Identification (Kind of event): Serious Incident
Date: 19. August 2003

Place: Frankfurt/Main

Aircraft: Airliner manufacturer/model: Boeing/B747-430 (Lufthansa registration is D-ABTH. B747-430M)

Personal injury: without any person hurt or damage to property

Airliner: minor damages

Third party damage: nil

Source of information: Investigation by BFU of circumstances

Summary: Around 12:30L on approach to runway 25L in Frankfurt/Main (about 3 to 5 minutes before landing) white smoke from behind the left lower cabin side panel of passenger row 24 was noticed and pointed out to the cabin crew. A member of the cabin crew immediately accomplished an examination and discovered a fire source within the lower part of the cabin liner beside passenger seat 24A as well as heavy smoke development. The fire was extinguished after the cabin crew used two portable extinguishers following removal of the lower side panel lining. The flight crew declared a Mayday. Since all announcements to the cockpit by the cabin staff indicated that the situation had been controlled and there was no smell in the cockpit, the approach and landing was accomplished problem-free. After the landing no further smoke development arose, so the flight crew decided to taxi to parking and to disembark the passengers normally.

The aircraft was in the airliner public transport category and the total period of operation of the airplane amounted to 60,512 hours at that time. Installation of the Inseat video Entertainment System into this airplane, which offers to the passengers in First and Business Class a video transmission (television) (the necessary monitors are stowed away and/or fastened in the arm rests of the seats), was carried out by a development enterprise assigned by the airline in the context of a minor modification conceived and later built into the airplane by way of an STC modification. While the main individual units of this system came from a Japanese electronics firm, the "Purser's control center" was supplied by an American enterprise.

This Entertainment system is not incorporated in tourist class. With this Entertainment system the video signals and the current supply (115 V alternating voltage) pass through a distribution box (AREA distribution box/ADB) under control of the flight crew station to a wired-in video-audio Seat electronics box (VASEB) in the seats. The last terminating VASEB must be provided in each case with a final resistance plug (P/N RD-AA2902), in order to lock the DATA LINE. The plugs' contact pins mate with the female ports of the plug admission sockets. Universal plug connectors, similar to Submin railways 15, are used with Gemischtbestueckung(1 HF line). In the tourist class there are no VASEB's in the seats.

With this airplane the final resistance plug with the change incorporated by the STC was extended on 11.07.2003 from the Business Class into tourist class. However this was accomplished via a plug-in connection, which was located at the left lower cabin side panel at pax seat row 24. The plug connection is thereby in a position where sub-lining condensation is expected to be encountered.

Meteorological information prevailing at the time of this serious incident in Frankfurt/Main was in accordance with the flight crew's briefing. The following weather conditions prevailed:

Daylight, wind variable at 7kts, visibility 5kms

Cloud Cover: 1 to 4 eighths in 500 ft

GND precipitation: nil

temperature: 19 deg C

Navigation aids: standard runway 25L ILS approach to Frankfurt/Main via a serviceable instrument landing system (ILS).

Radio traffic: Radio traffic between the approach control place Frankfurt/Main and the crew was available to the BFU as a tape transcription. It contained data on the airport Frankfurt/Main (EDDF) elevation 364 ft asl. The length of runway 25L (concrete) is 4,000 meters and the width is 45 meters. The radar data recording of the flight route as noted by the German air traffic control (DFS) was placed at the disposal of the BFU for evaluation . The evaluation of this data did not cast any fresh light on the incident.

Accident scene: Examination and statements made to the investigators at the aircraft resulted in a conclusion that a fire at the patch cord had arisen(cabin-sided plug admission Eqm NO. D19003J / final resistance plug P/N RD-AA2902). The fire was located behind the left lower cabin liner beside passenger seat 24A The final resistance plug and the plug admission ports showed clear fire-effect traces. Likewise the lagging of the cabin lining in this vicinity exhibited clear fire effects as well as smoke-damage. In this location two different insulation blankets were used: a green and a yellow type, in each case with an attached transparent foil film. The technical investigation of the plug connection parts and the insulation blankets concerned was carried out in a laboratory under supervision of the BFU.

