It has been a good year, but the figures mask insidious threats

Unless there are two or more dramatic air accidents in the last two weeks of December 2003 is going to be the best in commercial air transport history. There were fewer hull loss accidents and fatal accidents by mid-December than in any year since 1984, and 20 years ago the number of flights was roughly half what it is now.

But a study of the categories of accident that continued to happen in 2003 is worrying, and one category in particular - fatal accidents directly or partially caused by maintenance error - could bode ill. Also, controlled flight into terrain (CFIT) continued to be one of the two main causes of the year's fatal accidents. In 2002, CFIT accidents represented the largest single fatal accident category, while in the year just ending it has been - so far - the second largest. Unusually, the largest category, has been technical failure. At least two of the technical failure events are already known to have been caused by maintenance error, and others may have been - only the final accident reports will resolve these.

But safety figures at 1 December show a continuation - even an accentuation - of the steady downward drift in accident numbers and rates over the past three years. The trouble with single-year figures as an indicator in the global aviation industry is that the numbers are so small that a single major accident can transform the year's statistics. From one year to another the airlines' measured safety performance can be reversed dramatically, as they were in 1984-5. The safest year in air transport history at that time was immediately followed by the most disastrous.

But because the overall trends have sloped remorselessly toward better safety performance - particularly in accident rates - ever since records began, it is tempting to believe that, with the exception of a few annual spikes in the graph, it will continue that way - and it could. The question the industry has been asking ever since the end of the 1980s is: when will the downward trend hit the law of diminishing returns, then flatten out for the foreseeable future? By about 1989, the belief was that safety could not get much better, and in the first few years of the 1990s it looked to be true. But then the steady improvement resumed in 1997 and it has continued ever since.

During the 1990s, technology was beginning to produce safety returns for the airlines, both directly through improving reliability, and indirectly through diagnostic maintenance systems. Increased use of ground proximity warning systems (GPWS) have had a significant effect, and finally the increased availability of hard data from quick access recorders and digital flight data recorders has enabled the industry to become more focused in identifying and tackling the underlying causes of accidents.

Challenges identified

The industry began to be able to identify its challenges and, led by the Flight Safety Foundation, set up working groups to study accident category data in detail and began to come up with solutions or strategies.

Meanwhile, the enhanced GPWS (EGPWS) has virtually terrain-proofed a large percentage of the world's heavy airline fleet, mostly in North America and western Europe, where it is compulsory, but also for many major carriers elsewhere. No aircraft fitted with EGPWS has ever been involved in a CFIT accident, and that includes inadvertent collisions with flat terrain when aircraft get too low on a runway approach without the crew realising it.

If these events - which can be disastrous - occur close to the runway's end they are usually known as undershooting, and are often not thought of as CFIT although the generic circumstances and causal factors are identical. But they are CFIT, and the EGPWS has stopped those too.

The combined revelation that no EGPWS-fitted aircraft has ever had a CFIT accident, and that CFIT has continued to be the main cause of air accident fatalities over the last two years, contains the clue as to where future improvement may come from: the unequipped fleet which remains. Meanwhile, all new aircraft manufactured by Airbus and Boeing are fitted as standard with EGPWS - now generically known as a terrain awareness warning system (TAWS), so gradually the proportion of the world's TAWS-equipped large jet fleet is increasing.

But there is more to be done. Research by Honeywell's Don Bateman, known in the industry as the father of EGPWS, shows that the CFIT risk to the western European and North American large airliner fleet is now down to one CFIT accident per 91 million flights and still reducing, whereas the figures for the world excluding those two regions is one in 16 million and "stagnant".

Bateman says that the standard GPWS and the TAWS have together been a major factor in reducing CFIT as a cause of fatality to a 50th of the risk that existed in 1991 and 100th of the risk in 1975. But, he adds, "the risk has hardly changed in many parts of Asia, South America and Africa since 1975".

The other big cause of fatal accidents in 2003 has been technical failure of components. In two cases - both US-registered Raytheon Beech 1900Ds - the cause was mis-rigging of the pitch control system. The mistake in each case was not identical. The two aircraft were in different airlines, and the second event happened more than six months after the first, which had by then been publicised by the National Transportation Safety Board.

Maintenance cuts

It is just enough to stir a concern that maintenance investment cuts at this time of airline financial hardship may be influencing safety. The same concern might arise over pilot training quality. Two Boeing 737 fatal accidents followed an engine failure, but if an engine failure is found to have been the only causal factor, the question would arise as to why the pilots failed to maintain control of the aircraft.

The industry should look at the kind of accidents that happened in 2003 and wonder whether something is going wrong. Because it would be a pity if, after such a commendably safe year overall, the airlines were beginning to lose their concentration in a couple of vital areas.

For a more detailed accident analysis, see Flight International's annual world airline safety review to be published in the 20-26 January 2004 issue

Source: Flight International