ATM Modernization, Business & GA, Commercial, Military

Safety in Avionics: Time to Stop Fuel Tank Explosions

By David Evans | June 1, 2001
Send Feedback

It is often said that it takes two and a half accidents to spur the U.S. Federal Aviation Administration (FAA) to corrective action. By this standard, the feds are late, sunk even deeper into a Rip Van Winkle-like torpor of regulatory inactivity. And until they act, every pilot of a transport-category aircraft is flying a potential bomb, although no clues are provided by current cockpit instrumentation.

Not a bomb planted by terrorists. The bomb is the potentially explosive energy contained in the fuel/air vapors in the fuel tanks. All it takes is a tiny spark somewhere in the tank to set off the fatal "kaboom!"

The most recent explosion occurred March 3, when the center wing tank (CWT) of a Thai Airways International B737 exploded while the airplane was sitting on the tarmac at Bangkok’s main airport. The explosion set off a spreading fire that, 18 minutes later, caused the right wing tank to explode.

It was the third center wing tank explosion in the past 11 years. In 1990, the CWT exploded on a Philippine Airlines (PAL) B737 at Manila, and in 1996 the CWT exploded on a Trans World Airlines (TWA) B747 shortly after takeoff from New York (see sidebar).

These three fatal fuel-tank explosions have certain elements in common: All three airplanes had only a small amount of fuel in the wing tank. The air-conditioning packs, located in non-vented bays directly under the CWT, had been running before the explosions. And the outside air temperatures were quite warm.

The air-conditioning packs generate heat, which contributes to fuel vaporization in the non-insulated tanks. On the Thai jet, the packs had been running throughout the previous flight, and then for about 40 minutes while the jet was on the ground. They were running for about the same amount of time on the ground before the PAL jet exploded. On the TWA jet, packs had been running for about two and a half hours before the doomed airplane took off.

Outside air temperatures in both the Thai and PAL cases were in the high 90s (degrees Fahrenheit). The TWA jet had been sitting on the hot tarmac at New York’s John F. Kennedy (JFK) International Airport in July. The ambient air temperature was probably much more than the 82� F high reported for the day.

In the exhaustive TWA investigation, a bomb was ruled out for lack of the telltale microscopic evidence of pitting and "gas wash." The tank’s front bulkhead, blown forward, was a clear sign that the fuel/air vapors in the tank exploded. On the Thai jet, the CWT forward bulkhead also was blown forward.

Moreover, the sound signature on the cockpit voice recorded (CVR) of the Thai jet was similar to that of the PAL jet in 1990. Neither recording included the precipitating sound of an initiating explosion that may have ignited the fuel tank vapors.

The Thai Airways accident brings to 17 the number of fuel tank explosions that have occurred on jet airliners since the first event on a B707 in 1959. All told, about 550 people have been killed. Although the reigning philosophy has been to accept the presence of flammable vapors and to vigorously hunt down and eliminate all potential ignition sources, the toll exacted thus far is stark proof that the current design philosophy is a failure.

Two points bear emphasis. First, the three CWT explosions in recent years demonstrate the repeatable nature of flammable fuel/air vapor explosions. Although these tragic events are infrequent, an increasingly aware travelling public may become apprehensive about the danger of fuel tank explosions (to say nothing of more informed pilots).

Second, the explosion of the Thai jet occurred, one might say, right on schedule. After the TWA jumbo jet blew up in 1996, the FAA convened an Aviation Rulemaking Advisory Committee (ARAC) to determine what could be done do mitigate the hazard posed by flammable vapors in fuel tanks. The ARAC’s final 1998 report predicted gloomily that if the industry did nothing, flammable vapors in center wing tanks could be expected to explode at an average rate of about once every four and a half years (once over 54 months). Coincidentally, the CWT on the Thai jet exploded 56 months after the TWA jet.

