Flight 800 Mystery Now A Mechanical Puzzle -- Boeing Engineer: `If We Knew What Made It Explode, I Think We Can Fix It.'

On a warm evening last July, a Boeing 747-100 took off from New York's Kennedy International Airport bound for Paris, carrying vacationers, business travelers and a high-school French club from Pennsylvania.

Eleven and a half minutes into TWA Flight 800, things went terribly wrong. Vapors in the empty center fuel tank exploded, twisting a structural beam into the belly of the plane like a can opener. In a matter of seconds, a tear raced up to the row of windows on each side. The plane's forward section, with its distinctive hump, broke off and fell away.

But the engines kept roaring, arching the decapitated plane forward. About a half minute later, the wing tanks peeled open, releasing tens of thousands of gallons of jet fuel into the atmosphere. The fuel erupted into a ball of fire seen for miles. All 230 people on board died.

It was the second-worst air disaster in U.S. history. And, at a time of terrorist fears surrounding the Olympic Games, it invoked memories of the bombing of Pam Am Flight 103 over Lockerbie, Scotland, in 1988. There were suspicions TWA Flight 800 likewise had been brought down by an act of terror.

Now ten months later, those suspicions have been all but laid to rest by federal officials. In one of the most intensive crash probes ever conducted, they agree that the 25-year-old jet likely broke apart because of "catastrophic mechanical failure," as FBI Director Louis Freeh put it last week, acknowledging what the National Transportation Safety Board (NTSB) had been saying for months.

"The (center) tank is what opened up the airplane and started the events. . . ." said Bernard Loeb, director of aviation safety for the NTSB.

Boeing also acknowledges that the center fuel tank, a cavernous structure about the size of a living room, exploded. But the company won't concede there was a mechanical failure.

"The center tank exploded," said Steve Hatch, the company's 747 chief project engineer. "We've known it for a long time. The real question is what made it explode? If we knew what made it explode, I think we can fix it."

Ignition theories: frayed wires, static electricity

Since January, NTSB investigators have meticulously reconstructed the fuselage of the 747 from thousands of burned and mangled parts in a Calverton, N.Y., hangar. The evidence has convinced them that a mechanical or electrical malfunction caused TWA Flight 800's catastrophic explosion.

Now with the first anniversary of the crash approaching, they are running a series of elaborate tests to try to isolate the ignition source.

Boeing 747s are frequently flown with the 13,000-to-16,000-gallon-capacity center fuel tank nearly empty, because the extra fuel is needed only for very long flights. Even when the center tank is filled, the fuel it holds is burned first to keep the plane balanced.

An empty center tank is never completely dry. Fuel pumps inside the tank cannot draw out the final 100 or so gallons, leaving a potentially dangerous mix of fuel and air. When such a mixture is heated to the right temperature, it forms a highly combustible vapor capable of being ignited by the tiniest burst of electricity or heat.

NTSB investigators have determined that the TWA jet had been sitting on the tarmac for several hours in hot weather before its final takeoff. The air-conditioningunit on the underside of the center tank was left running full blast. Tests showed heat from the unit heated the fuel and air inside Flight 800's tank into a volatile vapor.

Aviation documents confirm there are numerous potential ignition sources in a 747's center fuel tank. Worn-out parts inside the center tanks of older 747s, in particular, are more likely to overheat, emit stray sparks and generate static electricity, the documents show.

Last January, the Federal Aviation Administration (FAA) issued an emergency order calling for the 433 oldest Boeing 747s still in service to be inspected for chafed electrical wires that could ignite a fuel-tank explosion.

The FAA in February ordered repetitive inspections of fuel pumps on 747s and 757s for corroded or frayed electrical connectors.

An early round of inspections found defects or deterioration in about one out of every 25 fuel pumps. Sixty-seven pumps were replaced. The February inspection order was not directly related to the TWA Flight 800 explosion but underscored a longstanding concern about the potential for leaking fuel pumps to spark fires on 747 and 757 jets.

"You've got all the conditions in there for something bad happening," said Sullivan Marsden, professor emeritus of petroleum engineering at Stanford University and an expert in oil-tank accidents.

A Boeing 747 jumbo jet stores jet fuel in seven tanks - three in each wing, plus the center tank - connected by a network of fuel lines, venting pipes and electrical pumps and sensors.

