On 6 January 2011 an Airbus AIRBUS A-319 being operated by Easyjet on a scheduled passenger flight from Liverpool to Belfast Aldergrove experienced the sudden onset of cabin air contamination with a smoke-like appearance as the aircraft cleared runway 25 at destination after a night landing in normal ground visibility. After the smoke appeared to be getting thicker, the aircraft was stopped and an evacuation was carried out during which one of the 52 occupants received a minor injury.The aircraft was undamaged.
A Field Investigation was carried out by the UK AAIB. It was noted that the aircraft commander was experienced in command on the aircraft type and had been PF for the incident flight. It had been noted from the Automatic Terminal Information Service (ATIS) en route that the air temperature and dew point at destination were both -3º, the wind was calm , the visibility was 2900 metres in mist with no significant cloud and runway 25 was reported as ‘wet’ throughout its length. The runway and taxiways at destination had been treated with potassium acetate and urea during the day so as to prevent ice formation overnight but this information was being not communicated to inbound aircraft. The landing was normal with pre-selected autobrake ‘LO’ and idle reverse used. After landing, ATC asked the aircraft to continue to the end of the runway to vacate because the usual intermediate exit was not available and followed this up with a request to keep the speed up prior to vacating. Idle reverse thrust remained selected and as the aircraft vacated, the First Officer selected air conditioning pack 2 to ‘OFF’ and started the APU. It was noted that the aircraft was the first one to use have used the exit taxiway for some time and deposits of de/anti-icing products had been present on the taxiway. Almost immediately “a smoke-like substance started filling the cabin from the overhead vents” It was subsequently described as “appearing along the entire length of the cabin and being either brown or black in colour”. It was impossible to establish with any accuracy what the volume or density of the smoke had been but it was clear that visibility within the cabin had been affected. The associated smell was described as “being reminiscent of a bonfire or electrical burning”.
The two cabin crew in the rear galley reported preparing to don their smoke hoods and while the cabin manager in the forward galley advised the First Officer that there was “smoke in the cabin” using the crew interphone, their colleague at that location ran part of the way down the cabin, shouting to passengers “get your heads down”, before returning to the galley. The Captain had not heard the cabin manager’s report and at about the same time as he was told of it, he was selecting the thrust levers from idle reverse to idle, by which time the aircraft had travelled about 270 metres along the taxiway and its ground speed had reduced to about 12 knots.
The cabin manager then saw that the smoke was becoming thicker and called the flight deck again to say that the smoke was becoming thicker and that they thought an evacuation was necessary. The First Officer partially repeated back the statement and the cabin manager got an ‘okay’ from both pilots. The aircraft was brought abruptly to a stop and the parking brake set. The First Officer made a MAYDAY call to advise the evacuation but had inadvertently left his transmit selector on the crew interphone channel so ATC did not hear it. As the flight crew commenced their evacuation check list as a prelude to ordering an evacuation by PA from the flight deck, the cabin manager issued their own evacuation command over the PA but did not activate the evacuation alarm. Upon hearing the forward doors opening behind the flight deck, the Captain then immediately shut the engines down and as the engine-driven generators went off line, all lighting in the flight deck extinguished making it necessary to complete the evacuation check list from memory. ATC were eventually advised and the Rescue and Fire Fighting Services despatched. The partial passenger load was quickly evacuated using three of the four slides; the slide at 2R was not used because it was reported to have been too dark for the supervising cabin crew to confirm that it had deployed correctly and the handle cover of the rear right over-wing escape hatch had been removed but the exit had not been opened. No attempt had been made to open any of the other over-wing exits. The passengers and crew were eventually taken by bus to the Terminal.
Early attention was focussed on runway and taxiway de-icing. The runway and taxiway used had been treated with de-icing chemicals on several occasions in the 24 hours prior to the incident. Late on the previous evening, ‘Isomex3’, a Potassium Acetate-based thickened liquid, had been applied at a rate of 22gm/m² and then the next day, in both the early morning and early afternoon, Urea in the form of ‘prills’ or small pellets had been applied at a rate of 100 gm/m². The Urea pellets had not been wetted immediately prior to application which has sometimes been recommended. Following the incident, the Airport Duty Manager reported identified partially dissolved prills on the final third of Runway 25.
