Accident occured on 5 November 2000 at Paris Charles de Gaulle (95) to the Boeing 747-200 registered TJ-CAB operated by Cameroon Airlines


This report presents the conclusions reached by the BEA on the circumstances and causes of this accident.

In accordance with Annex 13 of the Convention on International Civil Aviation, with Directive 94/56/EC and with Law N° 99-243 of 30 March 1999, the investigation is intended neither to apportion blame, nor to assess individual or collective liability. Its sole objective is to draw lessons from the occurrence which may help to prevent future accidents.

Consequently, the use of this report for any purpose other than for the prevention of future accidents could lead to erroneous interpretations.


This report has been translated and published by the BEA to make its reading easier for English-speaking people. As accurate as the translation may be, please refer to the original text in French.



Auxiliary Power Unit


Automatic Terminal Information Service


French bureau for investigation and analysis for safety in civil aviation (Bureau d'Enquêtes et d'Analyses pour la Sécurité de l'Aviation civile)


Calibrated Air Speed


Principal flight crew medical test centre


Cockpit Resource Management


Cockpit Voice Recorder


Digital Flight Data Recorder


French directorate fro civil aviation (Direction Générale de l'Aviation Civile)


Distance Measuring Equipment


Exhaust Gas Temperature


Final Approach Index


Federal Aviation Regulations


Flight Control Unit


Flight Engineer


Flight Mode Annunciator




Instrument Guidance System


Instrument Landing System


Joint Airworthiness Requirements




Tower frequency control position


Meteorological Aviation Report


Precision Approach Position Indicator


Pilot Flying / Pilot Not Flying


Primary Flight Display


Altimeter setting to obtain aerodrome elevation when on the ground


Universal Time Co-ordinated


VHF Omnidirectional Radio Range


Date and Time


5 November 2000 at 20 h 57[1]

Boeing 747-200

registered TJ-CAB


Site of Accident


Paris Charles de Gaulle Airport (95)

Cameroon Airlines


Type of Flight


Public transport of passengers

Douala - Paris Charles de Gaulle

Flight UYC070

Cameroon Airlines


Persons on Board


187 passengers (*)

3 Flight crew

13 Cabin crew


Towards the end of the landing roll on runway 09L, the aircraft suddenly deviated to the right of the runway centre line and left the runway between taxiways Z6 and Z7.

When the aircraft passed over a rainwater collector tank, the nose gear was torn off and damaged the fuselage at the level of the electronics bay. The aircraft came to a stop on the concrete structure over the tank.

The thrust reversers on both left engines were found to be in the retracted position, those of the right engines in the extended position.




Third Parties










Very seriously damaged







 (*) A difference of four passengers was noted relative to the data initially provided by the airline, which had stated that there were 183 passengers.


1.1 History of Flight

On Sunday 5 November 2000, the Boeing 747-200 registered TJ-CAB operated by Cameroon Airlines was carrying out the scheduled service between Douala and Paris Charles de Gaulle under the operational call sign UYC070. Take-off from Douala took place at 14 h 25 about five hours late. The stopover at Yaoundé was cancelled due to the operational delays accumulated the previous day. The flight crew was made up of the Captain, one Co-pilot and one Flight Engineer.

Cruise was performed at flight level 350. The beginning of the approach was unremarkable. Two female passengers invited by the Captain were present in the cockpit during the last part of the flight and occupied the two jump-seats.

After establishing contact with Paris Charles de Gaulle approach control, the crew prepared for an ILS approach to runway 09L. The final approach was performed at night at an average speed of 160 knots. The values on the landing card were: IRF 140, wind 140 degrees 20 knots.

Since the leading edge slats 3 did not extend pneumatically, the crew extended them electrically. The landing gear was extended, the flaps set at 30° and the speed maintained at 158 knots.

N.B.: after the event, the Captain indicated that the auto-throttle and the auto-pilot were disconnected on final approach.

Touch-down occurred at 20 h 57 at a speed of 151 knots.

The Captain, pilot flying, moved the thrust reverser levers. During braking, the spoilers did not deploy automatically and the auto-brake system disarmed.

