British European Airways Flight 548 was a scheduled air service operated by British European Airways (BEA) from London Heathrow Airport to Brussels Zaventem Airport. It was operated by Hawker Siddeley Trident airliners. At 1611 hours local time on Sunday 18 June 1972 the service, operated by a Trident 1C registered G-ARPI, crashed in the town of Staines less than three minutes after departing from Heathrow. All 118 people aboard — three pilots, three cabin crew members, three members of another air crew and 109 passengers — died. The accident was the biggest air crash in Britain until the Lockerbie disaster in 1988.
The Public Inquiry held to determine the causes of the accident was highly controversial among British pilots and the public. Many felt that extremely poor industrial relations at BEA were the real underlying cause of the accident and that the flight crew, headed by an experienced and respected senior captain, was wrongly assigned the role of scapegoat. Even the minutiae of what happened remain controversial, an elaborate but unofficial version of the events being that poor maintenance of a stall recovery system and unreliable speed sensors misled the flight crew.
The Public Inquiry's recommendations led to the mandatory fitting of cockpit voice recorders on heavier British-registered airliners. Within a year of the accident, BEA ceased to exist and became a division of British Airways.
Meanwhile, BALPA was also in an industrial dispute with BEA, concerning pay and conditions issues. The dispute was highly controversial, with clearly defined protagonists (mainly but not exclusively younger pilots) and antagonists (mostly but not only older pilots). At the time of the accident, BALPA had organised a confidential postal ballot to ask its members at BEA whether they wanted to strike.
At the time of the accident, a group of 22 members of a grade of BEA Trident co-pilots known as Supervisory First Officers (SFOs) were already on strike against BEA, citing their status and workload. To compensate for a temporary shortage of fully qualified copilots, SFOs were instructed to occupy only the third flightdeck seat of the Trident and to act in the capacity known as “P3,” involving operating the aircraft’s systems and assisting the captain (known as “P1” on the BEA Trident fleet) and the co-pilot (known as “P2”) who between them handled the aircraft. In other airlines and aircraft, the job of BEA Trident SFO/P3s was usually performed by flight engineers (usually the least experienced flightdeck crew members). As a result of being limited to the P3 role, BEA Trident SFOs/P3s were denied experience of aircraft handling: something which led to their losing some pay and which they resented. In addition, their status led to a regular anomaly: more experienced SFO/P3s could only assist while less experienced copilots actually flew the aircraft.
Tensions and hazards resulting from the positions in which BEA Trident SFOs and young copilots were placed came to the fore soon before the accident. On Thursday 15 June, a captain complained vociferously that the inexperienced copilot whom he had been assigned "would be useless in an emergency. Upset, the copilot committed a hazardous error on departure from Heathrow. This was noted and remedied by the SFO who later related the event to his colleagues as an example of avoidable danger. At the time of the accident, this event was becoming known among BEA pilots as "the Dublin Incident". A mere hour and a half before the departure of flight BE548, its captain-to-be Stanley J L Key was involved in a quarrel with a First Officer named Flavell while the two were in the BEA crew room at Heathrow’s Queen’s Building. The subject of Key's outburst was the threatened strike which Flavell supported and Key opposed. A witness described Key’s outburst as “the most violent argument he had ever heard.” Both of Key's flightdeck crew members on BE548 witnessed the altercation. Shortly afterward Key apologised to Flavell and the matter appeared closed. Key’s robustly anti-strike views had won him many enemies in the weeks before the accident. Graffiti directed personally against him had appeared on the flightdecks of many BEA Tridents.
While technically advanced, the Trident (and other aircraft with T-tail arrangement) had dangerous stalling characteristics. If its speed was insufficient, and particularly if its high-lift devices were not extended at the low speeds typical of climbing away after take-off or of approaching to land, it could enter a deep stall or “super stall” condition from which recovery was practically impossible.
