Iberia A320 at Madrid on Jul 3rd 2018, fumes and smoke in cockpit

Last Update: June 14, 2021 / 13:18:12 GMT/Zulu time

Bookmark this article

Photo of Iberia EC-JFN, Airbus A320 — Iberia EC-JFN, Airbus A320 (Photo credit: flybyeigenheer / Flickr / License: CC by-sa)

Incident Facts

Date of incident
Jul 3, 2018

Classification
Incident

Airline
Iberia

Flight number
IB-3214

Departure
Madrid, Spain

Destination
Brussels, Belgium

Aircraft Registration
EC-JFN

Aircraft Type
Airbus A320

ICAO Type Designator
A320

An Iberia Airbus A320-200, registration EC-JFN performing flight IB-3214 from Madrid,SP (Spain) to Brussels (Belgium), was climbing out of Madrid's runway 14R when at about 3500 feet MSL the crew noticed a strong burning odour in the cockpit and subsequently noticed smoke emanating from the air conditioning vents. The crew stopped the climb at FL070, donned their oxygen masks and returned to Madrid for a safe landing on runway 18R about 16 minutes after departure. At the time of touch down the smoke had almost completely dissipated. There were no injuries.

Spain's CIAIAC reported on Jul 26th 2018, that maintenance inspected the aircraft and found an obvious loss of hydraulic fluid from one of the engines (CFM56). The occurrence is being investigated by the CIAIAC.

On Jun 14th 2021 Spain's CIAIAC released their final report concluding the probable causes of the incident were:

The investigation has determined the incident was most likely caused by an inadequate maintenance practice, which led to the contamination of the bleed air from engine 1 and caused smoke to enter the cabins through the air conditioning vents.

The air was contaminated most probably by the residues of grease, oil and bird remains, which were burned during the routine operation of the engine, having not been removed during the maintenance carried out after the bird ingestion during the preceding flight. Oil spills left by prior maintenance procedures could also be a factor.

The CIAIAC analysed that weather had no role in the incident.

The CIAIAC analysed the flight:

Until the cockpit smoke event occurred, the flight was operating normally as per the established procedures. There was no evidence of any operational anomaly. According to their testimonies, up until that point, the crew believed that take-off had been carried out without incident. However, while taxiing for take-off, they noticed "a slight burning smell, which was difficult to pin down" but, as the captain indicated, smelled similar to "burnt glue". Nevertheless, he didn’t think it was important because there were other aircraft nearby, and he assumed their exhaust fumes had entered the air conditioning system.

The climb to 3,500 feet took place according to the usual parameters, and there were no ECAM warnings or abnormal values on the instrumentation panels. It was at this altitude that, on connecting pack 1, smoke began to come out of the CM2 A/C. The burning smell was noticeable for 15 minutes during the taxi for take-off. According to the captain, when the smoke began to filter in, the smell became more pronounced. We can, therefore, assume that it was the same smell detected while taxiing but not regarded as significant until the smoke appeared in the cockpit.
Given that the smoke had thickened and rapidly invaded the cockpit, the captain decided to declare an emergency and return to LEMD as soon as possible.

The captain believed the most suitable distribution of flight-deck functions for the immediate return to LEMD was for him to act as PF, taking control of the aircraft and communications, and for the co-pilot to perform the functions of PM, completing the QRH procedure. He instructed the co-pilot to use the emergency mask and he put his on at the same time.

Given the circumstances, this was the correct decision. According to his testimony, his priority was to return to LEMD as soon as possible while simultaneously dealing with the smoke in the cockpit as quickly as they could.

With the emergency masks on, the captain went through the MEMORY ITEMS, OEB, ECAM, NORMAL CHECKLISTS, RESET COMPUTER and finally, the QRH SMOKE / FUMES / AVNCS SMOKE out loud. He, therefore, followed the procedures correctly.

The captain instructed the co-pilot to locate and read this last list, but by the time he started to read it, the smoke was clearing. They were unable to communicate properly with the masks on, and when the smoke had dissipated entirely, he told the co-pilot to remove his mask whilst he did the same. Given that they were just minutes away from touching down, they abandoned the QRH because the captain needed the co-pilot to reset the FMGC data for the approach.

While this was happening, the CABIN-to-COCKPIT communications channel had been activated, but the captain decided not to answer as he was “busy flying the aircraft”. When the cabin crew attempted to make contact a second time, he responded. The purser informed him of the presence of smoke in the passenger cabin, and he replied that they would be back in Madrid in around 10’.

From that moment on, the operation was controlled and carried out according to procedures. He levelled the plane to 7,000 feet, declared "MAYDAY, MAYDAY, MAYDAY due to SMOKE IN CABIN", and requested vectors. After receiving clearance from ATC, he turned right to return at his discretion to runway 18R and began the manoeuvre.

Once everything was configured, he transferred the communications to the co-pilot. The purser called to report that the smoke had almost disappeared. The captain told him to instruct the cabin crew, inform the passengers and prepare for a PRECAUTIONARY LANDING.

Given the situation, they informed ATC that the circumstances were more PAN PAN than MAYDAY so that they could update the emergency services on the ground.

He went thought the approach checklist, and the rest of the landing sequence was carried out without incident.

The passengers and crew disembarked normally after waiting approximately 20‘ for the stairs and buses required to transfer them to the terminal. The captain used this time to speak to the passengers, reassuring them and informing them that the operator would ensure they could continue their journey as soon as possible.

