Air France B772 at Paris on May 22nd 2015, 100 tons missing in inserted takeoff weight

Last Update: January 22, 2019 / 16:54:02 GMT/Zulu time

Bookmark this article

Photo of Air France F-GUOC, Boeing 777-200 — Air France F-GUOC, Boeing 777-200 (Photo credit: airlines470 / Flickr / License: CC by-sa)

Incident Facts

Date of incident
May 22, 2015

Classification
Incident

Airline
Air France

Departure
Paris Charles de Gaulle, France

Destination
Mexico City, Mexico

Aircraft Registration
F-GUOC

Aircraft Type
Boeing 777-200

ICAO Type Designator
B772

An Air France Boeing 777-200 freighter, registration F-GUOC performing freight flight AF-6724 from Paris Charles de Gaulle (France) to Mexico City (Mexico), was accelerating for a balanced takeoff from Charles de Gaulle Airport's runway 26R when the crew detected the aircraft wasn't accelerating quickly enough and firewalled the engines. The aircraft took off without further incident, the crew discovered their takeoff weight had been programmed 100 tons below actual takeoff weight resulting in incorrect takeoff power and reference speed computations, and completed a safe landing in Mexico City about 11:15 hours later.

The airline reported: "Air France confirms that a data insertion error on take-off of cargo flight AF6724 operated by a Boeing 777 F from Paris-Charles de Gaulle to Mexico on 22 May 2015 led to insufficient acceleration at the beginning of take-off. The crew, who noticed this slow acceleration, immediately reacted by applying full thrust. The aircraft took off normally and the flight continued to its destination. The crew spontaneously declared this event by ASR (Air Safety Report) and informed their superiors." The French BEA was informed.

On Jun 2nd 2015 the French BEA reported in their weekly bulletin, that the crew used 243 tons of takeoff weight instead of 343 tons for computation of their takeoff performance, the resulting speeds were input into the flight management system. During rotation for takeoff the crew noticed the aircraft did not become airborne, firewalled the engines, established maximum pitch possible without tail strike, lifted off, climbed out to safety and continued to destination. The BEA rated the occurrence a serious incident and opened an investigation.

On Dec 21st 2018 the BEA released their final report in French only (Editorial note: to serve the purpose of global prevention of the repeat of causes leading to an occurrence an additional timely release of all occurrence reports in the only world spanning aviation language English would be necessary, a French only release does not achieve this purpose as set by ICAO annex 13 and just forces many aviators to waste much more time and effort each in trying to understand the circumstances leading to the occurrence. Aviators operating internationally are required to read/speak English besides their local language, investigators need to be able to read/write/speak English to communicate with their counterparts all around the globe).

The final report concludes the probable causes of the serious incident were:

The overflight at low height of the runway end during takeoff was the result of a take-off undertaken with erroneous parameters (low take-off speeds, flap setting and insufficient thrust).

The false parameters entered in the FMS and used for the take-off were the result of a performance calculation based on a mass of 100 tons less than the real mass of the aircraft.

The error of 100 tons was the result of erroneous calculations by the crew as well as entering the wrong mass into their performance tool (EFB)

The following elements have contributed to the absence of detection and the propagation of the error of 100 tons:

- The manipulation by the crew in various formats, tools and forms for the calculation of take off data

- The non-realization of orders of magnitude, partly related to the increasing use of performance tools

- The procedures include various elementary checks which proved not sufficiently robust, not taking into account sufficiently the operational context and the working procedures of the crew. These procedures rely on a double calculation supposedly independent whereas a simple verbalization can default the independence. The procedures do not include a means of detection of gross errors or simultaneous checks of the three media using mass data. (Load Status, Performance Tool and FMS)

- The absence of a system for detecting and preventing such gross errors and to warn the crew that the measured performance during take-off roll is insufficient, such as available on most commercial airliners, on the aircraft in question.

The BEA reported the flight crew consisted of captain (63, ATPL, 21,331 hours total, 5,272 hours on type), being pilot monitoring for the departure, and three first officers. The first first officer (49, ATPL, 9,717 hours total, 1,328 hours on type) was pilot flying for departure, during departure preparation these two pilots remained in the cockpit, while the other two first officers conducted the walk around.