The danger first come to light in a near-crash during a 1962 test flight when Hawker Siddeley pilots Peter Bugge and Ron Clear were testing the Trident's stalling characteristics by pitching its nose progressively higher, thus reducing its airspeed: "After a critical angle of attack was reached, the Trident began to sink tail-down in a deep stall." Eventually it entered a flat spin and a crash "looked inevitable". Only luck saved the test crew. The incident resulted in the Trident being fitted with an automatic stall warning system known as a "stick shaker" and also an automatic stall recovery system known as a "stick pusher" which automatically pitched the aircraft down to build-up speed if the crew failed to respond to the warning. These systems were the subject of "one of the most comprehensive stall programmes on record," involving some 3500 stalls being performed by Hawker Siddeley before the matter "was squared off to the satisfaction of ... the ARB" (Air Registration Board). Due to their very nature, however, the stall warning and recovery systems tended to over-react: of ten activations between the Trident entering service and June 1972, only half were genuine. This led many BEA Trident pilots to distrust them. Questioned informally, more than half of them said they would disable the systems in case of activation, rather than let them do their job.
The propensity to deep stalling resulted in the crash of Trident 1C G-ARPY on 3 June 1966 near Felthorpe in Norfolk during a test flight, with the loss of all four pilots on board. In this event, the crew had deliberately switched off the stick shaker and stick pusher to perform their tests. The Confidential Human-Factors Incident Reporting Programme, an experimental voluntary non-attributable and informal system of reporting hazardous air events introduced within BEA in the late 1960s and later taken up by the Civil Aviation Authority, brought to light two earlier near-accidents involving the Trident’s deep-stall characteristics.
In one, the captain of a Trident 1C departing Paris Orly Airport for London in December 1968 attempted to improve climb performance by retracting the flaps shortly after take-off. This was a non-standard procedure (the Trident had a sluggish climb performance which earned it the nickname of “Gripper” or “Ground Gripper” among its pilots). A short time later, he also retracted the leading edge devices. This configuration of high-lift devices at a low speed would have resulted in a deep stall. The co-pilot noticed what had happened, increased speed and re-extended the leading edge devices; the flight continued normally. The event became known as “the Paris Incident” or "the Orly Incident" among BEA pilots.
In the second near-accident, a Trident 2E climbing away from London Heathrow for Naples in May 1970 experienced what was claimed by its flightdeck crew to have been a spontaneous uncommanded retraction of the leading edge devices which remained unnoticed by any of them. The aircraft’s automatic systems sensed the loss of speed and lift and issued two stall warnings. Since the crew could not detect anything wrong, they disabled the automatic devices. At that moment, they claimed to have noted and immediately remedied the error. The flight continued normally. Investigators into the event stated that the aircraft had “just about managed to stay flying”; they found no mechanical malfunction which could have caused the premature leading edge device retraction. The layout of the controls came under some suspicion, however. The event became known as “the Naples Incident” or the "Foxtrot Hotel Incident" (after the registration of the aircraft concerned) at BEA.
An accident affecting the particular Trident operating BE548 had occurred on 3 July 1968. An Airspeed Ambassador freight aircraft, G-AMAD, deviated from the runway on landing and struck G-ARPI and its neighbouring sister aircraft G-ARPT while they were parked unoccupied near Terminal 1. The freighter's six crew died. G-ARPT was destroyed, but G-ARPI was repaired at a cost of £750,000 and performed entirely satisfactorily ever since.
Flight BE548 was due to depart at 15:45 hours; the flight crew arrived on board at 15:20. It comprised Capt Key (see above) as P1, Second Officer Jeremy W Keighley as P2 and Second Officer Simon Ticehurst as P3. The Captain was aged 51 and had some 15,000 flying hours' experience, of which 4000 on Tridents. S/O Ticehurst was 24 and had accumulated over 1400 hours, including 750 hours on Tridents. S/O Keighley was 22 and had joined line flying a month and a half earlier, accumulating 29 hours as P2.
