Malaysia B738 at Sibu on Apr 8th 2017, runway excursion, nose gear collapse

Last Update: May 11, 2020 / 21:51:54 GMT/Zulu time

Bookmark this article

Incident Facts

Date of incident
Apr 8, 2017

Classification
Accident

Cause
Runway excursion

Airline
Malaysia Airlines

Flight number
MH-2718

Departure
Kuala Lumpur, Malaysia

Destination
Sibu, Malaysia

Aircraft Registration
9M-MXX

Aircraft Type
Boeing 737-800

ICAO Type Designator
B738

A Malaysia Airlines Boeing 737-800, registration 9M-MXX performing flight MH-2718 from Kuala Lumpur to Sibu (Malaysia) with 63 passengers and 6 crew, landed on Sibu's runway 13 in heavy rain at about 22:10L (14:10Z) but went off the left side of the runway and came to a stop on soft ground about 1200 meters/4000 feet past the runway threshold with the nose gear collapsed, the aircraft heading about perpendicular to the runway center line with the nose towards the runway edge. No injuries are being reported, the aircraft received substantial damage.

The airline reported all occupants evacuated safely via two slides, no injuries occurred.

The airport reported the runway suffered a number of holes as result of the accident, that could be repaired only after the aircraft had been moved off the runway on Sunday (Apr 9th) by about 5pm local time. The runway was subsequently repaired and returned to operation on Monday Apr 10th by about noon. A total of 48 flights were cancelled as result of the accident affecting about 3700 travellers.

Malaysia's Aircraft Accident Investigation Bureau (BKSU) have opened an investigation into the accident and dispatched investigators on site.

On May 11th 2020 the final report releeased by BKSU at an unknown date was discovered. The report concludes:

CAUSAL FACTORS

- A sudden increase in the intensity of rain while approaching the runway at night resulted in the significant reduction of the PF‟s visual reference. Under these conditions and without the runway centerline lights, the PF did not detect the lateral movement of the aircraft in time to correct the displacement from the runway centerline.

- Pilot induced oscillation resulting in the progressive input of roll angle to the right of up to 6 degrees during flare maneuver without any considerable left rudder input. This resulted in a drift in the aircraft heading towards the right side of the runway. The resultant drift angle recorded on touchdown was 4 degrees (Runway heading is 129o while the touch down heading was 133o ).

- The PF had likely lost his positional awareness with reference to the runway edge due to the degraded visibility, hence, did not exert sufficient and timely rudder application to regain the runway centerline before departing the surface of the runway.

CONTRIBUTORY FACTORS

- Continuous rain with changing intensity over the airfield throughout the approach and landing.

- Inadequate risk assessment on the prevailing weather conditions made by the flight crew through the established TEM briefing as stipulated in the MAB OM (A).

- The elevated pitch attitude after touchdown may have caused the reduction of visual reference to the runway. This would reduce the effectiveness of braking and cornering ability of the aircraft due to reduced weight on the main landing gear.

- Lack of assertiveness by the PM in getting the attention of the PF to the developing and impending deviation from centerline. Currently, there is no standard callout stipulated in the MAB OM (A) with reference to the calling out of centerline deviation.

- Absence of centerline lights at night in the reduced visibility conditions due to the heavy rain made it difficult for the PF to acquire the required visual reference to conduct a safe landing and roll out.

- The “Black hole effect” is prevalent during night approaches into Sibu due to the lack of lightings and visual reference surrounding the airport. This could result in an optical illusion leading towards a false pitch or bank perception, especially during approaches with reduced visibility such as in heavy rain or fog.

- Rubber deposit on the runway could have caused the runway centerline markings to become less discernable, especially when the runway surface was wet and in reduced visibility condition.

- Inadequate updates of weather reports that were provided by the ATC controller to the pilots as and when considerable changes to the weather conditions over the airfield were observed.

- Use of single RVR readout as means of reporting the visibility in heavy rain or thunderstorm at night did not alert the pilots on the severity of precipitation at the airport. The transmissiometer only measures the horizontal visibility at a specific range around the unit located near runway thresholds.

The captain (46, ATPL, 8,438 hours total, 1,551 hours on type) was pilot flying, the first officer (27, CPL, 1,911 hours total, 1,711 hours on type) was pilot monitoring.