Findings following inspection of the aircraft for residual evidence of the contaminant were largely ambivalent but clear evidence was found of significantly more potassium acetate on seat headrest covers taken from the incident aircraft than from similar covers taken from other aircraft and was considered consistent with previously high levels of potassium acetate in the cabin environment. Since both potassium acetate and urea based de-icing products had previously been applied to the runway and taxiway and they considered the most likely source of the smoke/fumes. It was thought likely that “de icing chemicals were ingested into the engine, before passing through the air conditioning system and entering the cabin though the overhead vents” and it was noted that “although evidence of potassium acetate was found in the cabin, information from (other similar events) suggested that urea pellets may also have been the source of the smoke”.
It was considered that:
“Two factors probably influenced the production of smoke. First, the taxiway onto which the aircraft turned had been regularly treated during the period preceding the event but very lightly trafficked, so there was probably more de-icing product on this taxiway than on other parts of the manoeuvring area. Secondly, it is likely that the prolonged use of reverse thrust increased the volume of these products delivered to the cabin and the thickness of the smoke (given that) the appearance of the smoke was coincident with the use of reverse thrust on the taxiway”.
A reference to the risk of ingesting runway de-icing fluids was found in the Operations Manual:
“Engine ingestion of freshly treated runway with potassium acetate/urea may occur causing a nontoxic mist in the cabin. This mist can be misidentified as smoke. Therefore, consider briefing the Cabin Crew prior to landing.”
The Operator also reported two other similar reports of runway de icing fluid appearing to temporarily contaminate the air conditioning system in the same month as the investigated occurrence, another event at Belfast Aldergrove and one at Budapest.
Following a query, the Operator advised the Investigation that a check of OFDM data had disclosed that the incident aircraft commander “routinely used idle reverse thrust beyond the landing roll and onto the taxiway down to approximately 15 knots to save brake wear”. It was noted that his use of reverse thrust had apparently not provoked comment during recurrent training or checking and had not been identified through the OFDM programme even though it was specifically contrary to SOP as stated in both the Operations Manual and the Flight Crew Training Manual (FCTM) because of a desire to minimise the intake of Foreign Object Debris.
The issues raised by the confusion in communications between the cabin manager and the flight crew were reviewed. It was noted that the failure of the cabin manager to use the available ‘Emergency’ interphone channel rather than the standard one or to use the evacuation alarm prior to commencing an evacuation were both unhelpful. However, the importance of less ambiguous flight crew responses to reports from the cabin was also noted and observed to be already stressed in manuals used by both cabin crew and flight crew. The subject of when cabin crew should initiate an evacuation on their own rather than defer to the Captain was considered and it was concluded that documentation of pertinent cabin crew SOPs was adequate. The cabin manager’s response to the smoke was driven by a concern that fire or smoke in an aircraft may be very hazardous to the occupants; the cabin safety procedures manual listed ‘uncontrolled cabin fire/smoke’ as an example of a catastrophic situation which may require immediate independent evacuation action by the cabin crew. At the time the cabin manager began the evacuation, it was considered that the fact that the engines were running and they were already in communication with the flight crew meant that “it would have been appropriate for the cabin manager to wait for the commander to initiate the evacuation, according to the laid down procedure”.
On the absence of flight deck lighting after the emergency engine shut down during the incident, it was noted that the subject was covered in the FCTM with the entry “…on ground with engines stopped, only the right dome light is operational and the three positions (BRT, DIM, OFF) of the dome light switch remain available, allowing the emergency evacuation procedure completion.”
Safety Action taken by the Operator to review relevant sections of the Operations Manual, enhance pilot training in the use of reverse thrust, add a new OFDM event to track use of reverse thrust and introduce a new cabin crew briefing format was noted as were Airport Operator actions to improve their response to emergencies..
The Final Report of the Investigation was published on 10 May 2012. No Safety Recommendations were made.