WHAT THE FAA KNOWS
What the FAA Knew.. and When
ABOUT AGING WIRING
John D King 12/26/97
With the NTSB's TWA 800 week long hearing behind us now the public is left the disturbing news that aging aircraft not only suffer from corroded airframes but aged wires as well. Many now await the FAA's promised June 1998 response as they claimed to be surprised by this phenomena. The NTSB's wire panel expert, George Slenski cast the most light on this and its clear from his posted exhibits; "Causes Of Aircraft Electrical Wiring" and "Development And Analysis Of Insulation Constructions For Aerospace Wiring Applications" [NTSB Exhibit 9C, page references 000005 through 000033] that he has extensive experience.
In response to a question, Mr. Slenski also acknowledged the work of other industry colleagues; FAA's Patricia L. Cahill was one. As with Mr. Slenski's exhibits written in the early 90s, Ms. Cahill's work PREDATES this hearing even further, by about 5 years.
Ms. Cahill's reports are even more descriptive than Mr.Slenski's; so much so, I wonder that since she is an FAA employee, will the FAA be including her reports as they prepare for that next June report? I appreciate what the FAA has already done and has already known for some time, excerpts of three of Ms. Cahill's reports follow: REPORT NO. DOT/FAA/CT-88-4\line AIRCRAFT ELECTRICAL WET-WIRE ARC TRACKING. Author, Patricia L. Cahill and James Dailey. August 1988. Final Report 21 pages, Sponsoring Agency, FAA Tech Center, Atlantic City, N.J.
"Electrical wet-wire arc tracking is a phenomenon that has been known for many years. This can occur when leakage currents on a wet insulation surface are great enough to vaporize the moisture, resulting in the formation of dry spots. These dry spots offer a high amount of resistance to current flow. In turn, an induced voltage will develop across these spots and result in the occurrence of small surface discharges. Initially, these discharges will appear as scintillations at the insulation surface. These discharges produce highly localized temperatures on the order of 1000 degrees Centigrade. Temperatures of this magnitude will cause thermal degradation of the insulation material, the nature of which depends on the insulation material.
The Federal Aviation Administration [FAA] conducted a series of bench scale tests which demonstrated that the ability of an aircraft wire to resist wet arc tracking and possible flashover is highly dependent on the composition of the wire insulation. In addition, the conductivity level of the electrolyte may influence the time and the type of failure [arc track or open circuit] that can occur.
Editors Note: In BENCH tests, fluids commonly found on the aircraft and when collecting on aircraft wiring lead to failure of the wire insulation. Certain commonly used insulations fared less well than others.
REPORT NO. DOT/FAA/CT-89/21 FLAMMABILITY, SMOKE AND DRY ARC TRACKING TESTS OF AIRCRAFT ELECTRICAL WIRE INSULATIONS. Author Patricia L. Cahill July 1989 Final Report 21 pages. Sponsoring Agency, FAA Tech Center, Atlantic City, N.J.
John D King 12/26/97
With the NTSB's TWA 800 week long hearing behind us now the public is left the disturbing news that aging aircraft not only suffer from corroded airframes but aged wires as well. Many now await the FAA's promised June 1998 response as they claimed to be surprised by this phenomena. The NTSB's wire panel expert, George Slenski cast the most light of this and its clear from his posted exhibits; "Causes Of Aircraft Electrical Wiring" and "Development And Analysis Of Insulation Constructions For Aerospace Wiring Applications" [NTSB Exhibit 9C, page references 000005 through 000033] that he has extensive experience. Horizontal and vertical: of course everyone knows that a roll of burning newspaper will burn fastest when held vertical. As in the earlier Cahill tests, WET or DRY - the arc problem PERSISTED in certain older but commonly used wire insulations.
REPORT NO. DOT/FAA/CT-TN94 / 55 ELECTRICAL SHORT CIRCUIT AND CURRENT OVERLOAD TESTS ON AIRCRAFT WIRING. Author Patricia L. Cahill March 1995. Final Report 13 pages. Sponsoring Agency, FAA Tech Center, Atlantic City, N.J.