The plug (P/N RD-AA2902) exhibited a one-sided fire hole within the range of being between pin 1 and 2. This application uses a permanent AC voltage of 115 V, but max current is limited to 5 amperes. The plug housing was molten and the plastic mounting for the pins was carbonized. Pins 1 and 2, which were embedded within this plastic surface together with further pins, were missing (i.e. melted away). The edges of the holes, where under normal conditions pin 1 and 2 would be mounted, showed clear evidence of molten metal. These pins had been in the sockets of the female plug (Eqm No. D19003J). The plug ports exhibited clear fire-damage in the area of the sockets for pin 1 and 2. The plastic surrounds for the sockets of pins 1 and 2 were similarly carbonized. In these sockets (contact bushes) is pin 1 and 2 of the plug. Pin 1, which had melted into the socket, showed clear fusion traces. Pin 2 showed white deposits and with possible evidence for shorting over a limited range . The investigation of the plug admission port under the scanning electron microscope (SEM) showed melting and/or fire-caused oxide formation at both pins. Electrical breakdown characteristics could not be determined. Following withdrawal of the two casings and pins from their plastic surrounds, it was clearly recognizable that both pins had melted electrically.

An analysis of the deposits found at the pins, in particular with pin 2, as well as within the surrounding range showed, among other things, bromine. That obviously originated from the fire-extinguishing agent sprayed on the back of the contact holders of both the plug socket and power-plug. No other non fire-related deposits were found. In addition a plate tongue of the plug safety retaining device was bent (see photo 3). The insulation blankets found in the airplane at the fire's location were submitted to a fire test in accordance with FAR 25.853. With this test the mats were subjected to 12 seconds exposure of a temperature of 845 ˚C with a Bunsen burner . The following observations resulted:

A): Within zero seconds of flame application to the mat with transparent foil, the insulation blanket becomes green and shrivelled with the foil showing no fire effect,

B): yellow insulating blanket with transparent foil insulation, the blanket did not burn, but the foil burned, with flame applications over a period between 38 and 22 seconds.

Thus the insulation blanket tested under para B above does not fulfill the latest valid airworthiness specifications. These were not not yet in force at the time of type certification of the airplane model. The age of the insulation blankets found in the airplane can not be ascertained since these parts are not series-pursued and logged. The aircraft manufacturer uses a material for insulation blankets in accordance with the aircraft manufacturer's specification BMS 8-48 and for the foil, a material in accordance with BMS 8-142. An indication concerning the temperature at which the thermal-acoustic insulation material would begin to burn could not be given. The technical investigation of the distributor box (P/N RD-AA20001-60) in accordance with CMM ATA 23-32-51 with the maintenance enterprise of the aviation enterprise did not result in any technical objections. Visual examination of exit port J5 at the distributor box leading to the burnt plug connection disclosed no associated damage.

It was determined (in the context of the investigation carried out locally) that the 15-Ampere circuit protection device, which guarantees the current supply to the distributor box, had not tripped. The maximum permitted output current from the distributor box to the respective plug connection may not exceed the value of 5 amperes in accordance with the equipment specification. This is supposed to be guaranteed by a special circuit within the distributor box. The installation of the video Entertainment system was classified for the development enterprise assigned by the airline as a "minor change" (i.e. minor modification) (LuftBO § 12). Thus the Federal Office of Aviation was not responsible for granting approvals for this system. i.e., the permission, the installation and the acceptance of this system were all accomplished by the installing and designing airline on their own recognizance. The LBA became aware of this modification in the context of an airline announcement. Upon becoming aware of the modification, the LBA is always within its rights to ask for further information, but did not do so in this case.

Fire: The fire was contained within the proximity of the left lower cabin cowl beside pax seat 24A and was promptly extinguished during the flight's final approach by exhausting two portable fire extinguishers. On 26.07.2002 additional information was provided by the manufacturer for the video Entertainment system. A service bulletin (SELF-SERVICE) with the number RD-AA2902-23-001 was published for prevention of short-circuits by avoiding penetration of air-con condensation into the plug connection.

As a primary measure, the airline concerned decided that as operators, it should accomplish and respond to the bulletin (SELF-SERVICE) No. RD-AA2902-23-001 of 26.07.2002 of the video Entertainment system manufacturer. This required that the plugs (P/N RD-AA2902) should be fully inserted and locked in place. The modification required the sealing of pin 1 and 2 at the rear side of the plug with silicone. A silicone disk (P/N: insert R6HR1508ZA) was also required between the plug socket and plug. The silicone disk is required to be pushed fully onto the pins. Subsequently, a modification label with the designation "MOD 1" is to be affixed to the plug. The Federal Office of Aviation reacted to this incident with the publication of an instruction for airworthiness (airworthiness directive 2003-270 of 19.09.2003). This required an inspection of the IFI system jumpers, patch-cords and connections in accordance with the engineering order (EO) NR. 140168 and 140351 by the maintenance department of the airline concerned. It required inspections, repairs and modifications to the system. The modification stipulates an exchange of the plugs P/N RDAA2902 for plugs with the same part number and the additional marking "MOD 1". The measures were to be accomplished within 7 days after publication of the airworthiness directive. Only airplanes of the type Boeing 747-400 were concerned with this airworthiness directive.

