THE YEAR 1996 SAW the largest number both of airline fatal accidents and of fatalities on record. Other serious worries for the air-transport community highlighted by 1996 include the number of deaths on the ground caused by crashes - also the worst ever - and some compelling trends indicating that increased attention needs to be paid to the safety of freighter operations.
Although the total number of fatal accidents, at 57, was only one more than that for the previous year, that 1995 figure had been the highest up to that point. Also significant is the fact that 20 of the crashes involved freight, positioning , or other non-passenger airline operations, a larger number than usual. Meanwhile fatalities, at 1,840, have exceeded the previous highest total of 1,801, which occurred in 1985.
These increases confirm an upward trend in simple totals, which became established with the 1995 figures (Flight International, 17-23 January, 1996, P24), even though trends for rates (risk levels) during the 1990s do not appear to be worsening.
These figures exclude those from fatal events resulting from hijacking and other forms of illegal interference with flight. With the "illegal-interference" category included, however, 1985 remains the worst year on record by a considerable margin, because of the massive number of casualties (329) from the sabotage of the Air India Boeing 747 which crashed into the Atlantic near Ireland in June that year. The only fatal security-related event in 1996 was that of the Ethiopian Airlines Boeing 767 hijack, which ended with an out-of-fuel ditching near the Comoros Islands (see tables), killing 128 passengers and crew. The comparison between the two worst years, using the all-categories annual totals, therefore shows 2,230 fatalities in 1985, while the equivalent 1996 total is 1,967.
Attempts to put the 1996 figures in context provide little consolation. The 1985 figures were a dramatic "spike" in the graph (exaggerated by the fact that 1984 had produced the still-standing lowest-ever fatal-accidents/fatalities totals), and 1996 will probably also be a "spike". Nevertheless, last year's 1,840 fatalities in57 accidents contrast badly with the 1986-95 decade annual averages of 1,084 and 44.
Now, however, the updated ten-year average (1987-96) has increased dramatically because figures for a safe year - 1986 - drop out and the 1996 figures go in (see chart). This puts the new decade averages at 1,207 fatalities and 47 fatal accidents a year.
According to International Civil Aviation Organisation figures, however, traffic in 1985 was at only about 66% of today's levels. So, if the 1985 accident figures were factored to take into account the traffic increase between 1985 and 1996, the statistical extrapolation would give 2,729 fatalities and 59 fatal accidents for last year. Comparing commercial air transport's two worst years, therefore, the statistical risk to passengers was lower in 1996 than in 1985.
The industry, however, has no inclination to crow about relative safety. All parties, from manufacturers to airlines, maintain that public perception of airline safety is based on absolute figures. Rates, it is repeated time and again, do not affect public perception: absolute totals do, and 1996 has been a disaster for the industry in this respect.
About 1.5 billion scheduled air journeys were made in 1996 and, as the real price of airline tickets progressively decreases and disposable incomes rise, air travel has the potential to reach an increasingly wider market. As this happens, popular-media interest in all aspects of air travel, particularly accidents, will increase.
Another world-record was broken disastrously in 1996. The total number of people on the ground who were killed by airline crashes was by far the highest on record, at 364, with serious injuries believed to approach the same number. About 300 of the fatalities and 250 of the serious injuries were caused by one event: the 8 January abandoned take-off at Kinshasa, Zaire (see accident tables). The Scibe Airlift Antonov An-32 ran off the end of Kinshasa's runway and ploughed 600m through a busy shantytown marketplace. This was the worst third-party accident in history. It is too late, at Kinshasa and at many other of the world's airports, to argue that residential development of any kind should not have been allowed to take place close to the end of runways.
The worst third-party-casualty accidents seem consistently to be associated with freight aircraft. In 1996, after the Scibe Airlift disaster, the next worst was the October Millon Air Boeing 707 freighter crash at Manta, Ecuador, which killed 30 and injured double that. Third in the grim league comes the February Lineas Aereas del Caribe Douglas DC-8 freighter accident after take-off at Asuncion, Paraguay, in which 20 died on the ground.
The previous worst third-party accident in peacetime, according to Airclaims' World Airline Accident Survey, was a December 1966 Da Nang, Vietnam crash involving yet another freighter, a Flying Tigers Canadair CL-44. While on a radar-assisted approach in fog, the aircraft came down about 1.5km short of the runway killing 107 people on the ground.