At 15:36 flight despatcher J Coleman presented the loadsheet to Key. Payload for the flight was at the maximum permitted value. Fuel load was 8200 kilos/18,080 pounds, giving a takeoff weight of 50,000 kilos/110235 pounds or 2000 kilos/4410 pounds under the maximum permitted takeoff weight. Three minutes later, Key requested engine start clearance and was granted it. As the doors were about to close, Coleman asked Key to accommodate a BEA flight crew which had to collect a Vickers Vanguard "Merchantman" aircraft from Brussels. The additional weight of the three crew members necessitated the removal of some mail and freight from the Trident and the loadsheet to be redrafted. This took place with the engines running. The “dead-heading” crew was led by Capt John Collins, an experienced former Trident First Officer, whom Key invited to occupy the observer seat on the flightdeck.
The doors closed at 15:58 and at 16:00 Key requested pushback. At 16:03 BE548 was cleared to taxi to the holding point adjacent to the start of Runway 28 Right. During taxi, at 16:06 the flight received its departure route: a routing known as the “Dover One Standard Instrument Departure,” involving taking-off to the west over the Instrument Landing System Localiser Middle Marker beacon of the reciprocal Runway 10 Left, turning left to intercept the 1450 radial heading to the Epsom VOR beacon (to be passed at 3000 feet or more) and proceeding to Dover. Key then advised the Tower he was ready for take-off and was cleared to do so. He then reported an unspecified technical problem and remained for two minutes at the holding point to resolve it.
At 16:08 Key again requested and received take-off clearance. Wind came from 2100 at a rather high speed of 17 knots. Conditions were turbulent, with driving rain and a low cloudbase. At 16:08:30, BE548 began its take-off run which lasted 44 seconds, the aircraft leaving the ground at an indicated airspeed (IAS) of 145 knots. The safe climb speed (V2) of 152 knots was reached quickly. The autopilot was engaged at 170 knots and 355 feet, after 19 seconds in the air. Even though the required initial climb speed was 177 knots, the autopilot’s speed lock was engaged at 170 knots.
At 16:09:44hrs (74 seconds after the start of the take-off run), passing 690 feet, Key commenced the turn towards the Epsom VOR and reported that he was climbing as cleared. At 1000 feet the flight entered cloud and encountered stronger turbulence. At 16:10 (90 seconds), Key commenced a standard procedure to cut community noise which involved reducing engine power. As part of this, at 16:10:03 (93 seconds) he retracted the flaps from their take-off setting of 20 degrees. Shortly afterwards, BE548 reported passing 1500 feet above ground level and was re-cleared to climb to 6000 feet above sea level.
In the very turbulent conditions of the turn, speed dropped to 157 knots, 20 knots below the target speed. The reduction in engine power might also have activated the warning light indicating low air pressure in the stall recovery system: a three-way air pressure valve was later found to have been one-sixth of a turn out of position, and the pin which locked it into position was found to be missing.
At 16:10:24 (114 seconds), the leading edge devices were selected to be retracted at a height above the ground of 1770 feet and a speed of 162 knots: 63 knots below the minimum speed specified in the aircraft manual for leading edge device retraction. One second afterwards, visual and aural warnings of a stall activated, followed at 16:10:26hrs (116 seconds) by a stick shake and at 16:10:27hrs (117 seconds) by a stick push which disconnected the autopilot, in turn activating a loud autopilot disconnect claxon (this claxon was not switched-off for the remainder of the flight). Key pulled the nose up and levelled the wings, most likely in accordance with a basic principle of airmanship which states that in an emergency one ought to concentrate on “flying the aircraft.” On this occasion, however, his action had the unfortunate effect of stalling the aircraft by slowing it down.
By 16:10:32 (122 seconds), the leading edge devices had stowed fully into the wing. The speed was 177 knots and height above the ground was 1560 feet, with the nose pitched up to its usual climb angle. A second stick shake and stick push followed in the two seconds following this. Key continued to hold the nose up and the wings level. A third stick push followed 127 seconds into the flight but no recovery was attempted. One second later, the stall warning and recovery system was overridden by a flightdeck crew member.
