Swift AT72 near Alicante on Sep 9th 2017, stall in icing conditions

Last Update: February 4, 2020 / 23:40:34 GMT/Zulu time

Bookmark this article
Incident Facts

Date of incident
Sep 9, 2017

Classification
Incident

Airline
Swiftair

Flight number
UX-4050

Destination
Madrid, Spain

Aircraft Registration
EC-KKQ

Aircraft Type
ATR ATR-72-200

ICAO Type Designator
AT72

A Swift Air Avions de Transport Regional ATR-72-212A on behalf of Air Europa, registration EC-KKQ performing flight UX-4050 from Alicante,SP to Madrid,SP (Spain) with 22 passengers and 4 crew, was cleared to climb to FL170 out of Alicante and was climbing through FL165 in icing conditions when the aircraft stalled. The crew recovered the stall, descended the aircraft to FL130 later climbing to FL150 again and continued to Madrid for a safe landing.

Spain's CIAIAC reported the aircraft suffered a stall while climbing to FL170 in an icing zone. The crew recovered the stall, declared emergency and continued the flight to Madrid. The occurrence was rated a serious incident and is being investigated by the CIAIAC.

The occurrence aircraft remained on the ground for 42 hours before returning to service.

On Oct 4th 2017 the CIAIAC released more details stating the flight was planned at FL170, up to FL150 moderate icing conditions were expected. While climbing through FL130 the crew activated all three layers of protection against ice. About 7 minutes later, just before the aircraft was about to level off at FL170, the aircraft entered stall, the autopilot disconnected, the aircraft experienced a pronounced loss of altitude and pronounced "warping". The aircraft was recovered by the crew. The crew declared Mayday and received priority to Madrid. During recovery there was a blockage of the rudder, which was later identified to be related to the rudder travel limiter. There were no injuries to the occupants, one passenger suffered from anxiety during the occurrence.

On Feb 4th 2020 CIAIAC released their final report concluding the probable cause of the incident was:

The investigation has determined that the probable cause of the loss of control in icing conditions was a deficient flight management by the crew and an inappropriate use of automation.

The first officer (33, CPL, 1,431 hours total, 421 hours on type) was pilot flying, the captain (55, ATPL/CRI/TRI/IRI/FI, 18,294 hours total, 4,497 hours on type) was pilot monitoring.

The CIAIAC described the sequence of events:

The temperature at the departure airport was 28º C, with light winds and few clouds at 2000 ft. At the Madrid airport, the forecast was for 20º C and visibility in excess of 10 km. The forecast for the initial 80 NM of the flight called for light icing conditions at FL140, worsening to moderate icing conditions at reporting point NARGO (35 NM away from Alicante). The first officer was the pilot flying.

Eight minutes into the flight, with the aircraft passing FL100 (10023 ft) and climbing to FL170 (its authorized level), the temperature TAT reached a value 7º C2. Two minutes later, with the TAT at 2.7º C and climbing at a speed of 175 kt3 at 12516 ft, the crew turned the anti-icing systems on (ANTI ICING ON). Within 28 s, the aircraft’s ice build-up caution was activated (icing illuminating amber on the central panel: ICING LIGHT ON), as a result of which, 5 s later the crew turned the de-icing systems on (DEICING ON). Two minutes after the de-icing systems were turned on, at 20:20:58, with the aircraft at 14883 ft and still climbing, the FDR recorded the ICING signal went off (ICING LIGHT OFF), only to turn on again 42 s later (ICING LIGHT ON).

The aircraft continued to climb and two minutes later, at 20:22:45, at 16067 ft and 169 kt, the DEGRADED PERF caution turned on. The vertical speed during this period had decreased from 1100 fpm to 500 fpm.

Reaching 16200 ft, the vertical speed was 0 fpm and the aircraft, unable to climb, maintained the altitude for a minute and a half. The captain selected the VS (vertical speed) mode with a rate of climb of 1100 fpm in order to climb up to the top of the clouds. At the same time, at 20:25:41 (16852 ft), the crew requested permission from the Valencia TACC to climb to FL190 to “clear” weather conditions.