N.B.: the Flight Engineer stated that he told the Captain that the four reversers were deployed. He also stated he had said "speed brake" and had pulled the lever back in order deploy the spoilers manually (see. illustration in §1.12.3).

After a landing roll of about 1700 metres, the aircraft veered off laterally to the right. It went onto the central area situated between taxiways Z6 and Z7. While passing over a rainwater collector tank, the nose landing gear was torn off and damaged the airframe at the level of the electronics bay. The aircraft came to a stop on the concrete structure of the collector tank.

When at a stop, the normal cabin lighting went off. The emergency lighting automatically took over. In addition the public address system and radio communications with the tower were no longer operational. The evacuation was carried out calmly.

1.2 Injuries to Persons

There were no fatal or other injuries in this accident; the three cabin crew who were slightly injured during the runway excursion were not counted, in accordance with international definitions.

1.3 Damage to Aircraft

The structure of the airframe was severely damaged at the level of the electronics bay; it was impossible for the aircraft to be returned to service.

1.4 Other Damage

The 747 was embedded in the concrete part of a rainwater collector tank, of which a part of the vault collapsed under the weight. The damage caused to the airport infrastructures was as follows:

three blue runway side lights were broken and one PAPI box damaged,

grass verges damaged all along the path of the aircraft after its runway excursion between taxiways Z6 and Z7,

complete destruction of an inspection point on underground tank n°3 for rainwater collection. 

The operations to lift and remove the aircraft took place continuously until Friday 10 November at 6 h 00. Given the need to work in the operations areas, most of the work took place at night, runway 09L/27R being able to be closed only between 22 h 00 and 6 h 00. A temporary road was constructed in order to push the aircraft back; it was impossible to pull it forwards towards Lima taxiway since it would have crossed the underground tank, whose vault cannot bear more than 40 tons (the aircraft's weight on landing was about 247.8 tons).

The accident had consequences on the airport's capacity. The presence of the immobilised 747 near runway 09L/27R prevented operations using category II and III ILS approach procedures.

1.5 Personnel Information

1.5.1 Captain

Flying hours

All aircraft types

Of which on B747




In the previous month



CRM training course in Douala from 5 to 6 June 1996. The operator also provided the date of 19 April 2000.

In the three months preceding the accident, the Captain had performed two flights to Paris Charles de Gaulle: on Sunday 15 October 2000 and on Sunday 13 September 2000.

1.5.2 Co-pilot

Flying hours

All aircraft types

Of which on B747




In the previous month



The Co-pilot did not remember having followed a CRM training course. The operator, however, provided the date of 17 February 2000.

1.5.3 Flight Engineer

Flying hours

All aircraft types

Of which on B747




In the previous month



The Flight Engineer did not remember having followed a CRM training course. The operator, however, provided the date of 27 January 2000.

1.6 Aircraft Information

1.6.1 Airframe

1.6.2 Engines

The table below shows the basic engine data:


Pratt & Whitney (P&W)








Serial number





Flying hours





Number of cycles





Installation date





1.6.3 Maintenance

There were no acceptable deferred defects declared on the day of the accident and searches over the previous three months relating to the following systems brought to light no anomalies:

An overhaul (type D) was performed from 4 to 30 June 2000 at 47,940 flying hours, when maintenance of the aircraft was transferred from South African Airways to Air France.

The last type A overhaul was on 20 September 2000.

1.6.4 Weight and Balance

For flight UYC070 on 5 November 2000, the estimated landing weight was 247.8 tons. Taking into account both the distribution of the passengers and the centre of gravity (26 %), the aircraft was within the weight and balance limits defined by the manufacturer.

1.6.5 Landing Performance

Calculations performed after the accident indicated, with or without thrust reversers and AUTOBRAKE on MIN, a roll distance 2,225 metres.

These calculations were made based on the following basic conditions:

Note: the roll distance calculated with the auto-brake system active with a MIN selection does not depend on the use of the thrust reversers. It is when MAX is selected that the application of reverse thrust has an influence on the roll distance.