At 16:10:39 (129 seconds), the aircraft had descended to 1275 ft and accelerated to 193kt as a result of the stall recovery systems pitching it down to gather speed. It was in a 160 bank to the right, still on course to intercept its assigned route. Key pulled the nose up to reduce speed to the required 175 knots. This further stalled the aircraft.
At 16:10:43 (133 seconds), the aircraft entered a deep stall. It was descending through 1200 feet, its nose was pitched up by 31 degrees and its airspeed had dropped below the minimum indication of 54 knots. At 16:10:55 (145 seconds) and 1000 feet, the Trident broke cloud, descending at 4500 feet per minute at an angle of 60 degrees. The impact with the ground came at 16:11 precisely (150 seconds after brake release).
The aircraft came to rest on a narrow strip of land surrounded by tall trees immediately south of the A30 arterial road and a short distance south of the King George VI Reservoir near the town of Staines. It just cleared high tension overhead power lines. There was no fire on impact. (A fire broke out during the rescue effort some time later when cutting apparatus was used.)
There were three eyewitnesses of the impact: two brothers aged 13 and 9 who were walking nearby and a passing motorist who stopped and called at a house to telephone the airport authorities and advise them of the accident. Air traffic controllers had not noticed the disappearance from radar of BE548. The authorities were unaware of the accident for some minutes; they remained incompletely aware of the precise circumstances of it for nearly an hour). First on the scene of the accident was a nurse living locally. An ambulance crew driving near the scene arrived soon afterwards. Altogether, 30 ambulances and 25 fire engines attended the accident. A male passenger who had survived the accident was discovered in the aircraft cabin but died from internal injuries without recovering consciousness on arrival at Ashford Hospital. No other survivors were found. The aircraft's two flight data recorders were removed for immediate examination.
The British air community was wary of Public Inquiries for several reasons. In such Inquiries, AIB inspectors were on an equal footing with all other parties and the ultimate reports were not drafted by them, but by the Commissioner and his or her Assessors. Proceedings were often adversarial, with counsel for victims' families often attempting to secure positions for future litigation. Deadlines were often imposed on investigators (pressure of work caused by the Lane Inquiry was blamed for the death of a senior AIB inspector who committed suicide during the Inquiry).
Investigations at the site of the accident were completed within a week. The wreckage of the Trident was then removed to a hangar at the Royal Aircraft Establishment in Farnborough, Hampshire, for partial reassembly aimed at checking the integrity and inter-operability of its componentry. An inquest was held into the 118 deaths on 27 June 1972, returning 118 findings of accidental death. The pathologist stated that Captain Key had suffered a potentially distressing arterial event caused by the raised blood pressure typical of stress. (This event was popularly interpreted by much of the press as a "heart attack.") It had taken place "not more than two hours before the death and not less than about a minute". In other words, Key could have suffered it at any time between the row in the crewroom and 90 seconds after the start of the take-off run or the instant of commencing noise abatement procedures. The pathologist could not specify the degree of discomfort or incapacitation which Key might have felt. The Captain's medical state contined the be the subject of "conflicting views of medical experts" throughout the Inquiry and beyond it.
The Lane Inquiry opened at the Piccadilly Hotel in London on 20 November 1972 and continued for 37 business days until 25 January 1973 despite expectations that it would end sooner. Geoffrey Wilkinson of the AIB opened it with a description of the accident. In addition, counsel for the relatives of the crew members and passengers presented the results of their private investigations. In particular, Lee Kreindler of the New York City Bar presented claims and arguments which many considered tendencious and inadmissible. They involved hypotheses about the mental state of Capt Key, conjecture about his physical state (Kreindler highlighted disagreements between US and British cardiologists) and allegations about BEA management, and were delivered using tactics "bordering on the unethical. The Inquiry also conducted field inspections, flew real Tridents and "flew" the BEA Trident simulator and the Hawker Siddeley Trident control systems rig. Its members visited the reassembled wreckage of G-ARPI at Farnborough. Throughout their movements, they were followed by the press. The bare facts being more-or-less uncovered soon after the event, the Inquiry was frustrated by the absence on the accident aircraft of a cockpit voice recorder.
