The Statistical Orphan of Aviation Safety

Statistics abound on aircraft incidents and accidents, except for “public use” fixed-wing and rotary-wing flying machines. We are left with anecdotal reports, such as the number of accidents in a period of time, but nothing like events per so many flying hours. Without such a denominator, the comparative record of public use operations cannot be usefully compared to the rest of the aviation industry.

This is just another way in which public use operations are an orphan, with no effective oversight by the Federal Aviation Administration (FAA), which has indicated it has “no intention” of changing the state of affairs. Moreover, there are no equivalent statistics maintained by the National Transportation Safety Board (NTSB) on lives lost and mangled aluminum.

Just to refresh memories, public use operations involve contract flights on behalf of federal and local government entities. For example, the “water bombers” flying in support of U.S. Forest Service firefighting efforts are in the unregulated public use category. Another example: helicopters flying for local entities as the Maryland State Police.

If one asks the NTSB for the number of public use accidents, one will get a stack of accident summaries, from which one can glean 17 dead and five accidents since 2007 involving public use aircraft flying for the Forest Service. To be sure, this is an horrific record, reflecting the price paid by the FAA’s hands-off attitude.

How this toll compares to the other main sectors of aviation remains unknown. What is needed for the public use category is an accident rate per 100,000 flight hours and/or per 1,000,000 miles flown.

Review the NTSB statistics (www.ntsb.gov/aviation/stats.htm). The searcher will look in vain for any information regarding the public use category. Yet there are ample, detailed statistics on every other category of aviation, presented in the 12 tables.

For Part 121, scheduled airlines: data as far back as 1990 on the number of accidents, fatalities, accidents per 100,000 departures, per 100,000 flight hours, per 1,000,000 miles flown, and more.

For Part 135, non-scheduled air taxi and on demand operations, less detailed information but still accidents per 100,000 flight hours.

For Part 91, general aviation, a real eye-opener on the mayhem in this category – such as 1,474 accidents in 2009 (the most recent year data is presented) with 474 people aboard killed. There is an average of 1.3 deaths a day in general aviation. More importantly, there is data on the accidents per 100,000 flight hours.

 Hence, useful comparisons between Part 121, Part 135 and Part 91 flying are easily made:

ratescompared

For public use operations? Nothing. No table. No data. No comparisons possible. I am willing to bet that public use operations have an accident rate somewhere above that of Part 135 and below that of Part 91. For instance, the “water bombers” flying for the Forest Service are generally being flown in old, converted military airplanes, with sketchy maintenance records and a substantial amount of metal fatigue and corrosion. These firefighting aircraft are operating in areas of high turbulence (updrafts from fires), which puts a strain on their structure (wings have been breaking off). Flying at low altitude in these conditions is not as safe as an airliner benignly cruising at 30,000 feet, but how much less safe has firefighting flying been? Statistically, we have no idea.

One cannot make policy decisions without good data. One cannot focus regulatory resources in the absence of data.

Senior FAA officials proclaim that their priorities will be “data driven.” In the absence of data about public use operations, it is not surprising that the agency ascribes a non-priority to oversight of public use aircraft and helicopter operations.

All operators of public use aircraft maintain records of flying hours per machine. The information is necessary because many maintenance actions are linked to accumulated flying hours.

The NTSB can provide the accident data, as reporting of even less-than-fatal accidents is required.

The hours flown divided by the number of accidents can be used to derive the rate per 100,000 flying hours. This information can be displayed on another line in the table above for public use. My wager is that public use operations will have a fatal accident rate 10 times greater than for scheduled airline operations. Is this an acceptable rate for the FAA to continue its regulatory neglect? Or would such a tenfold rate galvanize FAA officials into adopting the public use orphan now sitting on the curb, ignored by the government as a no-data non-entity?