Description

On 10 June 1990, the left-side windscreen of a BAC One-Eleven (G-BJRT) operated by British Airways on a scheduled international passenger flight from Birmingham to Malaga as BA 5390 suddenly departed the aircraft as it climbed through FL170 in day VMC. The captain was almost wholly ejected from the flight deck and became unconscious whilst his legs were being held in position by cabin crew. A MAYDAY was declared and a diversion to the nearest suitable airport, Southampton, was completed. Only then was it possible to recover the captain, who had regained consciousness shortly before landing, back into the flight deck. The ejected windscreen and some of its bolts were subsequently found a few miles southeast of Didcot. 

Investigation

A Field Investigation was carried by the UK Air Accident Investigation Branch (AAIB).The CVR and FDR were removed from the aircraft and downloaded, and relevant information was recovered from both recorders. 

It was noted that the 42-year-old captain, who was acting as PF when the windscreen departed the aircraft, had a total of 11,050 hours flying experience, which included 1,075 hours on type. The 39-year-old first officer had a total of 7,500 hours flying experience, which included 1,100 hours on type. The engineer who carried out the replacement of the windscreen and the corresponding certification was an experienced shift manager in his fifties, employed by the airline.

What Happened 

Once the flight was established in the climb at 300 KIAS with the captain having taken over as PF following the first officer’s takeoff in accordance with normal company policy to facilitate monitored approach procedures later in the sector, both pilots released their shoulder harnesses, and the captain loosened his lap-strap. Thirteen minutes after takeoff, as the aircraft was passing FL173, there was a loud bang. The aircraft was immediately filled with condensation mist, and it was clear to all crew members that an explosive decompression had occurred. The captain was sucked partially out of the aperture where his windscreen had been, and the flight deck door was blown onto the flight deck where it lay across the central pedestal. A male member of the cabin crew who had been working close to the flight deck door “rushed onto the flight deck and grasped the captain round his waist to hold onto him”. The SCCM then removed the door and associated debris from the flight deck and stowed it in the forward toilet whilst the remaining two cabin crew instructed the 81 passengers to fasten their seat belts. The remaining two cabin crew members then took up their emergency positions.

At the point of decompression, FDR data showed that the control column had moved forward and to the right, probably due to the captain’s movement into the windscreen aperture, which caused the aircraft to pitch 6° nose down and bank 25° to the right. The first officer took control, set both thrust levers to flight idle, and allowed the airspeed to increase to 340 KIAS as the aircraft descended at 4,600 fpm to FL110. This descent took just over 2 minutes, and after levelling off at FL 110, the AP, which had been disconnected by displacement of the control column during the decompression, was re-engaged. The speed was initially reduced to 266 KIAS and then further reduced progressively to 163 KIAS as flaps were extended in accordance with AFM limitations. It was not possible to pull the captain back into the flight deck because the slipstream had pinned his upper body to the flight deck roof, and he had lost consciousness (he recalled lying on his back before losing consciousness).

A MAYDAY was declared but communications on what was a rather busy ATC frequency were very difficult because of noise on the flight deck. The controller continued primarily dealing with other traffic and did not proactively seek to assist the flight as the first officer attempted to select a diversion airport. The flight was also not transferred to a discrete frequency, an option which applicable ATC procedures allowed for, and the controller’s delay in establishing the nature of the emergency meant that the ACC Watch Supervisor did not, as procedurally required, advise the aircraft operator of the situation. This delayed the initiation of British Airways Emergency Procedures.

On the flight deck the member of cabin crew holding onto the captain “began to suffer from frostbite, cuts and bruising” and was relieved by two remaining cabin crew. The first officer requested radar vectors to the nearest airport and was eventually turned towards Southampton and transferred to the approach radar frequency. Having verified that there was sufficient runway length available for landing (the LDA was 1,650 metres), the aircraft was positioned onto visual final for runway 02. A successful landing was completed before coming to a stop on the 1,723 metre-long runway 22 minutes after the decompression had occurred. The airport RFFS attended and recovered the captain, who had just regained consciousness, into the flight deck through the windscreen aperture. He was taken to hospital with multiple fractures, bruising, frostbite and shock. The first officer had believed during the diversion that the captain may have been dead and had reported as such to ATC.

The Investigation noted that having been confronted with “an instantaneous and unforeseen emergency”, the combined actions of the first officer and the cabin crew had “successfully averted what could have been a major catastrophe” and as such “the fact that all those on board survived is a tribute to their quick thinking and perseverance in the face of a shocking experience”.

Why It Happened

It was noted that windscreen on this aircraft type was not designed on the 'plug' principal, which allows cabin differential pressure to contribute to holding it in place. Instead, it was fitted from the outside of the aircraft and secured by means of 90 countersunk bolts, also fitted from the outside. It was observed that the reason for using so many bolts was to prevent leakage of pressurised air through the window seal, whereas the force of internal air pressure alone could be satisfactorily resisted using far fewer bolts.

The accident flight was the first since the left-side flight deck windscreen had been routinely replaced. Its sudden loss soon after takeoff as the differential pressure increased was entirely attributable to its incorrect installation. The installation had involved the unintentional use of bolts which were nearly all of a marginally smaller (0.66mm) diameter than those specified, along with a few which were of the correct diameter but 2.5mm shorter than specified.

The engineer who had fitted the windscreen was found to have done so working alone on his first night shift. He was a shift maintenance manager with multiple approvals who had been employed by British Airways for 23 years and was well respected by both pilots and his fellow engineers. The shift maintenance manager required mild corrective lenses to read small print or figures but habitually did not use his spectacles whilst performing the windscreen replacement or indeed generally. A factor in his use of securing bolts which were almost all smaller than those specified was therefore considered likely to have been visual impairment due to not wearing his spectacles at work, even during a night shift working under artificial light.