"Although three different laboratory-scale tests were evaluated in this wire program, only the sixty-degree test is currently required by the Federal Aviation Administration [FAA]. All test specimens with the exception of MIL- W5086/1-PVC nylon passed this test with average burn lengths within the 3-inch maximum and no flame time. The MIL-W-5086/1 samples marginally passed the 30-second flame time, and the average burn length was greater than the 3-inch maximum specified in the FAR. The smoke test method used in this program called for a straight pilot burner when testing insulated conductor specimens. However, data for multi-directional pilot burner were also included in this report. Large variations in Ds occurred between the two burners for ETFE constructions at both the 5- and 20- minute test points. The MIL-W-81381/21 aromatic polymide and the composite construction [Teflon/ polyimide/ Teflon or TPT showed no appreciable difference in Ds between the two burner types. Moreover, test duration did not affect smoke generation for these two samples. A direct correlation can be seen between dry arc tracking tests and wet wire arc tracking tests [DOT/FAA/CT-88/4]. The halogenated polymers formed no conductive chars upon thermal degradation and, therefore, no dry arc tracking. The MIL-W-81381/12 aromatic polyimide sample formed a conductive char upon thermal degradation, and severe arc tracking occurred. Extensive damage to all wires in the bundle occurred due to arc tracking propagation upon circuit-breaker resetting. The TPT composite construction performed well. No dry arc tracking
Editors Note: In another series of BENCH tests, certain new types of wire insulation greatly diminished the problem of the effects of arcing - burnt wire insulation, burnt wire bundles and varying degrees of smoke. One type of commonly used insulation greatly exceeded the FAA's own limits. Here, the wire was only held at an angle slightly above and between horizontal and vertical; of course everyone knows that a roll of burning newspaper will burn fastest when held vertical. As in the earlier Cahill tests, WET or DRY - the arc problem PERSISTED in certain older but commonly used wire insulations.
REPORT NO. DOT/FAA/CT-TN94 / 55 ELECTRICAL SHORT CIRCUIT AND CURRENT OVERLOAD TESTS ON AIRCRAFT WIRING. Author Patricia L. Cahill. March 1995. Final Report 13 pages. Sponsoring Agency, FAA Tech Center, Atlantic City, N.J.
"A direct correlation can be seen between dry arc tracking tests and wet wire arc tracking tests [DOT/FAA/CT-88/4]. The halogenated polymers formed no conductive chars upon thermal degradation and, therefore, no dry arc tracking. The MIL-W-81381/12 aromatic polyimide sample formed conductive char upon thermal degradation, and severe arc tracking occurred. Extensive damage to all wires in the bundle occurred due to arc tracking propagation upon circuit-breaker resetting. The TPT composite construction performed well. No dry arc tracking was evident. This construction behaved similarly to a halogenated polymer in this respect.
Editors Note: In another series of BENCH tests, certain new types of wire insulation greatly diminished the problem of the effects of arcing - burnt wire insulation, burnt wire bundles and varying degrees of smoke. One type of commonly used insulation greatly exceeded the FAA's own limits. Here, the wire was only held at an angle slightly above and between "This document describes the electrical short circuit and current overload tests that were conducted on wires used in commercial transport category aircraft. This testing was conducted to evaluate the fire potential that may result from electrical faults. Results of this testing showed that circuit-breakers may not protect wire from ticking faults but can protect wire from direct shorts. It also showed that circuit-breakers may not safeguard against the ignition of flammable materials by ticking faults. Preliminary testing also indicated that direct short circuits are not likely to start a fire and that direct short circuits do not erode insulation and conductors to the same degree that ticking faults do.
Editors Note: The aircraft internal protection, circuit-breakers, did the job well, but largely only for DIRECT short circuits. The more dangerous arcing and resultant insulation fires did NOT reliably trip circuit-breakers. An unknowing crew would many times simply reset these breakers and thus greatly INCREASE the chance of wire BUNDLE FIRES. So important was this information about the FIRE potential resulting from electrical faults on transport category aircraft, this Cahill Report, and as an illustration, included three fires that have occurred during the past several years. An excerpted and more brief description of each follows:
1. On January 18, 1990. a US Air MD-80, en route to Cleveland from Buffalo, was forced to return to Buffalo when the cockpit filled with smoke from overheated electrical wire insulation. The left generator tripped off-line and the captain turned the right generator control switch to the "Off' position. He selected emergency power and initially was able to clear the smoke. The captain then started the auxiliary power unit [APU] and the cockpit again started to fill with smoke. The A.P.U. electrical power was then shut off and the emergency electrical power was turned back on. The aircraft returned to Buffalo with no further reports of smoke. It was found that the left generator phase B power feeder cable terminal, which is connected to a plastic terminal strip, had melted from intense arcing . The terminal, approximately, 15 inches of the cable, and the terminal stud had melted. The second source of smoke came from a fire started by the molten metal that sprayed an area forward of, and below the terminal strip. The only circuit-breaker to trip was the cabin temperature control. This incident was caused by improper torquing of the phase B terminal.
