How safe is your plane?

This weekend tests were ordered on 2,000 Boeing 737s after several unexplained crashes. There are concerns, too, about the 747 jumbo jet. Christian Wolmar examines an industry that is losing the public's trust

A safety crisis is looming in the world's airline industry. It arises at least partly out of a realisation that there probably was no bomb aboard the TWA Boeing 747 which plunged into the ocean three months ago, killing all 230 people on board. Instead, the investigators are focusing on a mechanical defect as the likely cause. And that is bad news for the aviation industry, the airline manufacturers and all who fly in planes.

Meanwhile, over the weekend, the Federal Aviation Administration ordered that all 2,000 Boeing 737s in the world's fleet should have tests on their rudders within the next 10 days. Although the trigger for this emergency measure covering the world's best selling commercial jet was the finding that the control mechanism could jam in a laboratory experiment, it follows years of concern about its rudder, culminating in a devastating series of articles on the 737's rudder problems, published last week in Boeing's home town newspaper, the Seattle Times. Uncommanded rudder deployment, which can make the plane flip over in seconds, is the suspected cause of a number of accidents, including two unexplained crashes in Colorado Springs and Pittsburgh in the past five years.

As far as Flight 800 is concerned, it would be much easier for Boeing and the industry if a bomb had been responsible for its destruction. But the terrorism theory had been looking increasingly shaky, the longer no definite statement emerged from the investigation team into the disaster on 17 July. It would have taken only a small amount of by-product from an explosion to have enabled the investigators to conclude with certainty, as they did within days of the Lockerbie disaster, that terrorism had been involved. Yet no such finding has been made, and no group has claimed responsibility for blowing up the plane.

Bombs may be the act of men but they seem more like an act of God, out of our control and virtually impossible to prevent. They may deter a few people, especially American tourists, from flying, but, with a few promises of tighter airport security and the addition of a few more inane questions at the check-in desks, the airline industry normally bumbles along quite happily.

Technical defects are another matter entirely, as they raise fundamental questions about airline safety and the future of the industry. The cause of the Flight 800 accident has been traced to a fuel tank, located between the wings and beneath the cabin, which exploded. As the flight from New York to Paris was relatively short, the tank was nearly empty and something ignited the fuel vapours to cause the massive conflagration. The possible targets are either a fuel probe in the fuel tank - though Boeing says that it does not carry sufficient current to cause a spark - or a fuel pump which lay adjacent to the tank.

Lawyers in New York, acting for 30 of the bereaved families, have produced a 27-page analysis, written by two experts, of the likely scenario which led to the crash. It centres on a fuel pump which produced the fatal spark, but, more important, it raises questions about the use of older aircraft by the world's airlines. The report says that, while the world's safety authorities have focused on the structure of ageing airliners, "there were no programmes for old aircraft to evaluate systems, flight controls and fuel management. This appears to be a real void in the airworthiness standards of ageing aircraft. We respectfully suggest that the Federal Aviation Administration examine this problem."

There are an awful lot of older aircraft out there. Flight 800 was operated by a 25-year-old plane, the 153rd Boeing 747 to be built, out of a total approaching 1,100. Any changes that result from the investigation into TWA 800 could be prohibitively expensive, meaning that a lot of them may be grounded or scrapped. Even worse, the accident may have highlighted something wrong with the whole design of the aircraft, necessitating that all 1,083 built are modified.

Such a prospect would have catastrophic knock-on effects for the industry, which would grind to a halt: there are nearly 1,100 Boeing 747s - 700 of which are the Classic type involved in this crash - and they are the workhorses of the industry.

There is a recent precedent for a major redesign programme. In October 1992, an El Al 747 freighter smashed into a block of flats, killing at least 43 people on the ground, after two engines fell off the right-hand wing. The accident investigators found that metal fatigue in a fuse pin, a cylindrical part the size of a standard baked bean tin, had been the primary cause. The fuse pin is part of the mechanism which holds the engine on to the wing, and the take-off of the heavily laden aircraft in turbulent conditions caused it to crack and give way.

Only a few months previously, another 747 freighter had gone down in Taiwan, also after losing an engine, and there had been several other incidents involving engines dropping off 747s - as well as 707s, which have similar fittings. After a lengthy investigation, Boeing decided to replace the engine pylons on all 747s flying and to change completely the plane's safety philosophy. The fuse pins had been designed to fail deliberately under conditions of severe vibration, so that the engine fell off before the wing broke. With the new fittings, the engine should never fall off.

