GOL B738 near Sao Paulo on Oct 16th 2011, partial loss of instruments

Last Update: January 15, 2014 / 22:03:16 GMT/Zulu time

Bookmark this article
Incident Facts

Date of incident
Oct 16, 2011

Classification
Incident

Aircraft Type
Boeing 737-800

ICAO Type Designator
B738

About 12 hours after The Aviation Herald picked up on the final post CENIPA changed their web page removing the sentence, that no investigation had been conducted and no report was available, and adding the final report in Portugese instead. The report concluded:

It was found that the air hose of the left hand air data module had been installed however not been properly connected permitting disconnection and air leakage resulting in incorrect data produced by the air data module and incorrect indications of airspeed and altitude for the left hand (pilot flying) instruments.

After takeoff the pilot flying engaged Autopilot and Autothrust, which had dropped offline during the takeoff run, without considering the implications of the disagreeing indications of airspeed and altitude, which took the aircraft to a critical pitch attitude for this phase of the flight.

The prevailing weather conditions were instrument meteorologic conditions shortly after takeoff (600 feet AGL). It is possible that the crew had understood the situation more quickly in visual meteorologic conditions.

The tasks in the cockpit were not properly managed and distributed during the emergency. There was no clear definition who was flying the aircraft, who should identify the failures and who should communicate with ATC. The pilot in the right hand seat just accumulated tasks until being overwhelmed despite an additional pilot occupying the observer's seat assisting.

The training of unreliable airspeed had been performed during the initial simulator training, however, the failure that occurred with the aircraft presented itsself differently and was more complex than had been trained in the simulator, which probably hindered the crew to correctly identify the problem.

There was complacency and excessive trust in systems by both pilot flying and pilot monitoring evident by the use of autopilot and autothrust during the emergency.

The complexity of administrative and management structure of the maintenance bases, divided between Sao Paulo Congonhas,SP and Belo Horizonte,MG (Brazil), generated difficulty in communication, coordination and management of various tasks to be performed by the maintenance sectors of the operator, this situation deteriorating from a latent condition to an active fault in this incident.

Several internal processes of maintenance activities required the preparation of reports in English, the professionals reported however they did not receive training by the company and instead used translation tools available on the Internet to help and understand information on maintenance procedures.

CENIPA reported that the crew comprised of two captains, the captain in the left hand seat was pilot flying and being checked out by the training captain occupying the right hand seat and assuming the role of pilot monitoring. Both pilots held valid ATPLs. Upon the insistence of the check captain a third pilot, who wanted to keep his seat in the cabin, took the observer's seat in the cockpit.

The aircraft was brand new and had accumulated 24 hours of operation in 14 flight cycles so far, the aircraft had not undergone an inspection yet.

At the time of the occurrence the weather showed overcast cloud with the base at 700 feet, the winds came from 130 degrees at 5 knots.

CENIPA reported that after landing the aircraft, and especially the blackboxes, remained powered resulting in the loss of the cockpit voice recordings.

According to the flight data recorder following can be concluded:

While the aircraft accelerated through about 45 knots on takeoff, the air hose disconnected from the left hand air data module. As the flight data recorder did not record the right hand instruments it could not be established whether at that point the indications remained constant.

About 8 seconds after the takeoff acceleration began the aircraft accelerated through 80 knots KIAS, 75 knots over ground, and 45 knots indicated on the left hand air speed indicator. Due to the lack of cockpit voice recordings it could not be determined whether the airspeeds disagreeing by more than 30 knots were recognized. According to the flight crew operating manual the pilot monitoring should monitor the airspeed for any anomaly and should call out 80 knots as soon as the aircraft reached 80 KIAS, the pilot flying should check his airspeed indicator and call "VERIFY" if his airspeed indication agrees. In case of a disagreement, if the accuracy of either primary airspeed indication is in doubt, the stand by airspeed indicator should be referenced, another aid is the ground speed indication. An early recognition of an airspeed indication malfunction is important to make a stop/go decision.

