RSS

Ground Fire on Cargo Plane Exposes Vulnerabilities

Mon, Jul 13, 2009 — David Evans

Articles, Featured

The electrical fire on the airplane in 2008 had so many features in common with a catastrophic fire in 1998, given the paucity of corrective action since, one wonders if the airline industry is capable of learning from past experience.

At a recent National Transportation Safety Board (NTSB) hearing, the subject of a 28 June 2008 ground fire on an ABX Air B767 cargo plane was discussed. The crew escaped before starting the engines, just moments before the entire forward part of the airplane was filled with noxious smoke and flame. This B767’s self-immolation calls to mind the inflight catastrophe that occurred on a Swissair MD-11 on 2 September 1998 during a flight from New York to Geneva. An electrical fire in that accident burned into the cockpit, and the aircraft crashed in Halifax, Canada, killing all 229 persons aboard. The aircraft was descending through 10,000 feet and about 10-15 minutes from achieving its emergency landing when fire burst through the cockpit ceiling, disabling vital flight instruments and incapacitating the crew.

The Transportation Safety Board (TSB) of Canada investigated. On the tenth anniversary of the crash, the TSB reiterated that most of its significant safety recommendations have not been implemented.

Now comes the ABX fire at San Francisco, with many parallels to the Swissair tragedy. On the Swissair jet, the fire was initiated by electrical arcing behind the cockpit that spread into the so-called “attic area” above the ceiling panels. As the fire spread, it burned through the end cap of an emergency oxygen line. The result? An oxygen-stoked electrical fire, perhaps the worst kind of emergency imaginable for an airplane in flight. The fire roared unfettered and inaccessible in an area not covered by built-in fire detectors or fire extinguishers, its extent and spread masked by the cabin linings and the ventilation airflows.

The ABX fire occurred up behind the cockpit, similar to the fire that destroyed Swissair flight 111 a decade ago.

The ABX fire occurred up behind the cockpit, similar to the fire that destroyed Swissair flight 111 a decade ago.

In the aftermath, Swissair fitted its remaining MD-11s with infrared cameras to detect fire in the attic space, and it also moved wires and components to provide increased separation, so that a failure of one item would not have cascading effects (see http://tinyurl.com/nfudn9). These changes, known as the Swissair “Modification Plus” program, were not mandated by regulatory authorities throughout the fleet of passenger and cargo aircraft nor fleetwide on MD-11’s. If they had been, the severe fire on the ABX plane would have been detected quickly and suppressed by an on-board extinguishing system.

Consider the similarities:

* The ABX fire broke out behind the cockpit, in an area ahead of the main deck cargo bay that is configured to carry three passengers, with a small galley and lavatory. This area is called the supernumerary compartment. The fatal Swissair fire broke out behind the cockpit and spread upward.

Flight deck of the accident aircraft; note spread of fire damage from the top down.

Flight deck of the accident aircraft; note spread of fire damage from the top down.

Supernumerary compartment on a sister ABX B767. Cockpit door is on the left. Panel covers removed from oxygen mask stowage panel on right.

Supernumerary compartment on a sister ABX B767. Cockpit door is on the left. Panel covers removed from oxygen mask stowage panel on right.

 

After the fire, looking aft in the supernumerary compartment. The fire consumed ceiling panels, only charred portions of which were found on the floor. The crown of the aircraft had burned through.

After the fire, looking aft in the supernumerary compartment. The fire consumed ceiling panels, only charred portions of which were found on the floor. The crown of the aircraft had burned through.

* Above the supernumerary compartment were the oxygen masks and steel oxygen supply lines, which piped pressurized oxygen from a tank in the lower bay. The Swissair fire burned through an oxygen line, accelerating the fire. As the NTSB Fire Investigation Factual Report describes:

“The stainless steel tubing would run vertically upward from below deck to above ceiling height on the main deck in the area forward and outboard of the galley. The tubing then would have crossed the overhead area above the ceiling panels and then dropped down into the bulkhead containing the oxygen mask stowage boxes … The oxygen cylinder located below deck, which would supply the supernumerary compartment oxygen system, was found empty. The cylinder was reported to have a pressure of 1,580 psig during the pre-flight check.”

Steel tubing runs vertically from the oxygen bottle in the hold, then crosses the overhead area above the ceiling panel and drops down into the bulkhead containing the oxygen mask stowage compartments.