2. On March 17,1991,a Delta L-1011 en route from Frankfurt, Germany, to Atlanta, Georgia, was forced to make an unscheduled landing in Goose Bay, Labrador, Canada. About 7.5 hours into the flight, flames erupted from the base of the left cabin sidewall panel to the height of the seatback tray at the next to last row of passenger seats. The fire was extinguished and a precautionary landing was made. The ignition source of this fire was not determined; however, a possible source of ignition appeared to be an electrical fault. Some of the wires in a fifteen wire bundle located in the fire area exhibited evidence of arcing. Five circuit-breakers connected to this wire bundle had tripped.
3. On November 24,1993,an SAS MD-87 experienced smoke and a subsequent fire upon touchdown. The fire damage was severe, including a 1-foot-diameter hole in the skin. Investigation found that two wires, one 115 volts [V] and one 29V, had been pinched together and were arcing to the fuselage structures. Neither the 10-ampere [amp] circuit-breaker [28V line] nor the 15-amp circuit-breaker [115V line] tripped.
Editors Note: The normal aircraft wire protection, circuit-breakers, do not protect against arc tracking and arc track fires. Between 1991 and 1997 ELEVEN other smoke of fire related reports were easily found in the NTSB's accident database at Http://www.ntsb.gov. [CLICK "Site Map", CLICK "Aviation", CLICK "Accident Synopses", CLICK "Select Monthly Lists"] The Valujet type fire/smoke reports continue. Consider the following reports from that NTSB database.
1. NTSB Identification: CHI97IA195. Scheduled 14 CFR 121 operation of NORTHWEST AIRLINES, INC. Incident occurred JUL-05-97 at FLINT, MI. Aircraft: McDonnell-Douglas DC-9-30, registration: N963N Injuries: 52 Uninjured. On July 5, 1997, at 1544 eastern daylight time, a McDonnell-Douglas DC-9-30, operated by Northwest Airlines, Incorporated, as Flight Number 1446, and piloted by an airline transport pilot certified crew, declared an emergency when the cockpit filled with smoke. Instrument meteorological conditions prevailed at the time of the incident. The 14 CFR Part 121 flight was operating on an instrument flight plan. The captain, first officer, two flight attendants, and 48 passengers reported no injuries. The flight departed Milwaukee, Wisconsin, at 1350 central
daylight time. Flight 1446 landed at the Bishop International Airport, Flint, Michigan, without incident. The passengers were deplaned through the concourse jetway. According to the captain, he and the first officer noted smoke coming from under the center instrument panel glareshield. The captain said the smoke was billowing out from under the center panel. He said the first officer and he put on their smoke goggles and oxygen masks and proceeded with the emergency checklist. The smoke immediately stopped when the generators were taken off line, according to the captain. He said the cockpit smoke had cleared completely within about 4 minutes. The smoke was confined to the cockpit and did not enter the passenger compartment.
2. NTSB Identification: FTW97IA227. Nonscheduled 14 CFR 121 operation of SUN COUNTRY AIRLINES [D.B.A. SUNCOUNTRY AIRLINES, INC.] Incident occurred JUN-17-97 at DENVER, CO Aircraft: McDonnell-Douglas DC-10-10, registration: N572SC. 24 Uninjured. On June 17, 1997, at 2009 mountain daylight time, a McDonnell-Douglas DC-10-10, N572SC, operated by Sun Country Airlines as flight 1121, made an emergency landing at Denver, Colorado, International Airport after the crew reported smoke in the aircraft. There was no damage to the airplane. Two persons were admitted to Denver General Hospital for observation and released. Visual meteorological conditions prevailed and an IFR flight plan had been filed for the flight conducted under Title 14 CFR Part 121. The flight was a company repositioning flight from Minneapolis, Minnesota, to Los Angeles, California. The flight was carrying 9 flight attendants and 12 deadheading crew members with company material [aircraft parts] and 1795 pounds of mail in the cargo hold. The cockpit crew reportedly donned their masks, as per operating procedure. It is unknown if the passenger oxygen system was activated.
Since 1991, THREE other ValuJet, DC-9 events were also found at the following Month/Day/Years; 2/1/91, 1/1195, 2/20/97. For OTHER type aircraft check - 9/28/91, MD-82, 1/3/92, 737, 4/26/94, 737, 8/13/94, C-130, 6/6/96, Beech 1900, 6/25/96, 767, 10/10/96, DC-10and 12/11/96, 757.
NTSB DATABASE DELETIONS
The three illustrative events listed in the Cahill 1995 report above were not found ! Neither were the balance of the 264 such events referenced by FAA's Thomas McSweeny during his 1993 Senate testimony [Hearing, 103-397, 11/8/93:, prepared statement of Thomas McSweeny; "We also looked at 264 cases, from 1986 through 1992 that involved smoke, fumes or fire our review disclosed that in none of those cases was the lack of pilot visibility of the instruments or the outside of the aircraft a factor in preventing continued safe flight and landing."