1.7 Meteorological Conditions

In line with activity at altitude, the warm sector of an Atlantic disturbance reached the coast of Brittany at 12 h bringing moderate winds from the south to south-east sector. Its progress across the regions of north-west France was accompanied by light to moderate precipitation.

At the time of the landing, the meteorological conditions were as follows:

The runway was wet.

METARs between 20 h 00 and 21 h 00

LFPG 052000Z 15019KT 9999 FEW036 BKN040 BKN100 10/05 Q0989 RERA

BECMG 15025G50KT=

LFPG 052030Z 14019KT 9999 FEW036 BKN039 10/05 Q0988 RERA BECMG


LFPG 052100Z 15018KT 9999 BKN038 11/05 Q0987 BECMG 15025G50KT=

1.8 Aids to Navigation

The aids to navigation were working normally. No anomalies were notified by users, either before or after the accident.

1.9 Telecommunications

The crew contacted Paris Charles de Gaulle on the tower LOC-North frequency (119.250) at 20 h 55 min 03. The communications are transcribed below.

POSITION: Loc. North        FREQUENCY:119,250 MHz

Date: 5 November 2000    from 20 h 55 min 03 s to 21 h 08 min 50 s







20 h 55 min 03 s

De Gaulle Tour good day CAM 0 70




CAM 0 70 good day cleared for landing 0 9 left 150 20 maximum 26 knots




Ok cleared 0 9 left CAM Air 0 70.



20 h 56 min 11 s

De Gaulle CAM 0 70 the latest wind?




150° 17 knots gusting to 26.




17 (?) to 26, CAM Air 0 70.



20 h 58 min 32 s

CAM Air 0 70?



20 h 58 min 48 s

CAM Air 0 70?



20 h 59 min 15 s

CAM Air 0 70?



20 h 59 min 54 s

0 70?



21 h 00 min 27 s

0 70?



21 h 01 min 30 s

0 70?



21 h 02 min 59 s

0 70?



21 h 06 min 29 s

0 70?



21 h 08 min 50 s

0 70?

1.10 Aerodrome Information

Paris Charles de Gaulle is a controlled aerodrome open to public air transport. It is located twenty-five kilometres north-east of Paris, at an average altitude of one hundred and twenty metres. At the time of the accident, it had parallel runways 08/26 to the south oriented 088°/268° with a length of 4,215 metres for 26R/08L and 2,700 metres for 26L/08R, and two runways to the north oriented 090°/270° with a length of 4,200 metres for 09R/27L and 2,700 metres for 09L/27R. The airport is equipped with category 9 and level 9 rescue and fire fighting facilities available twenty-four hours a day.

The use of runway 09L/27R was the subject of a NOTAM valid from 7 September 2000 to 5 December 2000 due to work on construction of taxiways.

The principal operating instructions in place were as follows:

1.11 Flight Recorders

In accordance with the applicable regulations, the aircraft was equipped with two flight recorders:

These flight recorders were read out by the BEA.

CVR: The CVR was functioning correctly at the time of the accident. It was equipped with a 30-minute magnetic loop recording tape. The data relating to the accident was erased as the equipment was left on during the evacuation of the passengers.

Note: The flight crew, who stayed in the cockpit for at least thirty minutes, did not apply the instruction on the last line of the evacuation check-list (appendix 5).

DFDR: The graphs in appendix 4 show the speed of the aircraft during the roll, its trajectory, the evolution of the engine parameters as well as the phases of movement and locking of the thrust reversers.

Note: the movement of the thrust levers is not recorded on the Boeing 747; it can be deduced from the behaviour of the engines.

The DFDR records eighty-three parameters on sixty-four words of twelve bits. The main parameters are summarized in the following table. The phases of the landing procedure are represented in one column and the notable parameters are on the lines. Graphs are given in appendix 4.