By 20:26:03, as a consequence of the previous actions, the aircraft had lost speed and was stalling at 153 kt. For 33 s, the aircraft descended, from 17148 ft to 15487 ft. During this descent, a minimum speed of 151 kt, an angle of attack of up to 19.6º, a pitch angle of up to -11º, and left and right bank angles of 58º and 39º, respectively, were recorded.

Stall was recovered at 20:26:36, 15487 ft and 195 kt.

At 20:26:26, still descending to 15980 ft, the crew asked the Valencia TACC for an “immediate descent”. The controller cleared them to descend at their discretion, reporting that the minimum radar in the area was FL70. The aircraft continued descending and at 20:28:04, the crew informed Valencia TACC they were descending to FL70, holding on course and asked the controller if there were any significant weather phenomena, receiving a negative reply. At that point, ATC inquired about the reason for the descent, to which the crew replied; “we stalled due to icing”. The aircraft was at 14932 ft and descending (20:28:38).

After this information, the controller asked about their intentions and informed them of the location of the Valencia Airport with respect to their position (20:28:58).

At 20:31:32, at 13826 ft, the crew informed the Valencia TACC they were clear of the clouds and were proceeding to CENTA. This report was interpreted by the controller as meaning they had solved their icing problem, which is how the controller relayed it to the adjacent controller in the Madrid ACC, to whom he would soon be transferring the aircraft.

At 20:31:47, the aircraft was at 13752 ft and was stopping the descent.
At 20:33:23, with the aircraft at 13778 ft, the first officer took photos of the cockpit windshield and the right wing, showing the presence of ice on the aircraft (section 1.7.4). One minute later, the crew asked the Valencia TACC to continue direct to PRADO to avoid clouds, and initiated a left turn. At 20:35:25, the DEGRADED PERF caution cleared. It had been activated for 12 min 50 s. Three minutes later, ATC asked them to climb to FL150 due to the minimum altitude of the Madrid ACC sector, a flight level the crew accepted. At 20:41:27, the aircraft was transferred to the Madrid ACC.

At 20:42:26, after the controller asked if the icing situation was solved, the crew reported they had taken on a lot of ice, that something was wrong, the problem was still present but they were stable. Five minutes later, at 20:47:22, the crew reported they had not solved the problem yet and requested to descend. At 20:48:25, they reported a MAYDAY and requested direct to Madrid due to continuing “problems with the controls”.

At 20:50:29, ATC asked if they needed help on the ground, to which the crew replied no. After several conversations on which runway to use in Madrid, and keeping in mind the crew’s request to land on runway 32R, this runway was opened, even though the runway in use was 32L. At 20:57:31, once in contact with the approach controller, the crew asked if there had been any reports of turbulence in the area where they were flying, receiving a negative reply. The aircraft was cleared to make the 32R approach.

At 21:07:00 the aircraft touched down at a speed of 140 kt. The temperature was 22º C and the firefighters were on the scene. During the landing run, the captain, who had been the pilot flying since the event, was heard saying “the left pedal is not working, the pedals are jammed”. This problem did not affect the landing run or the taxi, and the aircraft cleared the runway and taxied under its own power.

After leaving the runway, the purser requested an ambulance from the crew for one of the passengers who had been given oxygen during the flight. The request was relayed by the first officer to the tower, which initiated the medical protocol.

The aircraft stopped after taxiing for 7 min, at 21:14:04, at remote stand number 4, where the passengers disembarked. The passenger for whom the ambulance had been requested was treated for anxiety by airport medical personnel.

The CIAIAC analysed:

Flying in icing environments requires crews to constantly monitor external conditions. The problem lies in detecting the severity of the icing, and specifically, in how to identify if severe icing is taking place. Due to design limitations, weather radar is unable to detect this phenomenon (as happened in this flight) and crews must rely on other detection methods. Two such methods are included in the procedure (SEVERE ICING):

- First, the visual observation of ice build-up in certain areas of the aircraft:

The crew’s evaluation of the ice that was building up during the climb led them to conclude that they were not flying in severe icing conditions. The photos taken, though 8 min later, in fact do not show extreme or remarkable conditions. From the point of view of the visual indications that are normally associated with severe icing, and which the procedures describe (ice covering much or all of the windshield, for example), these were not present in this incident or in the incident in the United Kingdom involving aircraft G-COBO.