It was concluded that “the windscreen fitting process had been characterised by a series of poor work practices, poor judgments and perceptual errors” each of which eroded and compromised the safety standards laid down in generally applicable British Airways engineering procedures. It was found that various cues had been available to the shift maintenance manager to draw his attention to the use of wrong bolts, but “all went unnoticed or unheeded." It was noted that the task did not require a duplicate inspection, and one did not take place. But the reported noted that if a duplicate inspection had been made, there would have been a high probability of detecting the error. It was also considered that errors would have been more likely in the probable presence of sleep deprivation and circadian effects associated with the end of a first night shift. All the evidence available indicated that the practices employed by the shift maintenance manager had permitted the errors and indicated an habitual failure on his part to observe promulgated procedures rather than being "one-offs."

It was found that the work of shift maintenance managers was not subject to review by any other manager, and so there was no likelihood than any inadvertent errors made would be detected. This situation was then compounded by a similar failure of the company engineering Quality Assurance process as applied at Birmingham to directly monitor the working practices of such employees. A particular issue was also identified in respect of reporting of occurrences and the appropriate company paperwork to be used, which appeared to have compromised the effectiveness of internal safety reporting amongst line engineers at Birmingham.

Finally, it was accepted that “the windscreen replacement task may have been unique in that it alone could accommodate the errors associated with its fitment, such that they were exposed so dramatically the first time that the windscreen was called upon to resist cabin pressure”.

Other Observations

1. The performance of ATC following the MAYDAY declaration was found to have not been entirely satisfactory. Even after eventually recognising and acknowledging the first officer’s MAYDAY call and establishing two-way communication, ATC transcripts showed that the London ACC sector controller involved “had never fully appreciated the nature of the emergency” and as a result, it had been unduly difficult for the first officer to obtain the diversion he needed. The ACC frequency had been busy, but the controller had made no attempt to get the emergency traffic onto a discrete frequency. After examining the arrangements for both initial and recurrent emergencies training provided to ACC controllers, the Investigation concluded that the training given to the controller involved in the handling of emergency situations had “probably been inadequate." However, once handover to Southampton APP had occurred, effective assistance was provided and diversion to Southampton was facilitated without further communication difficulties.
2. Regulatory supervision by the UK CAA of line engineering standards at British Airways Birmingham was found to have been the responsibility of their Flight Operations Inspectorate. At the time of the accident, it had been "approximately a year" since an inspection visit had occurred, and it had been completed within a half day and so would have been “necessarily superficial and only likely to have picked up gross discrepancies”.

Three systemic Causal Factors which had combined to create an environment conducive to maintenance error were identified:

1. A safety critical task, not identified as a "vital point," was undertaken by one individual who also carried total responsibility for the quality achieved, and the installation was not tested until the aircraft was airborne on a passenger-carrying flight. 
2. The shift maintenance manager's potential to achieve quality in the windscreen fitting process was eroded by his inadequate care, poor trade practices, failure to adhere to company standards, and use of unsuitable equipment, which were judged symptomatic of a longer-term failure by him to observe the promulgated procedures. 
3. The British Airways local management, Product Samples and Quality Audits had not detected the existence of inadequate standards employed by the shift maintenance manager because they did not monitor directly the working practices of shift maintenance managers.

A total of eight Safety Recommendations were made as a result of this Investigation as follows:

• that the UK Civil Aviation Authority should examine the applicability of self-certification to aircraft engineering safety critical tasks following which the components or systems are cleared for service without functional checks. Such a review should include the interpretation of "single mal-assembly" within the context of "vital points" and the requirements which include a waiver making the definition of "vital points" non­mandatory for aircraft with a Maximum Take-Off Weight Authorised of over 5,700 kg, which were manufactured in accordance with a Type Certificate issued prior to 1 January 1986. 
• that British Airways should review their Quality Assurance system and the relative roles of Ground Occurrence Report Forms and Quality Management Deficiency Reports be clarified and they should continue to educate and encourage their engineers to provide feedback from the shop floor. 
• that British Airways should review the need to introduce job descriptions / terms of reference for engineering grades including shift maintenance manager and above. 
• that British Airways should review the Product Sample procedure with a view to achieving an independent assessment of standards and conduct an in-depth audit into the work practices at Birmingham. 
• that the UK Civil Aviation Authority should review the purpose and scope of Flight Operations Supervisory Visits. 
• that the UK Civil Aviation Authority should consider the need for the periodic training and testing of Engineers. 
• that the UK Civil Aviation Authority should recognise the need for the use of corrective glasses, if prescribed, in association with the undertaking of aircraft engineering tasks. 
• that the UK Civil Aviation Authority should ensure that, prior to the issue of an ATC rating, a candidate shall undergo an approved course which includes training in both the theoretical and practical handling of emergency situations. This training should then be enhanced at the validation stage and later by regular continuation and refresher exercises.  

The Final Report was completed in January 1992 and published on 14 April 1992.

Related Articles

• Explosive Depressurisation
• Maintenance Error
• Maintenance Error Management System
• Maintenance Error Decision Aid (MEDA)
• Internal Occurrence Reporting
• Quality Assurance
• Safety Oversight
• Technical Competencies
• Guidelines for Dealing with Unusual/Emergency Situations in ATC
• Emergency Communications
• Pressurisation Problems: Guidance for Controllers
• Human Factors