Emergency Landing Highlights Fire Hazard in Closed Spaces

A recent belly hold fire once again raises the specter of aircraft vulnerability to conflagrations in areas not accessible by flight crews armed with portable fire extinguishing equipment. In this case, the fire was suppressed by Halon, and the incident aircraft was just 10 miles from its destination airport and from emergency firefighting support upon landing.

Nevertheless, the May 13 incident involving an Air Canada B767-300 with 185 passengers and crew on board is under investigation by the Transportation Safety Board (TSB) of Canada. The case has many of the same earmarks that characterized the fatal 1998 crash of Swissair (now SWISS) Flight 111, an MD-11, in which a runaway in- flight electrical fire in an overhead ceiling area doomed the aircraft. The TSB's investigation of that case is in its final stages. This most recent case occurred below deck, but it is significant nonetheless as another case where electrical system safety, materials flammability, and the adequacy of fire detection and suppression are major issues.

The overall threat of in-flight fire is serious. Just a week before the belly hold fire in the Air Canada jet, a China Northern Airlines MD-82 twinjet crashed May 7 in the bay at Dalien, China. Preliminary reports cite a raging cabin fire; all 112 aboard were killed and investigators from the U.S. National Transportation Safety Board (NTSB) are assisting Chinese officials in the investigation.

According to a preliminary TSB account, the Air Canada jet was on descent to a landing at Toronto on a flight from Vancouver when the crew noted an aft cargo bay fire warning. They activated the built-in extinguishing system, declared an emergency, and the aircraft was met upon coming to a stop on the runway by airport firefighters. The firemen, using infrared sensing equipment, did not detect any sign of fire, so the aircraft taxied to a position just 40 feet short of the gate to allow firefighters to open the aft cargo compartment for a detailed inspection. Smoke billowed out when the door was opened. The firefighters crawled into the belly hold and discovered that the fire had been extinguished by the airplane's Halon system.

Intense heat
The charred evidence and metal structure distorted by exposure to heat suggested to TSB investigators that "an intense but relatively small fire occurred, causing significant structural damage," according to the TSB preliminary report.

The damage included two areas where the thermal/acoustic insulation blanketing was burned, and aluminum structural web had burned through completely, resulting in "significant heat distortion of the floor beam." The cargo bay does not have a solid floor, per se. Rather, there is an aluminum webbing between the joists and rails supporting the cargo pallets. The fire consumed the aluminum web between these support members. Not so much as a congealed bubble of melted aluminum was found.

"The fire was hot enough to consume aluminum, not just melt it," said TSB investigator Mike Stacey.

"They were fortunate the fire was detected early," Stacey added. The charred evidence suggests that the fire already was spreading. "It looks like the insulation blanketing provided fuel for the fire," Stacey said. Two types of insulation were installed. Metalized Tedlar-covered insulation was at the bottom of the cargo hold. Basically, this material lines the bottom of the pressurized hull. Non-metalized Mylar-covered insulation was used in areas above the bilge. Neither of these materials meets the latest radiant heat and direct flame test for insulation material flammability (see ASW, Oct. 16, 2000).

The fire burned its way forward and also spread approximately 18 inches up the outer side of a fire liner placed in the cargo bay to contain a fire. The defense supposedly provided by that liner clearly was breached. Fortunately, the Halon was able to follow and suppress the fire, halting its increasingly dangerous spread.

Traced to the source
The ignition source was traced to an electrical failure of a heater tape used to prevent water lines from freezing. "We believe the source of the fire was on the potable water drain line," Stacey said. Similar damage was found on other aircraft (e.g., burned insulation wrap), but they had not been significant enough to cause a fire. Air Canada took "immediate steps to inspect aircraft and deactivate affected systems," according to the TSB's preliminary report. Investigators are looking closely at the kind of fluids and other contaminants that soiled the insulation blankets during routine servcie, reducing their fire resistance. The incident airplane was built in 1991; the metalized Tedlar was most likely installed after entry into service, as this material is more durable than the factory-installed insulation material, which tends not to hold up well, one source explained.

The latest case recalls the forward cargo compartment fire that occurred Sept. 17, 1999, on a Delta Air Lines [DAL] MD-88 twinjet. In that incident, the heat from failed electrical circuitry led to burned insulation material in the belly hold area. In this case, too, the insulation material served as fuel to the fire. The case prompted concerns that extended well beyond flawed circuitry in the static port heater, to include flammability of insulation materials, excessively tight bend radii that can cause insulation on electrical wiring to crack prematurely, and the need for a design review to reduce the potential for dangerous arcing.