- Second, the monitoring of the aircraft’s performance, specifically, the drop in speed or climb rate.

Three minutes before the event, the aircraft was unable to keep climbing, and 35 s before, the speed began to drop. These are clear indications, and are described as such in the procedure, that the aircraft’s performance was degrading. Moreover, in the 4 min before the stall, the crew were given cautions that ice was building up on the aircraft (ICING LIGHT ON) and that the drag was increasing (DEGRADED PERF ON).

In other words, even if the visual indications may not have been as expected for severe icing, the remaining information (cockpit cautions and aircraft’s behavior) provided a clear indication that the aircraft’s performance was degrading and that the aircraft was flying in uncertified conditions. The crew only considered the visual indications and did not take into account the rest of the information, continuing to climb in severe icing conditions. Modifications included on the icing procedures after the event changed the priorities.

The CIAIAC was critical of the climb management commenting the climb target speed of 170 KIAS was not maintained, instead 180 KIAS and later 176 KIAS were used. The crew used the PITCH autopilot mode rather than the IAS autopilot mode, although none of the conditions for PITCH mode existed. When the TAT dropped through +7 degrees C, the IAS mode should have been used and the anti-ice systems should have been turned on, but were turned on only two minutes later when the TAT was already at +2.7 degrees C. When the icing light illuminated the crew turned de-icing systems on within 5 seconds (satisfactory).

The CIAIAC continued analysis:

When the degraded performance caution turned on, the caution informing of ice build-up on the aircraft (ICING LIGHT ON) was already on. It took the crew 23 s to turn off the master caution that is triggered by any caution. Of the steps specified in the procedure:

- The first step was already done, since the de-icing systems had been on for 4 min.
- The second step was to maintain the speed above 159 kt, which was also satisfied since the speed was 169 kt.
- The third step was not done since the autopilot was not disengaged.
- The fourth step, which involves conducting the severe icing procedure, was not done. At that point, of the three applicable conditions, at least one was not checked:

- the crew stated that there was no severe icing,
- the speed could be maintained above the red bug + 10 kt (in fact, it was at red bug + 20 kt)
- the abnormal feel of the flight controls could not be checked since the autopilot was not disengaged.

In other words, even though they had indications in the cockpit that the aircraft’s performance was degraded, they did not carry out the associated procedure.

Sixteen seconds after the degraded performance caution turned on, the aircraft reached 16200 ft, and for 90 seconds it held that altitude, unable to climb further. During this time, the crew changed from PITCH mode (which is prohibited in icing conditions and which had been used to climb since takeoff) to IAS mode (the recommended mode) at 176 kt (not at 170 kt). In this new configuration, the aircraft was able to maintain the 176 kt set, but it was not able to climb.

This (the unexpected drop in climb rate) is one of the indications that is specified for determining that the aircraft is operating in severe icing conditions. Rate of climb This condition persisted for 1.5 minutes and was not taken into consideration by the crew. In this case, as noted in point 2.1, only the visual indications were considered, and the indications involving the aircraft’s performance were ignored.

Neither the severe icing nor the degraded performance procedure was carried out.

Not only was the SEVERE ICING procedure not conducted, even though the inability to climb indicated this condition, but the crew commanded a series of actions to force the aircraft to climb beyond its capabilities, which led to the stall. These actions were as follows:

- Changing the power management selector from CLIMB to MCT.
- Changing the autopilot flight mode from IAS to VS, which is prohibited during climb in icing conditions, selecting an initial climb rate of 500 fpm, and then, seeing that the aircraft did not react as expected, increasing the rate to 1100 fpm (above published 800 fpm).

By way of these actions, the crew attempted to climb through the cloud layer they were flying in. Their intention was clear, and was further confirmed by selecting a new flight level (FL190) and requesting a new clearance from ATC. According to the crew’s statement, they were about to clear the cloud layer and thought that climbing (and not descending) was the best option to escape the icing, a decision that was also present in the G-COBO accident. These two actions show that the crew were completely focused on this objective and ignored the cautions that the aircraft was providing:

- The aircraft was unable to climb.
- Ice was accumulating (ICING LIGHT ON).
- The aircraft’s performance was degrading (DEGRADED PERF ON).