The cost of this exercise, carried out quietly, without publicity, is a staggering $1,200m dollars, most of which is being borne by the manufacturer, although airlines are contributing some of the labour. The very fact that Boeing felt it necessary to spend such a large sum shows the extent of the company's concern. The replacement programme, by the way, is still in its infancy as only 190 aircraft have so far been fitted with the new pylons and airlines have until 2001 to ensure the work is carried out. In the meantime, the suspect pins are inspected and replaced much more frequently than previously.

This could happen again. Suppose the fuel pumps need to be replaced. Is anyone going to want to fly in a 747 until the work has been carried out?

And if the 747 has problems, so does the 737. The industry loathes unexplained accidents, but it has failed to get to the root cause of several mishaps involving 737s. The weekend's announcement by the FAA that it will test all 737s may be the first in a series of moves which could result in a major retroactive redesign of the rudder. The 737 is unique among modern aircraft types in that its rudder is all in one piece and is controlled by one mechanism. Other planes have several controls, or have rudders which consist of several different moving parts. The problem with the 737, therefore, is that if the control goes wrong, there is no fail-safe. According to the Seattle Times, several pilots have had desperate battles with their aircraft's controls after uncommanded rudder deployments. Recently, Boeing changed the design of new 737s so that the rudder has a more complex control system, but it has said that this decision was taken for "technical" rather than safety reasons. However, the FAA may now force Boeing into a massive redesign programme costing billions of dollars. If it has to do that for either the 737 or the 747, problems would be caused. If it were forced to do it for both, the industry would be in crisis.

Nor is Airbus immune from these risks. The new Airbus types, such as the 320, have had problems with pilots getting confused with "fly by wire" mechanisms and, as a result, crashing.

All these problems, and the ever-present "human factors" which contribute to between 75 and 90 per cent of accidents, mean that safety rates are no longer improving. According to Boeing's own statistics, the chances of a passenger on a commercial jet plane being involved in an accident in which the aircraft is destroyed (though the passenger himself may survive) have hovered around the 1 in 500,000 per flight for the past 20 years.

This failure to improve safety rates means that every year there are going to be more crashes. The number of aircraft movements is going up by 4 per cent annually which means that, without any improvement in the safety rate, by the early days of the next decade there will, on average, be a major accident every 10 days. Mike Willett, the CAA's director of safety regulation, wrote in a letter recently to the Parliamentary Advisory Council for Transport Safety that "it would be necessary for the UK fatal accident rate to reduce by a third over the next 10 years to ensure that the absolute number of fatal accidents does not increase".

That is a very difficult target. The CAA has a series of 80 programmes to improve future safety, such as trying to draw lessons from the 6,000 "mandatory occurrence reports" filed by British airlines every year, but Mike Bell, its head of technical services, admits that, while the CAA's target is to have "zero accidents", "it is getting more and more difficult to reduce the rate of accident, because there are very few and each one has individual characteristics." The lessons learned from previous accidents have been implemented, making it difficult to make further improvements. Mr Bell said: "Our team of six people finds it very difficult to agree on the primary cause of crashes. There are often seven or eight possible factors, and one can't decide which was the main one."

It's been a bad few weeks for aviation. In the past month, a 757 crashed at sea off the coast of Peru, a 707 freighter smashed into housing in Ecuador, and an Antonov 24 destroyed housing at the end of the runway in Turin, Italy. Only last Friday a Fokker 100 killed many people on the ground in Sao Paulo, Brazil. What makes these crashes even more worrying is that in all of them people were killed on the ground.

Predictably, all these accidents involve the riskier types of aircraft. Older planes, freighters, and Third World operators and airports, all have higher than average numbers of accidents. Yet many such planes - such as the Air Algerie freighter which crashed at Coventry in December 1994 - are allowed to fly into the UK. Jeff Gazzard leads a campaign against a second runway at Manchester airport and lives under the flight path. He said: "We are desperately concerned. It is time that certain types of aircraft, or some airlines, were banned from airports."

The one thing that would put paid to BAA's hopes of building Terminal Five at Heathrow, currently the subject of the longest ever planning inquiry, would be an accident in which Londoners are killed by a falling plane. It has been 24 years since the Staines disaster, when a Trident plunged into a field near the airport. Statistically, we are overdue for a London air disaster and, if one were to occur, a lot of questions would be raised about whether we should allow the industry to grow unfettered.

Aviation safety is about perception. As you belt up before thundering along the runway, your chances of being killed are tiny and no greater than at any time in the past two decades. But that misses the point. The industry can go on endlessly about how it is safer than other forms of travel but it will be to no avail if aircraft fall out of the sky with the regularity of the past few weeks. There will be a backlash. People will not only no longer want to fly, they will not even want to live near airports. The industry has to find a way to improve again on safety rates, or else events such as TWA 800 and the various recent accidents which have killed people on the ground will lead to a crisis of confidence.