After piecing fragments of Flight 800's fuel system back together in the Calverton hangar, investigators were able to see plainly that the heaviest fire damage occurred in the forward, right side of the center tank and the adjacent fuselage and wing sections.

Besides the center tank, the No. 4 reserve tank, located near the outboard right engine, burned internally, and vent lines traversing the right and left wings were burned from the inside as well.

As the reconstruction progressed, investigators could find no evidence of a bomb or of a missile striking the jet. Nor could they find any sign that engine failure, a loose cargo door or weak fuselage framing members - all problems that have occurred on 747s - played any role in the crash.

Homing in on the center fuel tank, one of the first things they explored was the possibility that static electricity, sparked by fuel leaking from a damaged or weakened fuel-line coupler, may have ignited the explosive vapors in the center tank.

A static spark from a broken fuel line was blamed for simultaneously igniting a wing and center fuel tank on an Air Force KC-135 tanker, a military version of the Boeing 707, at Guadalajara, Spain, on June 3, 1971. All five crew members were killed.

One possible source of a static charge inside 747 fuel tanks is the "crossfeed manifold" - sections of aluminum tubing, joined by couplers, designed to carry fuel from any tank to any engine. Fuel spraying from a leak in the manifold, perhaps from a worn-out coupler, could create a static charge. That could happen as bigger drops of spray fall faster than smaller drops, causing charged particles in the fuel to separate.

Boeing's Hatch said static electricity could provide an ignition source inside the fuel tank of any model jetliner, but he noted the rarity of fuel-tank explosions blamed on the phenomenon.

"With 14,000 airplanes out there, why haven't we seen more?" Hatch asked. "If it's there, we haven't seen it, but we'll continue to look."

Could fuel pump have overheated?

Another possibility is that one of the three fuel pumps mounted partially inside Flight 800's center tank overheated because it was turned on with little or no fuel to pump. That's what happened on an Air Force B-52 that exploded in flames Dec. 6, 1988, at Sawyer Air Force base in Michigan, seriously injuring the eight crew members.

A Boeing engineer, Carl Norris, conducted tests for the Air Force in 1989, demonstrating that the B-52's fuel pumps, supplied by Burbank, Calif.-based Hydro-Aire, did not meet a Boeing specification that they be capable of operating without overheating even while running in an empty tank.

Norris recommended that enough fuel be kept in the tank at all times to keep the Hydro-Aire pump submersed in liquid. The Air Force adopted Norris' recommendations and eventually replaced the suspect pump with another model made by a different manufacturer.

However, Hydro-Aire fuel pumps similar to the one blamed for the B-52 explosion continue to be widely used on 747s and were used on the TWA jet that crashed.

Boeing's Hatch said the company has explored ways to install Hydro-Aire pumps on 747s so that they are always submerged, but does not plan to make any changes.

Hatch said there were differences between the Hydo-Aire pumps used on the B-52s and those on 747s, but he did not specify what the differences were. He said Boeing is convinced the 747 pump could not fail in the same way as the B-52 pump.

The NTSB's Loeb said two of Flight 800's center-tank fuel pumps were retrieved from the ocean and examined at a NASA lab in Huntsville, Ala., about a month after the accident. Loeb said they showed no evidence of electrical arcing or overheating.

But the third pump, called the "scavenge pump" because it is designed to draw as much of the last bits of fuel from the center tank as possible, has never been found. Also missing are the center-tank fuel probes, long poles set in 13 positions inside the tank.

The probes house electrical sensors that measure the level of fuel, and are specifically designed to eliminate the possibility of stray sparks. They use a very small amount of electricity protected by double walls of insulation, Hatch said.

But a fuel probe could have served as a kind of lightning rod, attracting enough static electricity from a nearby leak to generate a spark, said Stanford's Marsden. Investigators are also checking out the possibility that overheating of the missing scavenge pump ignited the center tank.

New York plaintiffs attorney Lee Kreindler, who represents families of victims of Flight 800, has sent the NTSB two extensive studies outlining how a short circuit in the missing scavenge pump could have ignited the vapors.

Kreindler arranged to have two fuel pumps removed from a recently retired Continental Airlines 747-100, now parked in the Mojave Desert. Both pumps had frayed wires and one had burn marks, indicating it sometimes issued stray sparks.