Phase of flight Passage Touch-down

Three seconds after


Runway excursion
Parameters flaps 30
Time 20.54.48 20.56.52 20.56.55 Loss of synchronisation from 20.57.15


158 kt 151 kt 146 kt 103 kt


105 ° 87 ° 86 ° 101°
EGT engine   1 420 437 475 598
2 451 412 446 413
3 463 418 455 418
4 441 417 431 426

N 1 engine  
1 50 43 54 88
2 55 45 48 49
3 56 48 51 46
4 53 42 46 45

Position of

thrust reversers

Retracted Retracted  1 Retracted  1 Retracted
 2 Extended  2 Transit
3 Extended  3 Extended
 4 Extended  4 Extended

Aircraft trajectory on landing roll:

The points correspond to the aircraft's trajectory sampled every second. The spacing between the points indicates a lack of deceleration.

Aircraft Braking:

If at least one of the thrust levers is forward of the positive idle position after landing, the auto-brake disarms and the brakes are no longer applied automatically (see also 1.16.2). The longitudinal acceleration curve in appendix 4.1 shows variations which characterise direct action by the pilot on the brakes after touch-down. The curves for the previous flight are also shown in appendix 4.1. The two flights were performed with one active auto-brake selected on MIN. The longitudinal acceleration curve for the previous flight shows few strong amplitude variations, unlike that which was recorded on the accident flight, before the runway excursion.

Note: the sampling set at four seconds of data relating to the thrust reversers recorded by the FDR make it impossible to establish a more precise chronology in this dynamic situation.

1.12 Wreckage and Impact Information

1.12.1 Examination of marks

The calculations made based on the parameters show that the aircraft rolled about 1,700 metres along the runway (appendix 4.2).

The first marks identified were at the level of taxiway Z6, that's to say about 2,100 metres from the threshold of runway 09L. They were on the right side of the runway.

The marks from the main landing gear are continuous and showed traces of rubber, indicating heavy braking before the runway excursion.

The aircraft left the runway just after taxiway Z6 leaving deep marks on the clearance strips (see following photo). The distance travelled between the beginning of its deviation and its stopping position was about two hundred and ten metres (see appendix 6).

1.12.2 Examination of the Aircraft

The aircraft came to a stop at a magnetic heading of 170°. The tail was about one hundred and ten metres from the runway centreline. The nose was resting on the ground on the edge of the taxiway parallel to the runway.

On impact with the concrete structure of the rainwater collector tank, the nose gear, bent backwards, penetrated the airframe at the level of the electronics bay, causing extensive damage to the equipment contained therein.

The thrust reversers on engines 3 and 4 (right side) were deployed. Those on engines 1 and 2 (left side) were in a retracted position. The spoilers were retracted. The flaps were extended at 30°on both two sides. The slats were also extended on both sides.

None of the four engines showed any visible signs of damage.

Three emergency slides were deployed on each side of the aircraft.

1.12.3 Examination of the Cockpit

The investigators examined the cockpit the day after the accident.

The four thrust levers were found in the idle position. It was noted that the crew had fired the fire extinguishers and the APU. The electric switch for extension of the number 3 slats was found in the "Extend" position.

Note: the position of the buttons and levers in the cockpit should be taken with care, since the implementation of the "Passenger evacuation" check list was carried out at night.

1.13 Medical and Pathological Information

Not applicable.

1.14 Fire

No fire broke out following the accident.

1.15 Survival Aspects

At 21 h 00 the red alert was given by the control tower. At 21 h 05, a vehicle from the ATC office arrived at the intersection of runway 09L and taxiway Z7. The emergency services were already on the spot.

After the runway excursion, the Captain ordered the evacuation of the passengers. Doors 1, 2, 3 on the right and left sides were used for the evacuation. The aft right and left n° 4 doors were not used due to the position of the aircraft, leaning forwards, which implied a slope which was too steep for the use of the emergency slides at the rear.

After the evacuation of the passengers, one member of the cabin crew went back up an emergency slide to inform the Captain that the evacuation had been carried out successfully and that there were no injured. The chief flight attendant remained with the passengers.

During this phase, the flight crew remained in the cockpit. The passengers and the cabin crew waited, spread out around the aircraft, and were then put onto buses and driven to terminal 1.