When the aircraft accelerated through 92 knots over ground the master caution activated, at that point the left airspeed indicator showed 48 knots.

The aircraft rotated early when the aircraft had reached 138 knots over ground, the left hand airspeed indicator (ASI) showed 47 knots.

The aircraft became fully airborne at a ground speed of 153 knots, the left hand ASI showed 69 knots, the pitched reached 10.55 degrees nose up.

During the initial climb, at a height of 925 feet, ground speed of 168 knots, pitch angle of 16.17 degrees nose up, the autopilot was engaged, the left ASI showed 140 KIAS.

The pitch increased to 26.07 degrees nose up, the ground speed began to decay reaching 150 knots at 1793 feet, the rate of climb was 4811 feet per minute, the left ASI showed 188 KIAS, the engines were operating at 101% N1.

At 2931 feet AGL the pitch was 22.5 degrees, the left ASI showed 213 KIAS, ground speed was 133 knots and the climb rate reduced to 71 feet per minute.

The aircraft reached 3107 feet AGL then began to rapidly descend, pitch angle was 16 degrees nose up, the left ASI showed 199 KIAS, the ground speed was 133 knots, the rate of descent was 1077 feet per minute, a speed warning activated, the autopilot disconnected, the flaps were retracted and the engines were operating at 75.5% N1.

The rate of descent increased to 3069 feet per minute at 2409 feet AGL, at that point the ground speed was 160 knots, the left ASI showed 196 KIAS, the pitch was 9.49 degrees nose up and the engines accelerated to 102.9% N1.

At a height of 2168 feet AGL the aircraft stopped descending and engaged in a climb of 854 feet per minute at 11.43 degrees nose up, 184 knots over ground, 201 KIAS. Continuing the shallow climb the aircraft accelerated reaching 223 knots over ground at 3184 feet AGL and then increasing the rate of climb settling in a normal envelope of flight reaching 7741 feet MSL and 267 knots over ground with 341 KIAS shown during the diversion to Viracopos.

During the descent towards Viracopos the indicated airspeed dropped to 54 KIAS at 326 feet AGL with the speed over ground remaining near constant between 144 and 148 knots with flaps at 40 degrees, the ASI showed 45 knots during roll out.

CENIPA summarized the pilot interviews stating that after the aircraft became airborne the pilot flying reset the autothrottle despite seeing the airspeed disagree warning, after the autothrottles were reset the altitude disagree warning also illuminated. The observing pilot pointed out the speed trend was high aiming towards 250 KIAS, still with flaps at 10 degrees the engines were slowed. The stick shaker activated - the pilots could not recall if flaps had already been retracted or were still deployed -, the check captain attempted to calm the pilot flying stating "stay calm, she is flying!", the observing pilot responded "she's not flying, need engines", which resulted in the immediate reaction by the pilot flying to accelerate the engines again. During the approach to Viracopos the crew observed the left airspeed gradually decreasing until the stick shaker activated, the landing was performed with the stick shaker continuously operating.

When the check captain was asked whether he did cross check instruments he stated that he advised to not trust any of the instruments and reduced the intensity of the instrument lightings. The check captain did positively establish that indicated airspeed and altitude disagree warnings had activated.

CENIPA reported the most comprehensive report came from the observer. The observer did not recollect anything of the takeoff run, he was reading at that time. Upon the master caution, which was triggered by both engine electronic control units switching to alternate mode, he became engaged in monitoring the instruments and recognized that the airspeeds disagreed substantially at the 80 knots call. During climb out there were fluctuations of speed indications on the left hand side, he realized they were approaching stall by both indications and the onset of according aircraft vibrations. They were descending at about 2800 feet per minute at the time when he spoke out "it's not flying, need engines", the instant response by the pilot flying was to apply full thrust. The observer recalled that the active alarms were indicated airspeed and altitude disagree as well as electronic engine control in alternate mode. Some time into the flight the check captain assumed the role of pilot flying and performed the landing into Viracopos.