The first officer performed mental arithmetic to compute the takeoff mass and erroneously computed 243 tons for takeoff weight. He entered the 243 tons into his electronic flight bag and derived the according takeoff paramenters.

The captain correctly computed the takeoff mass at 343 tons, however, committed a typo while entering the mass into the electronic flight bag typing 243 tons.

First officer and captain subsequently compared the results of their electronic flight bags, found the numbers in agreement and typed the results into the FMS.

The crew prepared for a balanced takeoff using an assumed temperature of 37 degrees C, flaps 5 degrees and V1=143 KIAS. The captain expressed doubt about the speeds as they were 20 knots lower than computed by the FMS, the crew decided to postpone completing the speed data until fueling was complete.

After loading was complete the first officer recomputed the takeoff mass again committing the same error and reaching 241.5 tons of takeoff mass. He then compared the performance data and found them correct.

The captain compared the takeoff mass figures with the FMS and found them correctly entered into the FMS. The FMS had computed the figure from the zero fuel weight of the aircraft, which had been correctly entered, and added the fuel weight as measured by the aircraft.

Once fueling was completed, the crew resumed entering the speeds into the FMS. At this stage, the reference speeds calculated by the FMS were no longer shown and the crew tried several times to get them to reappear. Nobody in the crew understood why they had disappeared. The commander and FO once again entered their weights starting with the ZFW without result. Finally, the crew entered the speeds calculated by the performance calculator (V1=143kt, Vr=152kt, V2=156kt). The flaps were set at 5°. While one of the first officers was surprised by the figures he did not question them.

The aircraft taxied to runway 26R and commenced takeoff.

The rotation was initiated at 154kt. Right after rotation, the crew felt the airplane sink. Five seconds after the start of the rotation, the tail-strike protection activated with maximum authority. The main gear was still in contact with the runway and the pitch established at 9° to avoid a tail strike. 8 seconds after the activation of the tail strike protection, full power was applied and “TOGA” announced. The airplane was at a height of 16 ft with a pitch of 13° and speed of 189kt. The airplane accelerated brusquely. The commander was concerned about the height and vertical speed. He commanded the gear to be retracted once the vertical speed reached +1500 fpm. The pitch increased to 16° and the aircraft passed the runway end at 172 ft.

The stick shaker was not triggered during departure.

During climb the crew started analyzing the situation. The error was found when the First Officer and pilot flying rechecked his performance tool. He was relieved of his position when the airplane passed FL200. The crew subsequently discussed a return to base but decided to continue the flight as they felt up to it. After arrival they conducted a walk around to verify there had been no tail strike. They subsequently informed the company and were relieved after the flight.

The BEA analysed that the correct performance data for departure would have been: V1=167kt, Vr=175kt, V2=179kt, Flex Temp of 37°C, Flaps 15° and 98.5% N1 thrust, while the crew actually used: V1=143kt, Vr=152kt, V2=156kt, Flex Temp 58°C, Flaps 5° and 89.3%N1.

When the crew changed the figures in the FMC following the receipt of the final load sheet, the FMC was not able to calculate V-Speeds for Flaps 5 as for the weight entered flaps 15 would have been necessary. Consequently the message V-SPEEDS UNAVAILABLE was displayed. This should have alerted the crew that there was a problem with the configuration but failed to do so. The BEA reasons that had the FMS displayed speeds for the weight, the crew would likely have noted the difference of more than 20 kts.

The BEA analyzed that the way weight figures are calculated and handled during pre-flight procedures does not adequately rule out errors like the sequence of events which led to this incident. It points out that in recent years there have been several similar incidents of which two on the same type where errors of similar magnitude were made.

The BEA annotated that there is no system available on the Boeing 777 which alerts the crew to weight discrepancies. The procedures used by Air France to cross check weight entries in the performance tool and the FMS are insufficient.

The BEA issued 3 safety recommendations to Air France, one to France's DGCA and two to Boeing.

On Jan 22nd 2019 the BEA released their English version of the final report.