Flight into Danger
(Published New Scientist 13 May, 2000)
Could major air disasters have been caused by faulty wiring and how can we avoid this in the future?
Gerry Byrne investigates
A LOUD BANG behind a panel above the pilot's head was the first sign of trouble. Sparks showered into the pilot's lap as red "out of order" flags popped up all over the instrument panel in front of him. The aircraft was losing electrical power rapidly, and more warnings erupted: the aircraft's batteries were discharging, the navigation instruments went down, and the auto-throttle failed. With only a few instruments operational, the plane was in trouble.
It was just after midnight on 22 June 1998, and the aircraft, a Boeing 757, had just taken off from Larnaca in Cyprus, bound for Manchester with 217 people on board. A major disaster appeared to be in the making. The pilot alerted air traffic control, and without pausing to dump fuel turned straight back to Larnaca. Fortunately, he managed to land the heavily laden aircraft safely.
When technicians removed the panel where the explosion had occurred, they discovered that the insulation on two wires was damaged where they had rubbed against a support bracket. The fault had caused a short circuit, which led to the sparking and the near-catastrophic loss of power.
In an industry as heavily regulated as aviation, you might imagine that incidents like this are rare. But a recent study by the US Air Line Pilots Association (ALPA) suggests that last year, on average, one US airliner a day was forced to make an emergency landing because of sparks, smoke or fire.
Faulty wiring is the leading culprit. And while the 757 involved in the Larnaca incident had been flying for only five years, it is older planes that are most likely to give trouble. Usually this is little more than a nuisance, no more serious than a dodgy power lead on a coffee pot. But sometimes the consequences are potentially catastrophic.
Some organisations have taken drastic action to deal with the problem. The US Navy in 1987 ordered the removal of the most vulnerable wiring from its planes, and last year NASA grounded its entire fleet of space shuttles when a wiring fault led to a launch being aborted. Yet every day, millions of passengers are still carried by commercial aircraft that are equipped with old wiring that cannot be properly tested for faults.
In the US, the Federal Aviation Administration (FAA) has been mounting a probe into the problems that may afflict aircraft that have been flying for more than 20 years . The Aging Aircraft Program has been running since 1988, prompted by an accident in which part of the roof peeled off an elderly Aloha Airlines Boeing 737 in the sky over Hawaii. It has involved exhaustive investigations into the structural integrity of things like wings, fuselages, control surfaces, landing gear and engine mountings aboard older aircraft. But nobody ever thought of looking at the electrical wiring. Until recently.
In 1996, TWA flight 800 came down off the coast of Long Island, killing all 230 people on board. Faulty wires inside a fuel tank were blamed as the most likely cause of the explosion. In the wake of that crash, checks on other airliners around the world led to the discovery of several other potential "flying bombs" in which the insulation on ageing wiring leading to sensors in fuel tanks had rubbed away through vibration, or had been damaged during routine maintenance. Then, in 1998, 229 lives were lost when Swissair flight 111 crashed off Nova Scotia minutes after the crew reported smoke in the cockpit. The cause of that accident has not yet been pinned down, but faulty wiring is one of the leading suspects.
Long before these disasters, warnings of the danger of old wiring--and of one kind of insulating material in particular--had begun to emerge. At the time of the TWA 800 disaster, one question haunted Richard Healing: "I kept asking myself, what did we know in the Navy that the commercial aviation sector didn't?" Healing, a former US Coast Guard captain, now director of safety and survivability for the US Navy, soon discovered that the answer was a tremendous amount.
After a series of baffling mid-air fires that killed several pilots in the 1980s, the Navy traced the problem to wiring in exposed areas of planes that had seen service aboard aircraft carriers: places like wheel wells, flaps and the hinged sections of folding wings. Attention focused on an insulation material made from a type of polymer known as an aromatic polyimide, often referred to by its common Du Pont proprietory brand name, Kapton. At first glance, Kapton, and its relatives, seemed to be everything an electrical insulator should be. It was tough, light, had high fire resistance and gave off relatively little toxic smoke if it did burn. As a result, aromatic polyimide coated wiring was used throughout the aircraft industry. But the Navy soon discovered that Kapton and other aromatic polyimides had a dark side.