Steel tubing runs vertically from the oxygen bottle in the hold, then crosses the overhead area above the ceiling panel and drops down into the bulkhead containing the oxygen mask stowage compartments.

* The oxygen line featured an electrically heated internal spring, apparently to warm the oxygen for human consumption. One should ask why an electrical component is fitted inside a potentially dangerous oxygen supply line. Elsewhere in its report, the NTSB asserts the string was “inadvertently” energized, which raises questions of circuit breaker protection that are not addressed.

o2-hose

* In addition, electrical wires were found contacting the oxygen line on other ABX B767s. As Bob Swaim, the NTSB’s resident electrical systems expert said: “Conductive oxygen tubing needs protection: physical isolation, nonconductive sleeving, and a nonconductive coating.”

Proxity of oxygen tube to electrical wires; proper installation on another ABX B767 aircraft.

Proxity of oxygen tube to electrical wires; proper installation on another ABX B767 aircraft.

improper-4 

The accident airplane had none of these things. A similar lack of protections were found in the Swissair wiring.

* The oxygen system was plagued by leaks, and on the accident aircraft was serviced 50 times between January 2007 and June 2008. Oxygen leaks were located only four times, but chronic problems still existed.

On the day of the accident, the crew was powering up the aircraft for takeoff. Extracts of the cockpit voice recorder indicate what happened:

22:06:58      First Officer David Hughes      Before start checklist complete.

22:0:27        FO Hughes                              Hey, there’s something going on in back. (According to Hughes’ interview with NTSB investigators on 2 July 2008, he checked the lavatory to make sure nobody other then he and the captain were aboard, closed the cockpit door, sat down in his seat and buckled up. He then heard a “muffled bang,” immediately followed by the sound of air flowing. The noise was coming from behind the two pilots.)

22:10:32      Captain Randolph Brooks        Uh oh. (According to Brooks’ 2 July 2008 interview with NTSB investigators, when the two pilots heard a “pop,” they looked at each other and said “What’s that?” The captain heard a “crescendo” of sound” and shut down the air conditioning packs.)

22:10:38      FO Hughes                              We got a fire.

22:10:40      FO Hughes                              We got a big fire.

22:10:40      Sound of cockpit door closing. (Hughes had unbuckled his seat belt and opened the cockpit door. Dark smoke was coming down from the ceiling. There was a glow in the upper side of the supernumerary area. Hughes reported to the captain, “We have a fire and got to get off the airplane.” The captain was thinking cargo fire.)

22:10:42      Capt. Brooks                          (Unintelligible words) Expletive deleted.

22:10:43      FO Hughes                              Whoa.

22:10:44      Sound similar to window being cranked open (probably the FO’s).

22:10:49      FO Hughes                              Declare an er- or a (expletive deleted).

22:10:56      Capt. Brooks                          Alright.

22:10:56      Sound of lavatory smoke detector alarm.

22:11:00      FO Hughes                             Want me to call ground roll the equipment?

22:11:01      Capt. Brooks                          Yeah, absolutely.

FO Hughes exited his window, using the rope. The captain discharged the engine and APU fire bottles, and he then jumped out of his window onto airstairs that had been pushed into position. The captain later said the smoke was getting very bad very quickly. It was seeping through the cockpit doorway.

Firefighters arrived within a minute, and they used a penetrator to pump foam into the first officer’s window. This foam did not suppress the fire, as it was not being applied to the upper portion of the supernumerary area.

At the NTSB’s hearing, Member Robert Sumwalt remarked, “The extinguishing agent was going outside the aircraft, not inside – for 6 minutes. But the fire was behind the cockpit and the cockpit was closed.”

As Acting Chairman Mark Rosenker properly remarked, “Had the fire started when the plane was in the air, the result would very likely have been catastrophic.”

Yes, it would have been very similar to Swissair flight 111, although the parallel was not mentioned by anyone at the hearing.

The industry still has unfinished business emanating from the Swissair accident report; here are a few ABX-relevant items that, five years after the Safety Board of Canada’s final report, are still wanting action:

— Materials used in occupied areas (the supernumerary compartment in this case) need to be self extinguishing. No testing has been required for toxicity (both Capt. Brooks and FO Hughes commented on the strength of the smoke and fumes).

— The FAA-required systems safety analysis did not then and does not now include an assessment of a system’s failure as a result of a fire in progress.

— Regarding wire separation and routing, regulations require that a potential threat be minimized; they do not require that the potential threat be eliminated. The term “minimized,” by the way, is not defined.