"This was off an airplane right off the line, retired just a couple of months ago," said Kreindler. "If the mixture of air and temperature in that fuel tank was right, you would have had another TWA 800."

Loeb said the NTSB has no evidence supporting Kreindler's theory.

Boeing's theories focused on terrorism

Boeing engineers have offered several theories about the crash to investigators.

Early in the investigation, they produced studies showing why the 100 gallons of fuel in the jet's 13,000-gallon tank would have been too cool to ignite. That theory came apart when the NTSB discovered that the fuel and air in the tank had been warmed into a volatile vapor by the air-conditioning system.

Next, Boeing produced analyses seeking to demonstrate that even if fuel vapors exploded inside the center tank, the blast could not have been strong enough to cripple the jet. But the wreckage reconstruction showed the initial explosion was powerful enough to rupture the belly of the plane and split the jet apart.

Most recently, Boeing developed a theory that a shoulder-fired missile, which could not have reached Flight 800 at the altitude at which it was flying, might have come within several hundred feet of the jet and self-destructed, hurling a tiny piece of hot shrapnel into the center tank. Investigators have closely examined the theory but remain unconvinced.

"Boeing did some computer work, but this is an extremely unlikely scenario," Loeb said. "You'd have to get it into a unique position where it self-destructs at just the right moment. It's a very way-out scenario, on the order of a piece of meteorite or space junk re-entering the atmosphere."

Freeh's comments last weekend all but eliminated the missile theory.

Boeing spokesman Doug Webb said positions taken by Boeing that were subsequently disproved were based on the best information available at the time. "Assertions do change as you go along," Webb said. "You don't stay married to a theory."

FAA hesitates on safety-board recommendation

Meanwhile, the NTSB remains concerned about the safety of hundreds of aging 747s. It issued a series of urgent recommendations last December to make airliners less susceptible to destruction by fuel-tank explosion.

The NTSB called for installing a safety device that automatically injects an inert gas, such as nitrogen, into tanks to replace air as the fuel level drops. While oxygen and fuel can be explosive in the right mixture, nitrogen mixed with fuel is harmless.

Military combat aircraft have used inert-gas systems for decades to prevent bullets that pierce a fuel tank from igniting the tank.

Boeing has taken no position thus far on the safety board's call to retrofit all passenger jets with inert-gas systems. But the company clearly has reservations.

Hatch said that inert systems "add another level of safety to the airplane," but he questioned whether the benefit would be worth the cost and whether installation of an inert-gas system could inadvertently cause other problems.

"You've got to look at the cost/benefits versus occurrences and say, hey, it doesn't make sense if it's something that drives the cost up so much that ticket prices increase and less people fly," he said. "Some of these things have a tendency to work against you. It's really not going to help us."

Nonetheless, Boeing has begun preliminary design of an inert-gas system, Hatch said.

The safety board also called for measures to prevent air-conditioning units from warming the fuel in 747 center tanks, new techniques to allow only cool fuel to be pumped into the tanks and procedures to ensure that enough fuel is kept in the center tank to prevent formation of explosive fuel-air vapors.

Asked if he would assure the public that older 747s are safe, the safety board's Loeb, choosing his words carefully, said: "I'm not going to give you any assurance. I fly on those airplanes. We made recommendations to the FAA that I think would improve safety and diminish the risk. I don't know what more I can do.

"What I'm saying is, without ever determining specifically what that ignition source (of TWA Flight 800) is, I've got some fixes to the problem that will make it safer," Loeb continued.

The NTSB, a tiny federal agency, investigates airplane crashes and makes safety recommendations, but it must rely on the much larger Federal Aviation Administration to write and enforce safety rules.

The FAA responded by questioning the effectiveness of inert-gas systems even though it had formally proposed such systems for all passenger jets more than 20 years ago, only to drop the proposal after airlines complained about the cost.

"The FAA philosophy regarding flammable fuel vapors is that the best way to ensure airplane safety is to preclude ignition sources within fuel tanks," Darrell M. Pederson, acting manager of the FAA's large-aircraft-certification office in Renton, wrote last month.

Pederson said implementing the NTSB recommendations would "require a significant change in airplane design and/or operational practices currently in use." He questioned the benefits of such changes and described the cost of installing inert-gas systems as "prohibitive."

The FAA last month issued a request for comment and has done nothing since.