All the above documents plus the testimony to two of our industry wire insulation experts are available now. - John D. King FAA mechanic, lic #1552888 -
UNITED AIRLINES FLIGHT 965: ARC-TRACKING
Aviation Week, Feb. 9, 1998, page 77 refers to a confirmed Kapton "arc tracking" failure of Kapton wire aboard United Airlines 767 ETOPS Flight #965 which was en route from Zurich to Washington. It had to land in London because of an in flight electrical fire. AAIB investigated the January 9 incident and concluded that the problems began shortly after take-off from Zurich when a series of apparently unrelated electrical problems began occurring - according to the flight crew reports. The crew elected to abort the flight and divert to London as they were west-northwest of Paris. Cabin attendants reported smoke in the first class cabin and gallery. After bringing the aircraft to a stop on the runway at Heathrow, the captain ordered an evacuation. AAIB investigators found about 10 CBs open. It was not certain if they opened in flight or on the ground. They also found an approx. 7 inch long section in a bundle of more than 100 wires that were severely burned and melted. The bundle was in the electrical and electronics bay of the 767-322ER directly below the first class galley. About 36 wires in the bundle were damaged by heat or fire. The twisted strands in one wire were fused into a single strand of copper, indicating its exposure to sustained high temperatures. There were copper globules in the damaged area, indicating active arcing occurred there. All of the damage to the bundle occurred on the inside bend of the bundle's curve over the top and down the side of a refrigeration unit. There was no evidence that the fire or heat extended up around the circumference of the bundle to top of the bundle. The exterior of wires on another bundle about an inch away suffered thermal damage, as did foam on the rear wall of the refrigeration unit. The unit, an 86 pound chiller for the first-class galley, had recently undergone maintenance. Investigators are trying to determine whether mechanics may have nicked insulation on a wire in the damaged bundle in the process of maneuvering the chiller out of or into its perch in the electronics bay. Most of the wires in the damaged bundle used ETFE insulation, but all of the individual damaged wires used Kapton. NOTE: When I met with Bob Francis in Washington in early August, I asked him about this incident. He told me his staff contacted Boeing and Boeing reported that this was some kind of a "fluke" because they were not using Kapton anymore. In this instance, it is one of those cases where someone had wire on the floor and used it, not knowing it was Kapton."
737 Total Electric Power Loss
On December 15, 1998, about 1216 eastern standard time, a Boeing 737-232, N327DL, registered to Wilmington Trust Company Trustee, operated by Delta Air Lines Inc., flight 2461, as a 14 CFR Part 121 scheduled domestic passenger/cargo flight, experienced a reported total loss of electrical power on approach for landing at Orlando International Airport, Orlando, Florida. Visual meteorological conditions prevailed and an IFR flight plan was filed. The airplane sustained minor damage. The airline transport-rated pilot-in-command (PIC), first officer (FO), 3 flight attendants, and 51 passengers reported no injuries. The flight originated from Boston, Massachusetts, about 3 hours 6 minutes before the incident. The PIC stated they were cleared for a visual approach to runway 35 at Orlando International Airport. The FO was flying the airplane. All descent and approach checklists were completed. The APU was started on base leg. The FO called for the gear and flaps 15. The airplane experienced a total loss of electrical power as the gear and flaps were extended. The APU did not start, and the battery indicated between 17 to 18 volts. The normal checklist procedures were accomplished followed by the quick reference procedures. Electrical power was unable to be restored. A go-around was initiated to continue the checklist. All communications and electrical equipment failed. The flap indicator indicated an asymmetrical setting. A flight controllability check was accomplished with no anomalies. The flight was continued and landing was made to runway 35. A left main landing gear tire blew out on rollout. The airplane cleared the runway, stopped, and was towed to the ramp.
A Kaptonitis smoking gun? At that voltage (17 to 18V) it sounds as though my theory on sr111 just came to life. They'd have been history if it had happened at night or IMC conditions (or both). The "normal checklist" (I assume) would be all about checking CB's and attempting to reset Generators. After that (logically) would come the (second?) APU start attempt. It sounds as though they probably had a dead short in the electrics (i.e. wiring) associated with the GEAR or FLAP extension (however it may have had something to do with the switchover priority between the APU gen and Gens 1 & 2 - wonder why they start APU on base leg?). With a dead short to the batteries I just don't think you are going to get a Gen reset (i.e. relevant to sr111 crew rotating Smoke/Elec switch - deselecting one generator and never getting the one associated with the next selection because of the dead short to the battery caused by the wiring fault). This was a very lucky planeful.
List of Annexes