When exposed to a combination of salt air and the solvents used to wash down aircraft aboard carriers, these wires experienced what was, in effect, accelerated ageing. The insulation became liable to crack if it was placed under strain. Worse, when faulty wires short-circuited or arced, the material changed from being an insulator to a partial conductor. Under the right circumstances, a bundle of arcing aromatic polyimide-coated wires could explode into a searing fire, generating temperatures of more than 1000 C--so hot that the wire's copper core would melt and spray outwards.
On their own, such fires will usually burn out harmlessly. But they can get hot enough to rupture and set fire to hydraulic pipes, fuel lines, insulation or any other flammable material nearby. So the US Navy spent hundreds of millions of dollars ripping the aromatic polyimides out of vulnerable areas aboard P-3 Orion surveillance aircraft and F-14 Tomcat fighters. In Britain, the Ministry of Defence heeded the warnings and started a programme to remove them from RAF and Royal Navy aircraft wherever possible.
Healing says the military were quite open about the urgency with which they removed aromatic polyimide or Kapton-type wiring. But manufacturers continued to install the same wiring in civilian airliners: Boeing used it until 1993, and Airbus Industrie continues to use limited amounts of it even now. "Did we have a problem communicating our findings to the other parts of the aviation world?" Healing asks. "I had a bucketful of information and I tossed it over the fence without checking that anybody knew about it."
Healing has helped to set up an industry organisation to combat both the cause and effect of aircraft fires--the Aircraft Wiring and Inert Gas Generator Group. He is concerned about the effect of ageing on all types of insulation used for aircraft wiring, not just Kapton. "People don't fully understand the seriousness of that degradation," he says.
Ed Block also has a strong interest in aircraft wiring: he is vice-chairman of the International Aviation Safety Association and a delegate on the Aging Transport Systems Rulemaking Advisory Committee. ATSRAC is charged by the FAA with investigating the condition of wiring in older aircraft. Block has been sounding alarm bells about Kapton and other wiring problems for more than a decade.
He points out that many older aircraft are flying long past their original design life. This is made possible by frequent inspections under "accelerated maintenance" programmes, but aircraft wiring poses special problems. Wiring, Block says, is designed for a maximum flying life of 60 000 hours. "TWA 800 had 93 303 hours on the clock when it crashed," he says. "That's 33 000 hours overtime."
ATSRAC is tackling the issue of ageing wiring on two fronts. The first, a series of visual inspections of wiring in a sample of older aircraft, is already complete. This identified more than 3000 "anomalies" aboard 81 aircraft, representing 26 per cent of the US fleet aged 20 years or more. These anomalies occurred in wires with aromatic polyimide insulation and in wires with other insulators, and only 140 of them were listed in ATSRAC's final report as being significant. The report also concluded that none of the anomalies posed a danger to flight. However, in a minority report, Block registered strong disagreement with this view.
Part of the problem with surveys like this is that paper-thin aircraft wiring insulation is so delicate that it can be damaged even by gently probing a thick bundle. Many of the anomalies that have already been found are held to be the result of careless handling by maintenance crews over the years.
ATSRAC is now moving on to the second phase of its work, a series of more intrusive inspections on retired aircraft, where damage caused by the inspection process itself doesn't matter. Block fears this may turn up even more problems. "It is generally accepted you can only visually and non-intrusively inspect 25 per cent of the wiring aboard an aircraft."
His view is partly borne out by a March 2000 report into the condition of NASA's fleet of ageing space shuttles, which were grounded for a time last year following the discovery of a series of faults, including wiring problems. In one incident, an aromatic polyimide-insulated wire had shorted out on a burred screw head, knocking out a primary and a back-up engine controller and leaving two of the shuttle's three engines without any controller back-up.
The NASA report lists 818 wiring problems on three shuttles, many of them relating to Kapton wiring. It concludes that most of the damage arose during maintenance and recommends that intrusive inspections be limited in order to minimise damage to wiring, which becomes increasingly delicate as it ages. It also warns that 20 per cent of wiring cannot be inspected without dismantling a large part of the shuttle. This means, the report says, that the job is best done when heavy maintenance is taking place.
The report also carries an even more ominous warning. It focuses on the shuttle's circuit breakers--switches designed to isolate the craft's electrical components when they sense problems such as a short circuit. The report says that the circuit breakers installed in the shuttle fleet do not always protect against an arcing fault, where sparks jump intermittently between damaged wires, or to ground. Instead, they interpret intermittent arcing as a varying load, so they may fail to trip even when current spikes exceed 10 times the danger level.
