Description

On 24 June 1982, a Rolls Royce RB211-524C2 powered Boeing 747-200 (G-BDXH) being operated by British Airways on a scheduled international passenger flight from Kuala Lumpur to Perth as BA009 and in the cruise at FL370 in night VMC with no knowledge of a recent new volcanic eruption of Mount Galunggung entered what turned out to be an ash cloud just south of Jakarta. All four engines quickly failed and a 25,000 feet loss of altitude occurred until four successful engine restarts were achieved at FL120. A diversion to Jakarta (Halim) was then completed by visual navigation on three engines after one of the restarted engines had to be shut down when the aircraft briefly entered the ash cloud again. Significant damage was subsequently found to have been caused to all the engines, to parts of the airframe and to the systems required for many of the primary flight instruments to function.   

Editor's Note:

Since no independent investigation in accordance with ICAO Annex 13 was conducted, this summary article relies entirely on reliable and technically competent secondary sources authored by persons with access to internal British Airways and Rolls Royce which have critically re-examined the available data from this event and others to try and achieve a better understanding of the likely density of the ash clouds penetrated in relation to the exposure duration and the subsequently observed engine damage. Some key points from the event could also be corroborated by a direct account of the event from the Captain of the aircraft which was included in a pilot ash awareness training video released by Boeing in 1992 after a similar event over Alaska in 1989. 

Investigation

The only investigation into the event was the one carried out by British Airways, assisted by Rolls Royce and the Indonesian ANSP but the full report of it was never widely circulated even internally and neither were safety actions derived from this work disclosed publically. However, the fact that this event was both the first serious one of its kind and was well documented helped to raise the safety issue involved so that when a similar event occurred over Alaska in 1989 the subject was taken even more seriously and both the direct safety and resultant maintenance consequences of flying through volcanic ash have since been given much more attention. In the absence of a fatal accident, the absence of an Annex 13 Investigation was not uncommon in the early 1980s. The summary which follows is therefore the best account possible using the information available.

What Happened

The three-man minimum flight crew (standard for the ‘classic’ Boeing 747-200) had taken over at Kuala Lumpur and consisted of a 41 year-old Captain, a 32 year old Senior First Officer and a 40 year-old Senior Flight Engineer. As the flight passed over the Jakarta area, the night sky appeared visually clear ahead but there was no moon. When cloud was entered shortly afterwards, it initially appeared no different to ‘normal’ moisture-based non-convective cloud. At this time, the Captain was briefly away from the flight deck but was immediately called back. As he passed through the upper deck cabin towards the flight deck, he became aware of “smoke” appearing from the floor vents which had an “acrid or ionised electrical” smell and on reaching the fight deck was greeted by a magnificent display of St Elmo’s Fire on the windshields.

Prior to his return, the others had made a precautionary selection of continuous ignition for the engines and put the seat belts on. Looking back at the engines, both pilots then saw that the engine intakes were “glowing as if lit from behind the fan” which was creating a stroboscopic effect and giving the illusion that the engines were moving slowly backwards. 

The aircraft ground track showing the location of Galunggung and its ash cloud. [Reproduced from a contemporary unpublished source]

At this point, the St Elmo’s Fire changed to dashes which were described as having the appearance of tracer bullets and it became clear that the “smoke” the Captain had seen coming from the floor vents earlier was now entering the flight deck too. Oxygen masks were donned by the crew and only about two minutes after the Captain’s return to the fight deck, the first engine (no 4) began to surge and failed. No sooner than the QRH fire drill for this failure had been completed and with the altitude now around FL345, engine no 2 also began to surge and failed and this was followed almost immediately by no 3 and no 1 engines also failing in quick succession. It appears that the engines ran normally at cruise power for as little as 2-3 minutes after entering the ash cloud. The engine failure/fire drill was not followed for the other three engines as it had then became clear that restart attempts would be required on all engines to achieve enough successful restarts to avoid a need to attempt a night ditching.

The crew were somewhat surprised to observe that loss of all generators had evidently not occurred as a result of the engine failures as there was no evidence of the electrical load shedding that, in corresponding simulator training, had occurred in such circumstances. Given the workload of attempting successive engine restarts during a glide descent in IMC, a key result of this was that the Captain’s AP was still useable. 

The First Officer made a MAYDAY call to Jakarta ACC but static radio interference resulted in the quality of communication deteriorating as altitude continued to rapidly decrease and the call that “all four” engines had failed was initially mis-interpreted by ATC as “no 4 engine” had failed. A Garuda flight crew on the frequency then assisted in clarifying the message to ATC.   

Whilst overseeing repeated engine restart attempts, the Captain made a left turn back towards Jakarta and set up a stable descent. As the rate of descent increased, multiple attempts to restart the engines were made without success. It was agreed that if the number of successful restarts was not sufficient to safely clear the terrain which would be on track after crossing the south coast of West Java, which required an altitude of at least 10,500 feet, it would be necessary to turn around and attempt to ditch, a decision which it was decided would need to be taken on reaching FL120. At this time, indications on both INSs were showing that they were unusable. 

Passing FL260, the continuing absence of any pressurisation had resulted in the cabin altitude rising and the corresponding warning when it exceeded 10,000 feet followed by automatic deployment of cabin oxygen masks when it reached 14,000 feet. A problem with the First Officer’s oxygen mask then prompted the Captain to briefly increase the rate of descent in order to get into conditions conducive to relatively normal unassisted breathing conditions. Passing FL200, the primary airspeed indications were found to be 50 knots apart. The Cabin Service Director was called to the flight deck for a briefing but the flight crew’s use of oxygen mask prevented any meaningful communication.    