The crew forced this situation for one minute, getting the aircraft to climb 500 ft. This was followed by the last 35 s before the event, during which the aircraft performed in a manner consistent with an approach to stall:

- Sudden drop in speed from 174 kt to 151 kt.
- Increase in the pitch, bank and attack angles.
- Vertical accelerations ranging from 1 g to 0.9 g.

The lowest speed reached was 151 kt, 2 kt above the icing (red bug) speed. This is mentioned because the crew, in their statements, said that the flight had been maintained above this value during the entire flight. It should be noted, however, that on the one hand, the minimum speed to maintain in severe icing conditions is this value plus 10 kt; and on the other, speed is not the only parameter to monitor in this environment to ensure that the aircraft is free from the effects of icing.

According to their statements, it was the captain who changed the power management switch from CLIMB to MCT, even though the first officer was the pilot flying, in violation of the specified task assignments. In this case, the vast difference in experience between them (18000 h versus 1400 h) affected, as he himself stated, the first officer’s assertiveness to oppose or argue the actions or decisions made by the captain.

The CIAIAC analysed that the captain did not seem to identify the stall for 21 seconds providing nose up inputs while the first officer pushed the controls firmly nose down. The crew subsequently recovered the aircraft 33 seconds after the entry into the stall and after losing 1661 feet of altitude. The CIAIAC analysed with respect to the stall recovery procedure:

From this procedure:

- The first item, which involves pushing the stick forward firmly, was only done by the first officer. The captain made the opposite input.
- The second item, to lower the flaps to 15º, was not done.
- The power management (throttle lever to NOTCH and the power management switch to MCT) was also not done. The lever was placed in MAX PWR.
- The report to ATC involved a call by the first officer requesting an immediate descent.

This all indicates that the stall recovery procedure was not carried out. Since the cockpit communications were unavailable, it was not possible to identify what decision-making process led the crew not to carry out this procedure.

Although the crew recovered the aircraft from the stall in 33 s, they still were not in control of the aircraft. Even though the captain was operating the aircraft now, ordering the nose down using his control stick, the trim and power selected made the aircraft climb again at 33 s, which caused it to lose 29 kt of speed and increased its pitch angle to 10º.

In conclusion, the actions recorded suggest that the captain did not identify the stall until 21 s had elapsed, or that if he did identify it, he did not take the only essential action that is required, which is to push the stick forward to lower the angle of attack and recover speed, or any of the other actions in the stall recovery procedure.

The CIAIAC analysed with respect to the rudder trouble encountered:

The investigation was unable to determine the cause of the problems the crew encountered with the rudder while decelerating on the runway. The results of the inspections of the TLU components replaced after landing do not seem to provide an answer as to the jamming reported by the crew. The aircraft has not exhibited similar problems since, meaning that if it actually was caused by a physical issue, it was solved by replacing those components.

Another possible explanation is that the TLU may have been in the HIGH SPEED position, meaning its travel was limited, which gave the crew the sense that it was jammed. Considering how, until 25 s before landing, the speed did not fall below 180 kt, the transition from HIGH to LOW SPEED would have started practically at that moment. This may have limited the system’s response time, preventing it from transitioning to LOW SPEED in time.
Incident Facts

Date of incident
Sep 9, 2017

Classification
Incident

Airline
Swiftair

Flight number
UX-4050

Destination
Madrid, Spain

Aircraft Registration
EC-KKQ

Aircraft Type
ATR ATR-72-200

ICAO Type Designator
AT72

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.

Free newsletter

Want to know more and stay ahead? Get our free weekly newsletter and join 5354 existing subscribers.

By subscribing, you accept our terms and conditions and confirm that you've read our privacy policy.

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.

Virtual Speech 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.

Get updates

Never miss an article from AeroInside. Subscribe to our free weekly newsletter and join 5354 existing subscribers.

By subscribing, you accept our terms and conditions and that you've read our privacy policy.

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