cockpit crew confusion and overload.

— In-flight fires in inaccessible areas unprotected by fire suppression. Every square inch of public buildings is required to have fire detection and suppression; not so for the most crowded of public spaces, a modern jetliner.

— Electrical wire insulation that exceeded smoke, toxicity and flammability standards. The running of low-power circuits in the same bundles as high-power circuits aggravated the problem, as high-power arcing has led to overload of low power circuits, with resulting fire and explosions. The lack of separation and segregation standards for electrical circuits continues to this day; when electrical arcing occurs, the result is usually a sequence of cascading failures.

— Flammable thermal/acoustic insulation blankets installed throughout the fuselage for fire burn-through “protection.” The use of flammable insulation for burn-through protection remains a sick irony. Yet insulating material that resists the hottest fire is available, but continues to be unrecognized by the FAA and not used by the industry, either on newly-manufactured aircraft or retrofitted onto older aircraft during periodic overhauls.

— Electrical equipment cabinets that allow faulty arcing components therein to spew molten metal, thus enabling an otherwise contained electrical fire to transition to an uncontained fuselage insulation fire.

— Electrical heater tapes, used throughout all airframes for freeze protection, with faults that led to an uncontainable fuselage insulation fire, raging just a few feet below the passenger cabin floor. Following two Air Canada fire investigations by the Transportation Safety Board (TSB) of Canada, Boeing said, “Between June 1985 and June 2002, operators of Boeing aircraft made a total of 67 reports to Boeing of heater ribbon failures where thermal degradation was evident.” Note the wording: “reports to Boeing.” The FAA was clueless regarding the extent of heater-tape induced conflagrations.

Damage to a B767 airliner from a faulty heater tape and the resultant damage to insulation blankets

Damage to a B767 airliner from a faulty heater tape and
the resultant damage to insulation blankets

— Uncommanded rudder movements, or dangerous rudder deflections from one side to another, which contributed to at least three fatal crashes and an unknown number of incidents in flight from which a startled crew was able to retrieve the situation.

— Continued incidents of flight into freezing rain/freezing drizzle that aircraft are not certified to cope with, and for which the FAA has been egregiously lax in setting a standard. In-flight upsets from what is known as “supercooled liquid droplets” (SLD) continue to occur. In SLD, the water slaps onto the airfoil and instantly freezes. Unlike the larger Boeing and Airbus jets, with their much larger engines to supply hot air for anti-icing, the smaller regional jets have a lesser ability to resist icing conditions.

— Cockpit smoke impairing the pilots’ view of the instruments and of the outside world. The FAA and the National Transportation Safety Board own up to one such event, but a search of various databases yields at least six events.

These and other hazards persist because of “firewalls” that hinder recognition of safety hazards. There are many such firewalls, many of them self-imposed by regulatory bureaucrats, but here are five indicating the magnitude of the problem:

1. The Accident/Incident database maintained by the National Transportation Safety Board (NTSB) contains only 20% (or thereabouts) of these hazards, which are seen in such incident reports as the FAA’s Service Difficulty Report (SDR) database or counterpart agency reports.

2. The FAA’s Incident database is even more abysmally incomplete than the one maintained by the NTSB. For example, of some 67 uncontained engine explosions involving GE engines, the FAA Incident database captured just one. The gaping gaps in the FAA’s database makes one wonder: what is the purpose of “maintaining” such a record; to use a Swiss cheese metaphor, the holes are greater than the amount of cheese. For analytical purposes, the Incident database is utterly useless.

3. The SDR data itself is woefully incomplete. Although required by FAA regulations, some airlines comply, reporting gaps at other airlines range from 70% compliance to as little as 30%. The FAA does not enforce reporting compliance, so the SDR database is incomplete, and many reports that are submitted have missing data or “apparent causes”. Moreover, SDR’s are only required on incidents that occur in the air. Ground events are not part of the reporting mandate. The industry successfully lobbied to force the FAA to withdraw a proposal to include SDR-reportable events that occur while the airplane is on the tarmac. The SDR system is in need of a complete overhaul and stiff penalties for not reporting. Moreover, the reports must be systematically analyzed for trends by aircraft type and among airlines; the SDR database in its present form is scandalously incomplete and gathering dust as a “data morgue”, not rigorously analyzed to identify emerging trends.

4. Identified safety hazards remain hidden from public view with the manufacturers’ use of Service Bulletins (SB’s) and All Operator Alerts (AOL’s). FAA-issued airworthiness directives — available to the public — often refer to AOL’s and SB’s. In fact, AOL’s and SB’s often contain gritty details, which are referred to in only the vaguest terms in the airworthiness directives. The situation is analogous to reading about the efficacy, or lack thereof, of a medicine by reading the label on the bottle. The protocols, clinical trials and scientific studies remain behind lock and key. Suffice to say, published airworthiness directives, which reference these proprietary documents, make their associations to prior events and to safety trends simply impossible for members of the general public.

5. Certain FAA studies (on fuselage flaws, on uncontained engine explosions, on electrical wiring) also are not available to the public, thus obscuring trends and/or non-safety associations. The disclaimer on a study of the flammability of in-service materials is typical: “In general, data are proprietary to the applicant, and we cannot disclose those data to the public.” Here’s another regarding rotating engine parts: “The FAA proposes to designate the … data submitted to them to create the Rotor Manufacturing Induced Database (known as the ‘ROMAN database’) as protected from disclosure … under the Freedom of Information Act (5 USC552) and other laws. The FAA wants to encourage … suppliers that manufacture high energy rotating gas turbine engine components to voluntarily submit information into the ROMAN database.” Of course, voluntary reporting is likely to lead to as porous and incomplete information in ROMAN as in the “required” SDR database.

The question is why this state of affairs exists.

For one thing, costs to the industry are lower because AOLs and SBs, unless mentioned in FAA-issued airworthiness directives, require no mandatory action. Similarly, by not requiring reporting of all SDR data — occurrences both in the air and on the ground –the FAA lowers the cost to the industry of submitting reports.

To be sure, making available incomplete and selective data for public consumption gives a false impression of safety. The average John Q. Public is probably not aware of the appalling database incompleteness. Certainly the association between incidents involving injury and accidents involving deaths is obscured. Thorough SDR and Incident databases might reveal plenty of precursor problems leading to a fatal accident. Defining a fatal loss as an unfortunate “one time” occurrence lessens the role of industry in these losses, as it also lessens the role of the FAA’s sloppy certification system in such losses.

In sum, costs to the industry are minimized and facilitated by the use of partial, incomplete and limited public access to identified risks and related events.

By allowing this system of data truancy and industry-friendly policies, one has to ask: who is the FAA’s customer — the favored airlines and manufacturers, or the flying public, which falsely believes the feds really practice the doctrine of safety first.