— The proposed arc fault circuit breaker (AFCB) technology will improve arc detection, but the performance criteria do not limit the arc energy to a level below the ignition of flammable materials. Such testing criteria have been established for residential AFCB, but they have yet to be mandated for aircraft applications.

The NTSB concluded in the ABX fire:

“Crew descriptions about what was heard when the fire started, combined with Safety Board testing, revealed that the ignition source had to be within the oxygen hose. The Safety Board’s investigation determined that a short circuit to the supplemental oxygen system reached the oxygen hose. The design of the hose included an internal spring, which could be heated by the inadvertent application of electrical current, causing the plastic hose to ignite. Safety Board testing found that the hose design brought together the three elements for a fire: the coil acting as an ignition source, the hose material acting as a fuel, and the oxygen to promote burning ….

“The Safety Board also found that other ABX 767 aircraft’s supplemental oxygen system did not include positive separation between electrical wiring and oxygen system tubing. Electrical wiring that is near or in contact with oxygen system tubing creates the potential for electrical short circuits to reach the oxygen system hoses. The involvement of oxygen in a fire can significantly expedite its growth and severity.”

Among its dozen recommendations, these are salient:

— Require operators to replace electrically conductive combustible oxygen hoses with electrically nonconductive hoses so that the internal hose spring cannot be energized.

— Require airplane manufacturers and operators to ensure that oxygen system tubing in proximity to electrical wiring is made of, sleeved with, or coated with nonconductive material or that the tubing is otherwise physically isolated from potential electrical source.

— Develop minimum electrical grounding requirements for oxygen system components that include these requirements as part of the certification process for new airplanes and approved supplemental type certification modifications to existing airplanes.

— Require operators of transport-category cargo airplanes to install smoke detectors in the supernumerary or similar compartments of their airplanes.

Not a word about AFCB technology, which may well have prevented the arcing?

Not a word about how an oxygen hose can be electrically energized in the first place, and how this potential problem must be minimized or eliminated during the electrical system design phase. During this design period, a fault tree analysis is supposed to be performed and hazards must be eliminated. As the NTSB noted, this case featured an ignition source, combustible material and concentrated pure oxygen. One wonders if the design was subjected to a fault tree analysis or, if it was, how the design passed muster.

Not a word about fire suppression in the supernumerary compartment or in the cockpit. Yes, there’s a call for fire detection in the supernumerary compartment, and this would have helped alert the crew. But the fire grew quickly and was obviously not going to be suppressed by a hand-held extinguisher. The time from detection to evacuation of the aircraft was on the order of 90 seconds. In the air, without a built-in fire suppression system, a fire detection alarm would simply provide notice of an imminent death sentence. The airplane could not be safely landed in 90 seconds.

There are many latest and obscure initiators of cabin and hold fires. Grease or oil and oxygen is a known deadly combination, as is wiring. The last noteworthy B767 internal fire occurred in 2002, and was caused by an overheating water-line heater (see http://tinyurl.com/kt584f). In that instance, the TSB of Canada addressed the specific cause (as did the FAA after further such water heater fires), but no investigative or regulatory agency has thus far addressed the issue generically by recommending, or even suggesting, that fire detection and suppression should be de rigeur. It’s not as if there’s a fallback position of ejection, as a military pilot would have on a tactical aircraft.

Fires that develop as fast as on Swissair 111 and on this ABX B767 are an in flight killer from any perspective. If they don’t cripple the airplane, they can still incapacitate the pilots. Admittedly, the pilots can don oxygen masks, but fires generate thick smoke and pilots still need to be able to see their instruments and run checklists. Survivability becomes a matter of freakish chance once there’s an outbreak in an inaccessible area. For cargo pilots, the question becomes whether they should leave their seats and fight the fire (assuming the oxygen mask umbilical is long enough to permit such mobility). “Fight or flight” takes on an entirely new meaning.

The stubborn absence of a recommendation or consideration for fire suppression is mystifying, given the development in recent years of low-pressure water mist system of proven effectiveness in squelching aircraft cabin fires.

The NTSB had the opportunity in this investigation to offer sweeping recommendations regarding fire protection of both cargo and passenger airlines. It chose to issue a “tactical report,” if you will, focusing on the particulars of this accident. It missed the opportunity to issue sweeping recommendations, to the effect that not one square inch of a modern jet ought to be left unprotected by fire detection and suppression.


Comments are closed.

Nolan Law Group