The stall warning and stall recovery systems were at the centre of the Inquiry; it examined in some detail their operation and why the flight crew mght have overriden them. A three-way pneumatic valve which formed and important part of the stall recovery system was reported to have been poorly maintained: it was one-sixth of a turn out of position (in effect, exactly midway between two set positions) and was lacking the locking pin which fixed it into a set position. This unserviceability might have given the crew indications of failure in the system, offering Key solid grounds for mistrusting its warnings and interventions and ultimately overriding it. The failure indications might have begun to appear just prior to the take-off run and might have accounted for the two-minute delay at the end of the runway. No other unserviceability was found on the crashed aircraft and thus there were no other probable reasons why the take-off would have been delayed as it was.
The Inquiry's findings as to the causes of the accident, published in its report on 14 April 1973, were that the Captain failed to maintain the recommended speed; that the leading-edge devices were retracted too early; that the crew failed to monitor speed and configuration; that the crew failed to recognise the reasons for the stall warnings and stall recovery system operation; and that the crew wrongly disabled the stall recovery system. The underlying cause of the accident was stated to have been Key's heart condition. Recommendations included an urgent call for cockpit voice recorders and for closer cooperation between the Civil Aviation Authority and British airlines. Though the report covered the industrial scene at BEA, no mention was made of it in its conclusions, despite the feelings of many observers that it intruded directly and comprehensively onto the flightdeck of the stricken aircraft. BEA, on which "surgery without anaesthetic had been conducted at the Inquiry, ceased to exist as a separate entity by 1974, becoming a division of British Airways. Among the outcomes of the report was that cockpit voice recorders became mandatory on larger British registered airliners from 1973.
An issue treated as secondary at the Inquiry was the presence on the flightdeck observer's seat of Capt Collins. The Lane report recommended greater caution in allowing off-duty flight crew members to occupy flightdeck seats and aired speculation that Collins might have been telling jokes to his colleagues. Sources close to the events of the time suggest that Collins might have played an altogether more positive role by attempting to lower the leading-edge devices in the final seconds of the flight.
There were protests at the conduct of the Inquiry by BALPA (which likened it to "a lawyers' picnic") and by the Guild of Air Pilots and Air Navigators which "condemned" the rules of evidence adopted and the adversarial nature of the proceedings. Many observers also pointed to an unduly favourable disposition by the Inquiry to Hawker Siddeley, manufacturer of the Trident, and to the makers of the aircraft's systems. Much debate about the enquiry continued throughout 1973 and beyond.
An alternative reading of the events of the accident, apocryphal within the British air community since 1972, centres on an assumption that there was a discrepancy between the indications of the flightdeck airspeed indicators (ASIs) and the airspeed readings obtained from the flight data recorder (FDR): "In simple words: what the FDR records is not necessarily the same as what the captain sees on his panel. Though the Trident had advanced electronics, it was possible for them to react in a manner which assured Key that his speed was "far from catastrophically low. If the assumption is correct, then all Key's actions, including retracting the leading-edge devices, appear entirely correct. This version discounts any genuine incapacitation on the part of Key, who on three occasions had the presence of mind and physical strength to counter the stall recovery system. Elements of this alternative reading were examined favourably by the Public Inquiry but failed to find their way into any of its conclusions or recommendations.
Two memorials in Staines were dedicated on 18 June 2004.
The first is a stained glass window in St Mary's parish church, Church Street. The second is a slightly more accessible area of reflection with seating on the Moormede estate, close to where the accident occurred. The location of the memorial is in the park/play area found near the end of Waters Drive in the Moormede Estate.