1.16 Tests and Research

1.16.1 Examination of equipment

The aircraft and its systems were examined. Details of these examinations are given in appendix 1. No malfunctions were noted on the thrust levers, the thrust reversers or the control systems.

1.16.2 Behaviour of certain automatic systems on landing

In order to study the accident, behaviour on landing of the following systems should be noted:

If the auto-throttle is active during the landing, it disconnects automatically two seconds after touch-down of the main landing gear.

The auto-brake is armed by positioning the "auto-brake" switch on one of the three deceleration rates available (min, medium, max).

The brakes are automatically applied when the ground mode (information supplied by the landing gear) and a wheel speed are detected and the engine thrust levers are in the idle position.

If at least one of the thrust levers is forward of the positive idle position after landing, the auto-brake disarms and the brakes are not or are no longer applied automatically.

Any action by the crew on the brake pedals de-activates the system.

The spoilers extend automatically on landing if the following conditions are met:

The spoilers retract if thrust levers 1 or 3 are moved forwards (in relation to the positive idle position).

The thrust reverser mechanism is commanded by one lever located on each thrust lever. In the ground position and when the thrust levers are at idle, it is possible to pull the thrust reverser levers back to the "interlock" position which corresponds to reverse. This unlocks the reversers and initiates their deployment. There is blocking in the interlock position when the reversers are in motion or incompletely extended. The use of full reverse thrust is possible when the "in transit" lights are off and the "unlock" lights are on (see appendix 2) to indicate that the reversers are deployed and locked.

1.16.3 Similar Events

The two runway excursions described below have many common features with the accident to TJ-CAB (task sharing within the flight crew, positive thrust applied to engine 1, etc.). Accident on 13 September 1993 at Tahiti-Faaa to the B 747-400 registered F-GITA operated by Air France

The crew performed, with instrument confirmation, a visual VOR-DME approach with flight director and auto-throttle active in VNAV mode.

Track following was ensured manually by the co-pilot (pilot flying), while the auto-throttle managed the speed.

In accordance with the logic of the active mode of the automatic flight system, the latter initiated an automatic go-around on arrival at the "End of Descent" point and the signal to the FMA (upper part of the Primary Flight Display).

The pilot not flying announced the change in mode status on the FMA, with no comment or analysis.

The aircraft passed above the glide path and the speed increased (it reached Vref + 35 kt at a height of one hundred and fifty feet).

The pilot flying pulled back and held the thrust levers; he stated that he felt that the levers "were pulling forwards' and tried unsuccessfully to disconnect the auto-throttle.

About two seconds before touch-down, the n°1 thrust lever slipped out of his control, the auto-throttle still being active in "Go Around" mode. The thrust on engine 1 went to the full positive position, until the aircraft came to a stop, without the crew noticing this.

Consequently, on landing, the spoilers did not deploy, the auto-brake was disarmed and thrust asymmetry was high.

The aircraft veered off the runway and ended up in the lagoon, with no casualties (see the report on the Internet site at Accident on 4 November 1993 at Hong Kong to the B 747-400 registered B-165 operated by China Airlines

The report issued by the Hong Kong authorities indicated that the aircraft over-ran the runway following a procedure specific to the Captain for landings in strong cross-winds. The pilot concentrated on roll control while maintaining residual thrust on the engines, in particular on engine 1 whose EGT curve showed values higher than the other engines. The inappropriate positioning of the n°1 thrust lever de-activated the automatic braking system. The aircraft over-ran the runway and ended up in the sea.

1.17 Information on Organisations and Management

1.17.1 Operating Manual and Cockpit Resource Management

Cameroon Airlines uses the Air France Operating Manual.

The normal deceleration procedures on landing are in appendix 3. In the section on task sharing, it is specified that during the deceleration phase the Flight Engineer shall maintain the thrust levers on idle via the handles at the base of the levers until reverse thrust is established.

1.17.2 Feedback System

Despite various requests from the investigators, Cameroon Airlines did not provide information on the establishment of a feedback system. Such a preventive system enables crews to benefit from the experience of other crews who have reported unusual situations.