The BKSU summarized the sequence of events:

The investigation revealed a number of factors that have caused and contributed towards the incident that occurred during the night hours. Sudden increase in rain intensity at the Sibu Airport, i.e. from moderate to heavy at below 100 ft above ground level (AGL) had caused the flight crew to have reduced visibility of visual references and runway lightings. There was no centerline light available on Runway 13/31 Sibu. The Pilot Flying (PF) was informed of the crosswind component from the right during the final approach by the Pilot Monitoring (PM). The wind velocity had however reduced to less than 2 knots from the initial 6 knots as the aircraft was approaching the runway. The PF nevertheless applied crosswind technique for the landing by oscillating the control wheel to the right, consistent with the „perceived‟ right crosswind. This had introduced a heading drift of 4o to the right that resulted in a heading of 133o. In actual fact, the runway heading is 129o. The aircraft touched down at approximately 10 m to the right of centerline with 6o of bank angle. There was minimal rudder input to regain the runway centerline track. The aircraft left the runway surface at 720 m from Threshold Runway 13, approximately three seconds after the aircraft had initially contacted the runway.

In detail the BKSU described the sequence of events:

The crew had already gone around on their first approach when at about 600 feet AGL the crew could not see the PAPI lights.

At 2215:44 LT, PM announced “1,000 ft Stable” based on the stable approach parameters observed. A/P was disconnected at approximately 800 ft ASL. Approaching 500 ft, PM announced the current wind information which was “Right crosswind 6 kts”.

In response to mildly changing wind conditions, PF made frequent but small aileron adjustments to maintain the aircraft on the localizer of the ILS. This was evident from the frequent oscillation of the control wheel during the manual flying segment.

Below 300 ft radio altitude (RA), the FDR recorded a gradual reduction on vertical descent rate from 800 ft per minute (fpm) to 550 fpm over approximately 12 seconds. Wind speed was recorded at 3 kts at this altitude.

At the minimum descent altitude (MDA) of 290 ft, PM announced “Minimum - PAPI”, followed by PF‟s response to continue the approach with reference to the PAPI lights and runway edge lightings.

At 2216:44 LT, PM called “three whites” indicating aircraft was slightly high on profile. This occurred at approximately 180 ft RA.

PF immediately announced “correcting” to indicate his response to the call by PM, and increased the rate of descent up to 920 fpm to correct the profile.

At 200 ft RA, the FDR recorded glide slope deviation of - 1 unit (1 dot above profile), while PF continued with the profile correction.

At approximately 100 ft RA, there was a sudden and intense rainfall which caused visibility to decrease rapidly. According to the pilots, they were still able to see the runway edge lights and PAPI, hence the PF continued the approach to land. The wipers were operating at HIGH speed setting at that time.

This reduction in visibility was later confirmed based on the data retrieved from the RVR database, where the visibility dropped from 1,400 m to 900 m as the aircraft was about to touch down. The intensity of the runway edge light and PAPI was set at 100 percent by ATC controller. This was done following the go-around from the first approach as the crew declared that they were not able to see the PAPI or the runway edge lights. (Note: The visibility further reduced to down to 450 m over the next few minutes while evacuation was in progress.)

According to the FDR, the aircraft crossed over the threshold at 42 ft RA and was flying over the runway centerline.

At 30 ft RA, the aircraft was still over the centerline. PF initiated flare at 25 ft RA. Wind speed at that time had decreased to 1 kt.

Below 20 ft RA, a slight bank between 2° and 6° was progressively introduced (The aircraft touched down with 6° bank angle). A very slight left rudder input was evident from the FDR data just prior to landing (Refer to Appendix B for graphical interpretation).

The aircraft heading was offset by 4o to the right on touchdown.

From the interview, PF mentioned that he could not recall any bank angle being introduced.

PM observed that the aircraft was drifting to the right as he noticed the runway edge lights moving towards him just before touchdown. The CVR recorded PM saying “slightly left of centerline sir”, followed by “slightly right of centerline sir”. These calls were initiated by PM from 8 ft RA until the aircraft had touched down.