The EO NR. 140351 of 10.09.2003 served as a temporary solution, in order to switch the wires for the plug connection behind the lower side panel lining up to the junction box. The plugs (RD-AA2902) were to be attached so that, at the distributor box (ADB # 1 and ADB # 3) at the port J5, no great residual tension acted against the plug (see photographs behind the lower cabin liner (Dado panel)). This was to be accomplished regarding the plug (P/N RD-AA2902) in accordance with EO NR. 140168 of 01.09.2003 (published by the maintenance department of the airline concerned) with the default status being observation of the service bulletin No. RDAA2902-23-001 of the video Entertainment system manufacturer. Further amendment to this EO was provided in EO NR. 140185 which required that, additionally to the inhouse modification, the complete rear side of the plug back-shell and the plug socket ports were to be sealed with silicone (to exclude entry of condensation).

Analysis: The fire behind the lower cabin liner, which was noticed by a passenger and pointed out to the cabin crew, was not indicated to the flight crew in the cockpit (i.e. no tripped circuit-breaker). The fire started due to a short-circuit in the internal area of the plug connection (cabin-side of plug entry port Eqm NO. D19003J/final resistance plug P/N RDAA2902) for the video Entertainment system. This was located in the vicinity of passenger seat row 24 behind the lower cabin liner dado. Since this plug connection comes into contact with the insulation blanket between the external lining of the airplane and the cabin liner inside paneling, it can be exposed to heavy condensation flows. It was therefore always highly probable that condensation water would penetrate into the plug connection and that dirt deposits in this area would allow a short-circuit between pins 1 and 2 of the final resistance plug (permanently carrying an alternating current voltage of 115 V). Penetration of the condensation was guaranteed by the fact that both the cabin-side female plug and the final resistance plug's backshell were open and not sealed.

An arc or a flash led to a short-circuit between the pins and a hole was burned into the housing of the final resistance plug. This led to a fire in the general area of this plug connection. Entry of condensation into an unprotected plug carrying 115VAC virtually guaranteed a short and a fire. There are no recorded instances of problems occurring within this type plug connection in any circumstance unconnected with water ingress or condensation. Therefore this type plug, if sealed, is not considered to represent a problem.
To what extent the insulation blanket surrounding the area in question could have burned cannot be guessed at. Due to the fire effect seen on the insulation blanket, at least the foil fastened to the thermal-acoustic material was burned away. Up to the time of this serious incident these insulation blankets had, according to the FAR regulation NR. 25.853, only been tested for exposure to a temperature of 845˚C. In this case however due to the hole burned through the metal screen of the final resistance plug between pins 1 and 2, there was a high probability that temperatures present were well over 1000˚C. It was due only to the rapid and sensible reaction of the cabin crew that this fire was not able to spread further.

The Federal aviation administration (FAA) on 02.09.2003 strengthened the regulations for thermal-acoustic insulating material. Via its "Improved Flammability standard for Thermal/Acoustic Insulation of material - Used in transport Category of airplane" the American authority hopes to have resolved the problem of the flammability or continuous burning of insulation blankets in airliners. According to this regulation the insulation blankets now must meet a standard (among other things) of 30 seconds exposure to a temperature of 1038 ˚C +/- at an ambient temperature of 56 ˚C. In order to avoid further incidents of this kind, the Federal Office of Aviation immediately published the FAA instruction for airworthiness (airworthiness directive) as an NR . 2003-270 of 19.09.2003 and at the airline, its engineering Orders (EO) No. EO 140168, EO 140185 and EO 140351 including the changes in the service bulletins of the video Entertainment system manufacturer. Due to the measures mentioned herein - sealing with silicone at the rear side of the respective cabin-side plug entry ports and of the final resistance plug as well as inserting one Silicone disc between the female and male plugs - any otherwise highly probable penetration of condensation in this area should have been now excluded.

Even if in accordance with European aviation Safety Agency (EASA) part 21, an authorized developer (DAS) might consider that only a minor modification is involved; nevertheless this incident raises the question of "to what extent or latitude" the developer can be allowed to objectively make this estimate. The development process that establishes this estimate must always be justified on a purely technical basis.

Conclusions: The accident is attributed to the fact that humidity could readily penetrate into the plug connection behind the lower cabin cowl (dado). This water ingress led to a short-circuit between two pins in the final resistance plug, whereby it came to arc, and this process initiated a fire in this area.

Investigation leader Friedrich (co-operation Krupper / co-operation Pitz)

http://www.iasa.com.au/folders/sr111/747IFE-fire.htm

My thanks to IASA for the above information.