1.18 Supplementary Information

1.18.1 Summary of Flight Crew's Testimony

The members of the flight crew stated that flight UYC070 was initially scheduled to take off from Douala on 5 November at 8 h 00. Due to the late arrival of the previous flight, the departure was delayed until 14 h 25. A fuel leak noticed before the departure from Douala also caused a further delay.

No problems were experienced on the flight until the approach. The n°3 slats must have extended electrically, which put the aircraft in the normal configuration for landing.

The Captain was pilot flying. At one thousand feet the aircraft was, according to him, stabilised for final approach. The latest information on wind transmitted by the ATC was as follows: wind from 150° at seventeen knots with gusts to twenty-six knots.

The Captain stated that:

- the aircraft had a strong correction to the right on final due to wind,

- the flaps were extended to 30°,

- the auto-brake had been selected on minimum,

- the approach was stabilised on manual,

- the auto-throttle and the auto-pilot were on OFF,

- the FAI of 160 knots was displayed on the auto-thrust speed window on display P10.

He had planned to leave the runway via the exit taxiway for Z7 which is located at the end of runway 09L.

After the touch-down and the announcement of "four blues", (lights corresponding to the indication "reverser in transit") which the Flight Engineer said he had performed, the Captain said that he applied reverse thrust on all four engines. As far as he was concerned, the spoilers were extended. As to the Flight Engineer, he stated that he had seen the blue lights go out and the orange lights corresponding to the "reversers locked" indication come on.

While the speed of the aircraft was around one hundred knots, the Captain noticed that the aircraft was veering off of its trajectory and announced "we're going off" to the crew. He stated that he had retracted the thrust reversers at that moment and then moved the spoilers to the retracted position.

The Flight Engineer noticed that, after the touch-down, the speed was high. He stated that the "SPEED-BRAKE" handle disconnected and that he intervened to extend the spoilers manually, informing the Captain of his action.

After the runway excursion, as soon as the aircraft came to a stop on the central area, the crew, who were no longer in contact radio with the tower, launched an emergency evacuation. The interphone and the lighting were no longer working, with the exception of the emergency lighting. The Flight Engineer shut down the engines and fired the fire extinguishers on all four engines as well as that on the APU. The crew applied the emergency evacuation check-list (see appendix 5) and stated that they pulled the CVR circuit-breaker ten minutes after the aircraft came to a stop (action by the Flight Engineer).

Note: In the course of the second interview with the technical investigators, the flight crew mentioned the presence of two female passengers in the cockpit. They were sitting on the jump-seats to watch the landing. One of these passengers was contacted by the investigators but did not wish to give any information relating to the end of the flight.

1.18.2 Chief Flight Attendant's Testimony Boarding

Since the investigators noted a difference between the number of people listed on the various on-board documents and that provided by the operator after the accident, the chief flight attendant was questioned on this point.

He indicated that the passenger count on boarding at Douala was the responsibility of the airport staff. He did not remember being informed of the last update at the airport relating to the number of people boarded. Furthermore, he confirmed that no head-count of the passengers on board had been undertaken before take-off. The Evacuation

The chief flight attendant stated that the evacuation was carried out normally. With the exception of the flight crew, all of those on board left the aircraft via the emergency slides after it came to a stop.

1.18.3 ATC Office Agent's Testimony

The red alert was given almost immediately after the accident by the control tower duty chief. The ATC office agent went to the accident site knowing only that a Boeing 747 had had an accident near taxiway Z7 and that there was no further radio contact. The emergency services had already arrived. He took part in mustering the passengers while waiting for the buses. He then went on board the aircraft via an emergency slide to go to the cockpit. The flight crew were still present.

In accordance with the procedures in case of an accident, the agent stated that he asked the Captain to stop the CVR so as to preserve the recorded data. The Captain responded that "as there was no electricity on board, the flight recorders were thus stopped". The agent restated his question, saying that the CVR had a duration of thirty minutes. The co-pilot asked if this meant pulling the circuit breaker and the Flight Engineer said that the CVR circuit breaker had been pulled.