Similar circuit breakers are used in commercial aircraft, and American pilots are angry that the FAA is not taking their shortcomings seriously. Assuming that a circuit breaker will always trip out if there is a serious problem, pilots often reset tripped circuit breakers to see if a fault has rectified itself. If Kapton or other aromatic polyimide wiring is arcing, says Block, that action could be lethal. "Each time you re-energise the circuit you may be setting it up for an even more dangerous failure," he says.
In 1991, the FAA issued an advisory circular on the resetting of circuit breakers, warning that it could be dangerous. The FAA has left it up to pilots' discretion whether to continue with the practice, but in Britain the authorities have taken a stricter line. The Civil Aviation Authority has firmly instructed pilots never to reset circuit breakers, except in "exceptional circumstances". Airbus says it has issued similar recommendations for its aircraft. The ALPA is demanding that the FAA should tighten up its ruling, too.
The problem might be solved by installing different equipment. Arc fault circuit interrupters, a type of circuit breaker designed to detect and prevent arcing are now widely available for use in homes and cars. So why not put these breakers in airliners too? "There are some very unique things about aircraft power," says Jim Shaw, manager of the In-Flight Fire Project Team at ALPA. "You get what we call dirty power," he says. The current and voltage can jump when the pilot switches from one power generator to another. "That could cause all your arc protection circuit breakers to pop."
For Healing, that's not the end of the story, however. He says a research contract is being finalised between the US Navy, the FAA and two electronics manufacturers to miniaturise a smart circuit breaker that should be able to handle dirty power. A design may be completed within six months, though Block believes it could be up to two years before the regulatory authorities approve it for use on aircraft.
The main problem remains that there is at present no system that can report accurately on the condition of an aircraft's wiring. NASA routinely puts 1500 volts through some 115 volt space shuttle wiring systems in a bid to detect insulation faults. But it isn't foolproof, as defects only register if they are within a millimetre or two from a connection to ground.
A more thorough test involves soaking the wiring in a conducting solution and checking for stray voltages coming through the electrolyte. However, this system is mostly used for wiring that has already been removed from the plane, as operators are reluctant to spray corrosive conducting solutions onto their aircraft. And even if other insulation-checking systems become available, they may not give reliable results unless the testers also have baseline data for perfect, brand-new aircraft.
Much of the emphasis in the Aging Aircraft Program has been on planes which were designed and built in the late 1970s and early 1980s. Shaw warns that over the next few years much more complicated "fly-by-wire" aircraft, which have many more electrical and electronic systems, will start to come into the "ageing" bracket. With more than 300 kilometres of wiring in the average large jetliner, the problem may call for more radical solutions.
One answer, says Healing, might be to replace low-voltage wires--the sort suspected of causing the TWA explosion--with optical fibres or even to use VHF radio signals to transmit commands around the aircraft, doing away with most wires altogether. But refitting an old aircraft from scratch would probably cost more than the plane is worth.
Should all wires like those insulated with Kapton and other aromatic polyimides be removed from aircraft immediately? Safety advocates like Block certainly think so. He says the issue is a matter of life and death. "It's sad," he says. "We have come this far yet we are still tiptoeing around this subject because of its economic ramifications."
But others disagree. According to DuPont, there have been no accidents which upon analysis can be linked to Kapton. Airbus says it uses Kapton for cabin wiring because it is hard to ignite and, should a fire start, it gives off far less toxic fumes than comparable insulators. In the event of a fire it may improve survivability, says Airbus spokesman David Vailypilai. Healing agrees. "There are many places where it is the best solution provided it is not used in situations where it is vulnerable."
Whether the aviation industry changes its mind on the issue could depend on the results of ATSRAC’s wiring study, which will be completed by September.
Gerry Byrne is based in Ireland and is the winner of an IBM/STI Science and Technology Journalism Award
NASA's Space Shuttle Report is available at www.hq.nasa.gov/osf/shuttle_assess.html
The progress of the TWA 800 disaster investigation can be followed at www.ntsb.gov/events/twa800/default.htm
Details of the Swissair 111 investigation are at www.tsb.gc.ca
Comments from DuPont on Kapton insulation can be found at www.kapton-dupont.com/