We have therefore concluded that the flight, including the emergency declaration that involved the aircraft’s return to the departure airport, was carried out in a safe and controlled manner.

From the analysis of the evolution of the emergency, it is clear the captain thought the best way of managing the situation (smoke in the cockpit) was to change roles so that he could take charge of immediately flying the aircraft back to runway 18R, while the co-pilot assumed the role of PM.

Once the decision to return to LEMD had been made, the captain changed the crew roles and reorganised the cockpit workload so that he could concentrate on immediately flying the aircraft back to land on runway 18R.
He then ordered the co-pilot to put on his oxygen mask. According to the captain, from that moment on, he could not hear the co-pilot clearly and was unable to establish proper communications between them. They then carried out the first QRH instruction following an anticipated return to the airfield.

In the brief period after the crew had fitted their oxygen masks, the smoke began to clear. The co-pilot was already going through the QRH, and the captain was trying to establish communications with him but was unable to hear him properly. In the meantime, the smoke had disappeared. The captain removed his mask, telling the co-pilot to do the same. This stage of the emergency appears to have been slightly chaotic, and it seems both the captain and the co-pilot were somewhat slow to react. That said, they were finally able to resolve the situation satisfactorily because even though the smell remained, the smoke had dissipated, and given the compromised communications, they made the correct choice in removing the masks.

We have not been able to determine why the flight crew struggled to communicate with each other whilst wearing the oxygen masks. There is no evidence of failures in the intercom system, and we have been unable to ascertain whether the flight crew performed all the actions necessary to correctly establish and maintain communication whilst wearing oxygen masks.

The operator, IBERIA, obliges all new flight crew members and those already working for them to complete the in-house emergency and safety equipment course. In both cases, the course covers crew usage of oxygen masks in line with part D of the Operations Manual.

The initial emergency declaration was correct. Subsequently, with the situation under control, the captain downgraded the situation to a PAN PAN emergency so as not to allocate excessive airport emergency resources unnecessarily. This decision demonstrates a clear awareness of the level of severity and risks stemming from the situation, which, by that time, was already fully controlled.

He asked the co-pilot to programme the FMGC for the approach, and once configured, he transferred communications to him. This was the appropriate procedure.

Given the high workload of the flight crew, communications with the cabin crew were initially delayed. After the second call, the purser entered the cockpit to indicate that the smoke had almost disappeared. At that moment, the captain told him that he had declared an emergency, that he intended to return to Madrid and that he expected it would take 10'. He instructed him to prepare for a precautionary landing.

The crew stated that they carried out the approach and landing checklists without further incident and according to procedures.

After verifying that there were no ECAM warnings or related actions required, the QRH reading should have been the next immediate priority. Due to communication issues between the flight crew whilst wearing the oxygen masks, it was not performed adequately.

Despite the fact that, as a result of a glitch in the operator's course control system, the captain had not completed his CRM training within the period established by the Operating Manuals, we believe that, given the situation, he managed the resources available to him during the emergency correctly.

The CIAIAC analysed that a post flight inspection revealed a cracked engine driven pump for the yellow hydraulic system, however, the hydraulic leak could not have caused the fumes and smoke as the crack occurred after the flight with the hydraulic pressures remaining normal throughout of the flight until after landing.

The aircraft had undergone maintnance because during the last flight prior to maintance both engines had ingested birds. The subsequent maintenance actions were completed normally and in accordance with the troubleshooting manual.

The CIAIAC further analysed that the smoke appeared at the precise moment when pack #1, fed by the left hand engine's bleed air, was connected and concluded the only engine involved in the fume/smoke event was the left hand engine. The CIAIAC stateD: "In the maintenance intervention prior to the incident, work on engine 1 consisted of a boroscopic inspection to verify the ingestion of birds as reported by the crew. It was indeed confirmed, as was the fact that no damage had been sustained as a result. Once this was established, the AMM does not specify that the engine must be cleaned. We can, therefore, presume that there could have been bird remains inside the engine and that they would have been burnt during the subsequent flight, contributing to the cabin odour/smoke event."

The right hand engine had been replaced following the dual engine bird ingestion, hence both packs had never been tested together. The CIAIAC continued analysis:

As for the volume of smoke released, it would depend on the amount of residue in the engine and/or the ducting. According to reports, the smoke was thick, but dissipated rapidly in about 3 minutes, which is consistent with the hypothesis that multiple agents were burned in engine 1. These agents may have been grease and oil residues from maintenance, including possible spills on components that were not subsequently cleaned, as well as bird debris. All of which are combustible agents that produce dense smoke and smells, primarily of burnt oil.

The fact that no further smoke was generated after about 3 minutes is consistent with the fact that the design of the air conditioning system allows for the complete renewal of the cabin air in that time frame, and that the causative agents would have disappeared once consumed. It also tallies with the fact that no oil leaks were identified in the post-flight inspection, and that oil pressure was maintained during the flight.

Furthermore, this theory is consistent with the fact that because the QRH list was not completed, pack 1 was never disconnected. Therefore, the bleed air continued to feed the A/C, by this point being smoke-free but still contaminated by the products of oil pyrolysis. This would have caused the persistent burning smell described by the crew, and which had permeated the ducting and bleed-air filters of the A/C system and therefore lasted until the aircraft landed.