2.1 Accident Scenario

The sequence which led to the runway excursion can be broken down into three significant phases. Firstly, the engine 1 thrust lever was not in the idle position at the beginning of the deceleration (just after the touch-down) which inhibited the automatic braking systems. Then, this lever was inadvertently moved forward, which generated thrust asymmetry leading to the lateral runway excursion. Finally, the n°2 thrust reverser was retracted before the aircraft came to a stop.

2.1.1 Residual Thrust during Flare

The recorded parameters, such as the Fuel Flow and N1, show that engine 1, that's to say the outer left engine, was providing greater thrust than the other engines between the touch-down and the runway excursion (appendix 4.3), which indicates that lever n°1 was not in the idle position at the beginning of the deceleration. This hypothesis is reinforced by the absence of any auto-brake.

Equally, the spoilers were not extended automatically after touch-down. The Flight Engineer stated that he had announced "speed brake" and had intervened to extend them manually. Paragraph 1.16.2 showed that thrust levers n°1 and 3 must be in the idle position for the spoilers to extend automatically. The fact that they did not extend confirms that thrust lever n°1 was not in the idle position at the beginning of the deceleration.

It is difficult to explain this incorrect positioning of the thrust lever. The meteorological conditions at the time must certainly have taken up part of the crew's resources during the flare. The aircraft had a tendency to drift with the wind coming from the right, and the pilot had applied firm corrective control. At the time of the flare, he must thus have pushed down on the left rudder pedal to get the aircraft back onto the centre line and moved the control column to the right to continue to counter the effects of  the wind. This manoeuvre, which implies simultaneous action of the rudder pedals, the control column and the thrust levers, requires co-ordination. The incorrect positioning of thrust lever n°1 may have been the result of an involuntary and undetected action on this lever during the decrabbing manoeuvre.

Fatigue was perhaps a contributory factor since the flight was arriving at night and with a considerable delay. In addition, the presence of two people who were not members of the flight crew may have disturbed the crew.

2.1.2 Positive Thrust on Engine Number One and Runway Excursion

The Captain felt the aircraft veering off of its trajectory when the speed was around one hundred knots. The positive increase in thrust on engine 1 noted on the FDR parameters implies that the thrust lever was moved even further forwards. The uncontrolled thrust on engine 1 and the effects of the side wind from the south added to this to pull the aircraft right of the runway centre line.

The raisons for the increased thrust on engine 1 remain difficult to explain. It is not impossible that while reaching over to extend the spoilers, which are on the centre console next to the thrust controls (see photo in 1.12.3), the Flight Engineer may inadvertently have moved the n°1 thrust lever towards full positive thrust. A lack of coordination between the Captain and Flight Engineer, as well as an absence of joint control with the co-pilot, may have contributed to the failure to correct the situation.

2.1.3 Retraction of n°2 Thrust Reverser

After the accident, the investigators noted that the reversers on the two left engines were retracted. At the beginning of the deceleration, three reversers out of four were extended. The retraction of the engine 2 thrust reverser was instigated during the runway excursion (appendix 4.2).

This action, of which the crew was not conscious, can only be explained by the confusion which must have reigned at the time of the runway excursion.

2.2 Cockpit Resource Management

Cameroon Airlines was not able to supply the information requested during the investigation on the nature of CRM training provided for its flight crews. It is likely that only the Captain had undertaken CRM training. The absence of CVR data makes it difficult to study precisely tha task-sharing within the crew at the time of the landing. The preceding scenario nevertheless shows a lack of co-ordination, an absence of joint control and non-application of landing procedures.

2.3 Presence of Third Parties in the Cockpit

The Operation Manual, in the General/Operations section, sets out the conditions for access to the cockpit subject to authorisation from the Captain. In particular, it specifies that "in the interests of safety, the Captain ensures that admission to the cockpit does not create any distractions, nor interferes with the correct execution of the flight".

Persons with no aeronautical function may, despite their best intentions, interfere with the correct execution of the flight, especially during specific phases of flight such as take-off and landing. This has already been noticed on various occasions during investigations in France and abroad, as is illustrated by the following recommendation from the report on the mid-air collision which occurred on 30 July 1998 in Quiberon Bay (56) between the Beech 1900D registered F-GSJM operated by Proteus Airlines and the Cessna 177 registered F-GAJE.