Another freighter horror was the more recent (September 1992) El Al Israel Airlines 747 crash near Amsterdam, Netherlands, in which the aircraft hit two blocks of flats after a pair of its engines had separated from one wing. Although grim media speculation at the time forecast hundreds of casualties in these high-density dwellings, the final total was estimated to have been fewer than 50 fatalities.
This "freighter factor" needs examining. As in the Amsterdam El Al case, nearly all the 1990s in-flight separations of wing-mounted engines have involved freighters: 747s at Taiwan (China Air Lines, 1991) and at Anchorage, Alaska (Evergreen International, 1993) as well as the Amsterdam event, and a 707 which was diverted to Istres, France (Trans Air Services, 1992), with only its Nos 1 and 2 engines still attached and working. Only the Taiwan and Amsterdam accidents were fatal, however.
The UK Civil Aviation Authority's new Accident Analysis Group (AAG), tasked with identifying worldwide trends, has confirmed that the accident rate for freighters is "far higher" than for passenger aircraft. This is "one of the trends" which the AAG may study.
Meanwhile, the world's traditional low achievers in safety have again confirmed their position as such, but more dramatically than usual. Out of the 57 fatal accidents, 45 involved operators registered in nations which can either be described as "Third World" or as "developing economies". The latter description differentiates them from the "mature economies".
The significance of this is that the industry sector which carries about 12% of the world's airline traffic has produced 80% of the fatal accidents in 1996. In fatalities terms ,the figures look less bad: the "low-achiever" sector produced 37% of the on-board casualties.
The "high-achiever" sector, however, suffered some high-casualty accidents in 1996, including those of ValuJet (110 deaths), Trans World Airlines (230) and Saudi Arabian Airlines (312). North America and the Middle East are statistically two of the world's safest regions, but they have suffered painful reminders that there is no such thing as immunity from accidents.
As usual, there is such a diversity of accident causes that trends are difficult to identify. Another unsurprising truth is that the causes known so far are familiar ones.
Some, in particular, exemplify the potential results of careless departure from the most basic standard procedures. The Birgenair and AeroPeru 757 disasters, for example, were both caused by blockages of parts of their pitot-static systems, which disabled flight instruments.
Neither accident need have happened, say investigators, if time-honoured, basic, standard safety routines had been respected. The official report of the Birgenair accident concludes that the pitot covers had not been fitted during long-term parking of the aircraft at Puerto Plata, and one of the tubes had become blocked. The AeroPeru aircraft, according to the investigators who have examined the recovered wreckage, still had protective tape stuck over the static vents. It had been put there by maintenance crews who were cleaning the hull, and had not been removed either by them, or spotted by the line-maintenance or flightcrews during pre-departure walk-around checks.
More complex events
Other events, such as the still-unexplained TWA 747 accident, may prove more complex, both in the causal chain of events and in their repercussions for the industry. If the US National Transportation Safety Board's (NTSB) worst fears about the aircraft's centre-wing fuel-tank explosion are realised, however, then "empty" fuel-tanks (filled only with fuel vapour) may be more vulnerable, under certain circumstances, to ignition than had previously been appreciated; or perhaps than had previously been admitted.
The NTSB, quoting three previous incidents involving fuel-tank explosions (one in a 747), has recommended to the US Federal Aviation Administration that certain measures should be adopted - some immediately - to reduce the "vulnerability" of fuel-vapour-filled tanks to even low-power ignition sources (Flight International, 1-7 January, P9). The implications of these recommendations, whether or not they turn out to have direct relevance to the TWA accident cause, will be considered carefully by the whole air-transport industry.
The year provided its ironies. When the disastrous Saudi 747/Kazakh Il-76 mid-air collision occurred near Delhi, the "Western" media's interest was centred on the Indian air-traffic-control (ATC) system. Delhi's terminal area radar was initially represented as being "primitive" by virtue of being primary only (secondary radar was in the process of being installed there). A week later, in the USA, a United Express Beech 1900 collided fatally with a private Beech King Air A90 on landing at Baldwin Field, near Quincy, Illinois, which is an uncontrolled airfield: ie, there is no ATC at all, and users are supposed to communicate position and intentions by broadcasting on a common frequency.
The "mature economies" like the USA may have better flight-safety records on average than those of Third World nations, but the existence of uncontrolled airfields proves that at least some in the USA still consider that simple systems can work just as safely as sophisticated ones, provided that they are operated correctly and within their limitations. Increasingly, as the world's industry searches for solutions to the problem of the widening gap between the safety standards of the richer and poorer nations, it is becoming clear that money and technology alone will not save safety problems.