Broad implications
The May 13 case is likely to provide added grist to the TSB's earlier expressed concern about the need to remove flammable materials from aircraft. TSB officials maintain that if combustible materials were prohibited from use, in-flight fires would not occur (see ASW, Sept. 10, 2001). Indeed, TSB officials attach considerable significance to the Air Canada in-flight fire. They perceive a number of major issues:

Consequences of ribbon heater tape failures.
The maintainability of ribbon heater tapes.
The flammability of insulation blanket materials and the flammability of debris in cargo bays.
The degrading effects of age, condition and contamination on the flammability of insulation blankets (i.e., making them more prone to accelerate rather than retard an in- flight fire).
The potential for fires in inaccessible areas of an aircraft, where there is a lack of fire-fighting capability, to spread with lethal consequences. This case doubtless will help to beef up the TSB's forthcoming report on the Swissair Flight 111 disaster and the hazard of an fire in inaccessible areas that cannot be monitored from the cockpit.
Actually, one might go further in terms of implications. Terrorists may quickly perceive the vulnerability of belly holds to caustic chemicals and flammable materials, placed in baggage as well as freight containers, the deadly combination activated by a barometric timer. The terrorist threat provides an added impetus for improving the fire protection of airliner belly holds.

The Air Canada jet was saved by the Halon, injected into the cargo bay under the pressure of inert nitrogen gas. Engines and auxiliary power units (APUs) normally feature the option of combating a fire with a second Halon injection from another engine's bottle, should that be necessary. The absence of a second dousing shot of extinguishing Halon for belly holds may be a significant oversight. In the B767-300, three Halon bottles are plumbed to combat a fire in either the forward or the aft belly hold. Bottle 1 is supposed to extinguish the fire. After a 30-minute delay in flight, the Halon stored in Bottles 2 and 2A is metered to provide 195 minutes of ongoing fire "suppression" with a three- percent concentration of Halon. This feature is necessary should a fire warning occur during extended twin-engine operations (ETOPS).

However, Bottle 1 is supposed to "knock down" the fire. In the case of the B767-300, Bottles 2 and 2A not only supplement the Number 1 bottle during the remainder of the flight, they are rigged to discharge completely immediately upon landing, regardless of the time delay. This design feature may have had a significant impact in the Air Canada case, as the fire received the equivalent of a second "knockdown" of Halon just minutes after Bottle 1 discharged.

If the fire had broken out earlier in the flight, would the 195 minutes of suppression have been of any use if the outer skin had been pierced and the hold open to the airflow? In addition to a second-shot option, pilots might also benefit from closed circuit television (CCTV), enabling them to physically see the course of events when fire strikes an inaccessible area. Smoke detectors, once initiated, are useless in a smoke-filled environment. A CCTV would provide pilots the ability to ensure that the fire is out. To be sure, infrared cameras would be needed for the CCTV to provide the pilots with the best capability to locate the heat of flames concealed visually by smoke. Infrared cameras are being installed in Swissair's surviving MD-11s as part of the carrier's "Modification Plus" program (see ASW, July 30, 2001).

Moreover, water misting or galley/lavatory-water diversion, if available, would give pilots the option to "flood" the hold. Not least, simultaneously getting the power off all electrical circuits in the affected area certainly would aid in killing an electrical fire. Sources advise that when the fire extinguishing system on the B767-300 arms, power is cut to most of the electrical systems in the cargo hold, but not the heater ribbon tape. Many electrical fires can get a second breath after having been doused with fire suppressant the first time. Another question is whether it would be worth the cost of the extra plumbing to build in the capability to divert engine and APU fire-bottle output into the cargo holds.

In addition, if nitrogen enriched air (NEA) is available to inert the flammable vapors in fuel tanks, the capability could also be employed to use the same NEA to dampen a belly hold fire - or a fire in virtually any remote space in the airplane.

In the Air Canada case, the stark evidence of the fire's brief but rapacious appetite was plainly laid out, unlike the Swissair jet, which was shattered into thousands of pieces. "Fortunately, no one was killed" in the Air Canada fire, Stacey said. With the evidence so obviously at hand, he declared, "We have an opportunity to fix things."

Fire Propagated by Flammable Material
"The National Transportation Safety Board's investigation of the [Sept. 17, 1999] incident revealed that a spark from the right alternate static port heater ignited the fire and that the fire propagated by consuming the metalized Mylar covering the sidewall insulation blankets surrounding the heater. There was substantial damage to the insulation blankets and soot and heat distress to components in the area, such as some floor structure and a potable water tank.

"In response to the incident, Delta Air Lines [inspected] its entire fleet of MD-88s and MD-90s - 136 airplanes - and ... [found] that 11 percent of the airplanes had at least one heater installation that exhibited some kind of damage.

"Delta Air Lines removed the metalized Mylar-covered insulation blankets from around all of these heaters and ... plans to install metalized Tedlar-covered insulation blankets ... Delta Air Lines' Engineering Department further recommended that the heater be redesigned to provide a more generous bend radius to the thermostat lead wires and heater elements.

"The Safety Board also believes that the FAA should initiate a design review of the primary and alternate static port heaters on MD-80, MD-90 and DC-9 series airplanes and ... require design changes to reduce the potential for arcing."

Source: NTSB recommendations A-01-03 through-05, Feb. 6, 2001