That 1998 ARAC report concluded that inerting the tanks just was not cost-effective. To its credit, the FAA convened a second ARAC and charged this "Son of ARAC" to look harder at ways to inert fuel tanks. The second committee is slated to issue its final report next month. As of this writing (late April), the task force is looking at several promising approaches. The one it’s not looking at may be the most attractive. Here are the three options the task force has under consideration:

  • Ground-based inerting. The fuel tanks would receive a squirt of nitrogen-enriched air before pushback. Inerting would last for the taxi, takeoff and climb portions of flight, when the fuel vapors would be warmest.
  • On-board ground inerting. The inerting equipment is placed on the airplane.
  • On-board inert gas generation system (OBIGGS). With this "all-singing/all-dancing" system, the ullage in the fuel tanks would be inerted at all times, from pushback to docking. (Ullage is the vapor in the space between the fuel level and the top of the tank.)

A fourth approach not under formal consideration may represent the best all-around compromise in terms of cost vs. capability. Under the concept conceived by New York-based Aviation Safety Facilitators Corp., the airplane would be provided with a supply of high purity liquid nitrogen from a ground source. This super-cooled liquid nitrogen would be stored on the airplane in a vacuum-sealed insulated container, or dewar, and would be fed to the fuel tanks under low pressure (less than 20 psi.).The dewar saves valuable space. The 100 gallons (380 liters) of liquid nitrogen stored in the full dewar would provide some 9,300 cubic feet (263 m3) of inerting nitrogen gas. That’s enough to completely fill the 2,100 cubic-foot center wing tank of a B747 three times over. As such, the airplane could operate into airports without a local capability to produce liquid nitrogen (commonly used in the food packaging industry and in other industrial applications) and still have enough in the dewar for additional flying.

Oxygen sensors in the fuel tanks would provide feedback to the system’s computer control, which would meter the liquid nitrogen to keep the fuel tanks inerted for all phases of flight (departure through arrival). The liquid nitrogen also could supplement fire suppression in other areas of the aircraft (cargo holds, electronics and equipment bays, landing-gear wheel wells, etc.).

Pilots would have a small display in the cockpit that would provide them with potentially lifesaving information. For example, pilots presently don’t know the temperature in their fuel tanks (hence, they have no way to determine if the ullage is in a flammable temperature range). Ah, the illusion of safety through ignorance.

But with the liquid nitrogen system, they would know the tank temperature at all times, and they would know the oxygen content of the tanks. Thus, when the oxygen content of the ullage is at 7% or less, pilots would have positive assurance the tanks are inerted. They would have other information, such as the liquid nitrogen supply in the dewar, system status, alarm conditions, and so forth.

Nothing fancy, just the assurance that the ullage in the tanks finally has been tamed.

It’s nice to have a wonderful suite of avionics to tell pilots where they’re going; it would be even nicer to give them a display with proof positive they won’t get blown up along the way.

David Evans is the award-winning editor of Air Safety Week. Comments can be sent via e-mail to [email protected].

Three CWT Explosions of Doom

  • May 11, 1990, Manila, Philippines–Shortly after pushback, the center wing tank (CWT) exploded on this Philippine Airlines Boeing 737, killing eight people. The CWT had not been filled since March 9, 1990. Air-conditioning (A/C) packs had been running on the ground before pushback (approximately 30 to 45 minutes). Ambient air was 95� F (35� C).
  • March 3, 2001, Bangkok, Thailand–Center wing tank explosion on Thai Airways International B737 was followed 18 minutes later by an explosion in the right wing tanak. Residual fuel was in CWT. Air-conditioning packs had been running continuously since the airplane’s previous flight, including about 40 minutes on the ground. Ambient air temperature was in the 90s (degrees F).
  • July 17, 1996, New York–CWT exploded on this TWA B747-100 shortly after takeoff from JFK International Airport, killing all 230 aboard. CWT contained a slight amount of residual fuel. A/C packs had been running on the ground for 2.5 hours before takeoff. Ambient temperature was 82� F (28� C).

Source: National Transportation Safety Board

Receive the latest avionics news right to your inbox