A post flight inspection of the aircraft showed, that the air hose ASSY - P/N B394DDB0106D at the air data module had been installed but was left unconnected allowing a leak of air supplied from the left hand pitot tube. After the hose was connected and a leak test had been concluded the system showed no anomaly. Maintenance nonetheless replaced the left air data module including static ports and the left hand angle of attack sensor, swapped the right hand air data module including static port and the right hand angle of attack sensor with the one of another aircraft, performed another test of the air data systems with no faults found.

Subsequently simulation tests were performed with the hose being disconnected. There were scenarios in which there was no disagreement of speeds and altitudes, in other scenarios there were substantial disagreements exactly as occurred during the occurrence flight. The tests confirmed the disagreeing airspeed and altitude alerts were the result of that hose being unconnected.

Prior to the occurrence flight the aircraft had been flown from Seattle to Belo Horizonte. At Belo Horizonte the aircraft was inspected for receipt and some adoptions performed to adjust the aircraft to the operational standards of the airline. Subsequently the aircraft positioned from Belo Horizonte to Congonhas, during that flight the weather radar failed, maintenance identified incorrect software had been installed on the radar system.

To resolve this issue it was necessary to access the radar, which in turn required to unlock and disconnect the hoses from the pitot tube to the left hand air data module.

After the radar had been fixed the aircraft was dispatched for the incident flight.

CENIPA analysed that at the 80 knots call the pilot flying did not reject takeoff, the reasons for not rejecting takeoff could not be established. On the other hand, the pilot monitoring, actually commander of the aircraft, did either not recognize the speed disagreement or if he did detect the anomaly permitted the pilot flying to continue takeoff despite the substantial difference of indicated airspeeds. The correct decision would have been to reject takeoff, the runway conditions would have permitted a safe stop from about 75 knots over ground.

When the master caution occurred at 92 knots over ground the runway conditions would have made a rejected takeoff more critical.

During climb out the crew selected 190 KIAS into the master control panel and engaged autopilot and autothrust. Due to the erratic air speed indications now showing higher speeds than present the aircraft pitched up to very high nose up angles, while the crew retracted flaps at the same time. The retraction of flaps, decaying airspeed with the actual airspeed substantially lower than indicated and the high pitch angles resulted in a high angle of attack and the approach to stall, the stick shaker activated. At this stage the pilots still had not identified the actual problem and the cause of the indications.

CENIPA analysed that it was the pilot on the observer seat who identified the situation and with his remark "it is not flying, need engines" finally initiated the recovery of the aircraft to a stable flight envelope.
Incident Facts

Date of incident
Oct 16, 2011

Classification
Incident

Aircraft Type
Boeing 737-800

ICAO Type Designator
B738

This article is published under license from Avherald.com. © of text by Avherald.com.
Article source

You can read 2 more free articles without a subscription.

Subscribe now and continue reading without any limits!

Are you a subscriber? Login
Subscribe

Read unlimited articles and receive our daily update briefing. Gain better insights into what is happening in commercial aviation safety.

Send tip

Support AeroInside by sending a small tip amount.

Related articles

Newest articles

Subscribe today

Are you researching aviation incidents? Get access to AeroInside Insights, unlimited read access and receive the daily newsletter.

Pick your plan and subscribe

Partner

Blockaviation logo

A new way to document and demonstrate airworthiness compliance and aircraft value. Find out more.

ELITE Logo

ELITE Simulation Solutions is a leading global provider of Flight Simulation Training Devices, IFR training software as well as flight controls and related services. Find out more.

Blue Altitude Logo

Your regulation partner, specialists in aviation safety and compliance; providing training, auditing, and consultancy services. Find out more.

AeroInside Blog
Popular aircraft
Airbus A320
Boeing 737-800
Boeing 737-800 MAX
Popular airlines
American Airlines
United
Delta
Air Canada
Lufthansa
British Airways