Tyre marks on the runway indicated that the aircraft first touched down on the right wheel at a distance of approximately 540 m from Threshold Runway 13, and approximately 13 m to the right of the runway centerline marking. The FDR recorded the first touchdown on the right wheel at 2217:02 LT. The left wheels contacted the runway at 620 m (2217:03 LT). The nose wheel marking was only seen on the soft ground outside the runway surface. The FDR data showed the nose wheel contacted the ground 4 seconds after the main gear touchdown. Following the main gear touchdown, the pitch attitude was held approximately constant with a momentary increase in pitch attitude of up to 6o observed before the nose wheel contacted the ground.

The aircraft heading was offset by 3o – 4o to the right of the runway alignment. The FDR recorded heading drift from 130o to 133o. The runway heading was 129o.

Shortly after, at 2217:05 LT, PM called out “Go-around Captain, Go-around Captain”. The FDR data shows that the aircraft has started departing the runway surface by this time.

PF mentioned during the interview that he made an attempt to commence a go-around by pressing the TOGA switch. However, there was no response from the aircraft autothrottle system. The FDR data reviewed later showed that the TOGA mode was activated after the aircraft came to a stop.

The FDR data showed the ground spoilers extension upon touchdown and the autobrakes were subsequently engaged. The thrust reversers were not deployed until the aircraft was out of the runway.

PF mentioned that he did not realize the aircraft was heading off the runway until he felt the aircraft moving violently over the surface. He stated that the visibility was significantly reduced at this point.

Aircraft departed the runway surface at 780 m from the threshold and travelled approximately 480 m on the soft ground parallel to the runway approximately 20 m to the right of runway edge, until the nose gear collapsed. The aircraft then swung left towards the runway edge as the nose gear collapsed and came to a stop diagonally over the runway edge.

The flight and cabin crew members felt strong vibrationswaying to the left and right, and a loud “thud” before the aircraft came to a stop at approximately 1,260 m from Threshold Runway 13.

The Captain then shut down both engines in anticipation of a possible evacuation. Upon assessing the situation and discussing with the co-pilot, the Captain requested for the evacuation checklist. He commanded passenger evacuation at approximately 4 minutes after the aircraft came to a stop. The evacuation was carried out using the two forward slides (Door 1 Left and Door 1 Right). All passengers and crew members were safely evacuated in less than 90 seconds. No injuries were reported.

Inspection of the flight deck panels performed during the investigation found certain switches and flight controls settings were not in the correct positions as required by the evacuation checklist. These included Engine 1 and 2 Fire Switches, outflow valve position, flap handle and flap surfaces.

The cabin crew also reported that four Airport Fire and Rescue Services (AFRS) personnel climbed up the slide while the crew were still on board the aircraft. All passengers had evacuated by this time.

According to the cabin crew, the rain was getting heavier during the evacuation. Passengers were subsequently led by the cabin crew and AFRS personnel to the fire station. Head count was performed by the cabin crew at the fire station and all passengers were accounted for. No injuries were reported. Some of the passengers were transported to the terminal using vehicles provided by airport authorities, MAB engineering and other agencies within the airport. Most of the passengers walked to the airport terminal building.

The medical team who was waiting at the terminal performed medical check-up on all passengers. All passengers were found without any injuries from the incident or evacuation.

The aircraft was subsequently removed from the incident location on 09 April 2017, and parked at Bay 1A for further assessment and rectification.

The runway was closed for 34 hours following the incident to facilitate the removal of aircraft, repair and inspection of the runway.

Based on initial assessment, the aircraft sustained damages to the nose gear assembly and the lower fuselage aft of the nose gear, while areas around the flaps, engine cowling and fan bypass areas sustained minor damages.

The BKSU analysed:

Tyre track observed on the runway during the investigation revealed the following:

a. The right wheel contacted the runway first at a distance of 540 m from Threshold Runway 13. This was followed by left wheel touchdown at distance of 620 m from Threshold Runway 13. The FDR data confirmed 1 second lapse between the right and left wheel touchdown.

b. Lateral displacement of right wheel touchdown position from runway centerline was measured to be approximately 13 m to right of centerline. Given the wheel base of B737-800 of 5.7 m, the lateral displacement of the aircraft from normal touchdown position was approximately 10 m to the right.

c. The nose wheel touched the ground at approximately 4 seconds after the main wheel touchdown. The prolonged holding of the aircraft pitch following main gear touchdown prevented positive directional control that would have normally been achieved from nose gear contact on the runway. Based on the tyre tracks, the point of nose wheel touchdown was on the soft ground.

d. The elevated pitch attitude observed after the touchdown might have caused the pilot to have the reduced visibility of the runway situation. In addition, this maneuver had reduced the weight on main landing gear and thus decreased the braking and cornering effectiveness.