"Paragraph 100 of the OPS 1, sections a) and b), restrictively defines the conditions for access to the cockpit. However, section c) recalls that the final decision rests with the Captain, which may be interpreted as giving the latter the possibility of over-riding the provisions of the sections a) and b), even without any reasons of safety. In addition, especially in small cockpits, interference between the passengers and the crew may occur without there being any real access to the cockpit.

Consequently, the BEA recommends:

That the DGAC ensure that paragraph OPS 1-100 are correctly understood and applied by operators. The BEA believes that the provisions adopted should, in particular, specify:

that section c) allow sections a) and b) to be over-ruled only for reasons of safety,

that apart from the cruise phase, access to the cockpit be limited, apart from members of the crew, to the airline's technical personnel and representatives of official bodies, within the context of their official functions, and that the cockpit access door on public transport aircraft, where there is one, must then be closed."

Though this point could not be studied in detail in the case of the accident to TJ-CAB, given the loss of the CVR data, it cannot be ruled out that the presence of two female passengers may have distracted the crew at certain times.

3 - conclusions

3.1 Findings

3.2 Probable Causes

The initial cause of the accident was the incomplete reduction of thrust on the left outer engine at the beginning of deceleration. This caused the de-activation of the automatic braking systems and the non-extension of the n° 1 thrust reverser. The inadvertent selection of full thrust on this engine after the landing created high thrust asymmetry leading to the runway excursion.

The lack of co-ordination and of joint control by the crew members, perhaps aggravated by the presence of third parties in the cockpit, contributed to the development of this situation.


In accordance with article 10 of Directive 94/56/CE on accident investigations, a safety recommendation is intended neither to apportion blame, nor to assess individual or collective responsibility for an accident or incident.

In 1995 the ICAO amended Annex 6 concerning the operation of aircraft in order to encourage the establishment of initial and on-going CRM training for operators' crews.

The accident to TJ-CAB brought to light deficiencies in the domain of task-sharing and joint control. Consequently, the BEA recommends that

the Civil Aviation Authority of the Republic of Cameroon ensure that Cameroon Airlines:

Annex 6 recommends ²that from 1 January 2002, operators of aircraft whose takeoff weight exceeds 20,000 kg establish and maintain a flight data analysis programme in the context of their accident prevention and fight safety programme².

The existence of such a system within Cameroon Airlines could not be established during the investigation, due to a failure to reply to the investigators' requests. Consequently, the BEA recommends that

the Civil Aviation Authority of the Republic of Cameroon:

The CVR data was not saved following the accident. Consequently, the BEA recommends that

the Civil Aviation Authority of the Republic of Cameroon:


Summary of technical examinations of equipment

On 9 November 2000 primary checks were carried out on the thrust controls. The results were as follows:

On 29 November 2000 checks were carried out on reverser number one using the following methodology:

On 6 December 2000 additional checks were performed on the n°1 thrust reverser electrical circuit. All electrical connections were correct. The solenoid and the relays were operating correctly.

Note: only the electrical connections between the cockpit console and the wig root could not be checked due to the cable having been torn out at the level of box P252 in the electronics bay. With regard to the Air/Ground function, if one of the relays in question had not been supplied with electrical power, symmetrical reverse would also have been affected.

Supposing that the unchecked connections were in good condition before the accident, the checks performed showed no malfunctions in thrust reverser n°1. 


2.1 Diagram of thrust reversers

2.2 Thrust reverser - Controls

APPENDIX 3 Normal procedures (extract from the T.U.)

3.1 Conduct of flight

3.2 Deceleration on landing


FDR curves

4.1 Braking on the accident flight and on the previous flight

4.2 Calculation of the runway roll distance

4.3 Engine power


Evacuation procedure


Plan of accident site

[1] All times in this report are UTC except where otherwise specified. One hour should be added to obtain the legal time applicable in metropolitan France on the day of the accident.