After about 12 minutes of descent, the aircraft emerged from the cloud into VMC at FL130 still over the sea. By FL125, No 4 engine had been successfully restarted and after a further 1 minute and 20 seconds, the No 3 was then restarted followed quickly by the two remaining ones which made it possible to fly level at FL120. It was decided that the diversion to Jakarta could be continued but that greater than the minimum terrain clearance over the coastal mountains would be worthwhile and would also improve VHF communications with ATC. A climb to FL150 was therefore requested, approved and commenced. However, by the time this level was reached after about 3 minutes at climb thrust mainly in the ash cloud, St Elmo’s Fire had reappeared. Even though thrust was being reduced, engine no 2 then began to surge violently and was shut down. A short descent until clear of cloud again was made using the speed brakes, flaps and landing gear extension rather than reducing the cruise thrust setting and visual navigation then took the aircraft to the Halim overhead at FL100. 

A wide right hand descending turn was then made so as to arrive on the ILS LOC for runway 26 and after initial difficulty locating the runway at range, it was identified and the turn right onto final approach was completed. Having been informed that the ILS GS was of service, the First Officer and the Flight Engineer were ready to assist by calling out the radio altitude and DME distance to go so that an appropriate visual descent could be made, However, as soon as the aircraft was on final approach, it became apparent that almost all the external surface of the two main flight deck windshields had been so abraded by ash that nothing could be seen through them. Only by looking through a narrow vertical strip on the left edge of his windshield was it possible to make out the VASI on the left side of the runway. As touchdown neared, it became obvious that the landing lights, although on, were not illuminating anything and as the aircraft crossed the runway threshold, the crew reported that “the whole of the front windows were filled with a diffuse glare of light” as, in the fortunately calm wind conditions, they waited for the main gear to touch the invisible runway. After landing, exiting the runway and taxiing to the assigned parking position was reported to have been challenging because of the poor forward visibility through the abraded windshields.    

Aircraft and Engine Damage

On disembarking the aircraft, it was immediately obvious to the crew that the wing leading edges, the nose and the engine nacelles had, like the forward windshields, been subject to “sandblasting”. A similar effect on the glass of the landing lights was found to be what had prevented any light being projected from them (see the illustration below).

One of the removed landing light covers. [Reproduced from a contemporary unpublished source] 

Ash contamination of the pitot tubes was also evident as was damage to the engine turbine blades where the blade tips had been eroded away and where silicacious refractory material had sintered in contact with hot metal and fused itself to the blades. This would, by changing the shape and size of the blades, have seriously affected the engines’ efficiency. 

However, after the subsequent removal of the engines and strip down for internal inspection, it immediately became clear that the primary reason for engine failure had been a reduction in the throat area of the Nozzle Guide Vanes (NGVs) of the Stage 1 HP Turbine (see the illustration below). A comparison between the remaining ash residues in the three engines used to complete the diversion and engine no 2, which was not subject to attempted restarting after being shutdown following the second ash encounter, showed as suspected, a significant difference. It was deduced that a substantial amount of ash debris which had been taken above its melting point in the engine hot section had, during the initial 25,000 feet of descent then cooled and solidified when the engines ceased operating and that the subsequent restart attempts had led to deposits which had solidified and become brittle to be partially shaken free in this form. Continued operation at a much lower air density had then enabled them to continue operating during cruise and descent, albeit at reduced efficiency.

The NGVs on the engine no 2 HP turbine showing the characteristically greater ash deposits on alternate vanes (the ‘leading’ ones. [Reproduced from a contemporary unpublished source] 

Observations

Perhaps of most significance was the fact that neither the flight crew involved nor British Airways (or, apparently, ATC) were aware of the ash cloud in advance. In fact, it was from an eruption of Mount Galunggung which had only begun approximately two hours before the encounter. However, that this volcano was generally in an active phase was known - it had been intermittently active during the previous three months. As the new eruption continued, the ash cloud began to be visible on satellite weather pictures.   

This event and a similar one to a Pratt & Whitney-engined Boeing 747-200 which also entered an ash cloud from Mount Galunggung again at night but at the slightly lower cruise altitude of FL330 on 13 July 1982 (three engines failed but were eventually restarted and the fourth was reduced to flight idle) were the trigger for the ICAO to begin work to address the ash risk. In particular, a network of Volcanic Ash Advisory Centres was introduced in order to make it more likely that high density ash clouds could be avoided.  

This event started what remains an ongoing debate about what ash loading was encountered in the two episodes and whilst this was evidently enough to put continued high altitude engine operation at thrust levels significantly above flight idle at risk, it has yet to be possible to reliably establish even a range of densities encountered on the indirect evidence available. 

Related Articles

• Volcanic Ash
• Volcanic Ash: Guidance for Flight Crews
• Managing Volcanic Ash Risk to the Safety of Flights
• Volcanic Ash Advisory Centre (VAAC)
• Contained Engine Failure
• Engine Failure in Cruise
• Cabin Air Quality and Contamination
• B744, en-route, Alaska USA, 1989

Boeing Flight Crew Briefing Video about Volcanic Ash Avoidance

Read more

Volcanic Ash Avoidance

Further Reading

• Volcanic Ash and Aviation Safety: Proceedings of the First International Symposium on Volcanic Ash and Aviation Safety, US Geological Survey, 1994