The BKSU analysed with respect to Threat and Error Management (TEM):

PF mentioned during his approach briefing and later reminded the PM of his plan to discontinue the approach, if he or the PM was not happy with the approach conditions. While it may be presumed that the PF was aware of the potential negative consequence resulting from the adverse weather and was prepared to execute a go-around should it becomes necessary, this was not clearly communicated to the PM. Obviously the phrase “not happy” does not provide a clear reference or guidelines on what to expect and what actions are to be taken in case of an undesirable event.

There were several occasions where the crew had the opportunity to review their decision to commence and continue their approach to land in the midst of heavy showers prevailing at the airport. The crew were initially provided with weather information which led them to believe that there was no considerable threat to the landing. The weather update was given while on descent passing 13,500 ft certainly presented them with the opportunity to review their decision based on the newly identified threats, i.e. RVR 1,200 m and in moderate to heavy rain. Instead, the crew appeared to be relying on one aspect of the weather criteria (visibility) in their decision to continue the approach. Another opportunity was presented following the missed approach procedures and holding.

Therefore, it is evident that the flight crew did not use adequate risk management strategy in identifying all the potential threats that were related to the approach and landing in heavy rain and thunderstorm. In particular, the potential threat of wind shear, microburst, turbulence, or sudden drop in visibility during the approach, or landing on the runway potentially contaminated by standing water were not considered in its entirety.

While the crew did discuss on the plan to divert to Kuching if the weather conditions in SBW does not show any improvement, the crew did not evaluate and review the appropriate options or recovery methods from each of the identified risks. This is not consistent with company policy pertaining to the adaptation of “TEM” concept prior to commencing approach. As a result, the possibility of a go-around in the event of failure to maintain runway centerline was not anticipated, particularly in the absence of the runway centerline lights. Neither was the possibility of losing visual reference due to heavy rain.

From the CVR, the flight deck environment appeared to be conducive for effective communication between the crew members. The commander did not show any signs of hostility or power gradient that could hamper open communication. This is essential to ensure any abnormalities to the operation of flight are effectively communicated.

With respect to crosswind landing technique the BKSU analysed:

During the final approach, the PM called out the wind information as displayed on the flight instrument to inform the PF of the wind component that was affecting the aircraft flight path. According to the CVR, the PM called “right 6 tail 2” at approximately 760 ft during the approach, and again at 600 ft, indicating that there was 6 kts of right crosswind and 2 kts of tailwind.

Consequently the PF applied the necessary drift angle and control wheel input to maintain the desired track towards the runway. This was evident from the constant oscillation of the control wheel throughout the approach. However, the wind was gradually dissipating below 500 ft, with less than 2 kts recorded below 130 ft AGL.

The PF continued to apply wing low crosswind technique during the flare maneuver before landing, on the basis of “perceived” right crosswind. Hence, below 25 ft RA, as the PF increased the aircraft pitch attitude to begin his flare maneuver, a right bank was progressively introduced. This was consistent with FDR data which showed that the aircraft was flying over the centerline until 30 ft in wings level attitude. Subsequently, the aircraft was slowly drifting to the right from a right bank that was introduced of up to 6o, before the right wheel came into contact with the runway.

...

The PF mentioned during the interview that he employed a mix of sideslip and decrab technique during the said landing, where he intended to touchdown on the upwind (right) wheel first, followed by the left wheel. Hence, to achieve this, the PF induced progressive right bank during the flare.

It was raining heavily at this point and according to the PF, the visibility had considerably reduced. There was no turbulence reported by the crew during the approach.

FDR readout showed minimal rudder input employed by the PF to correct for the drift in heading, which resulted from the pilot induced roll. The final heading recorded on touchdown was 133o , while the runway alignment was 129o.

The PF explained that visibility had reduced significantly prior to the touchdown, such that he had lost his positional awareness in relation to the runway centerline. Although he could reasonably see the runway edge lights, he could not be certain of the amount of rudder that was required to correct the drift, in the absence of runway centerline lights.

The PF further explained that he was not able to react in time to bring the aircraft back to the centerline before the aircraft left the runway surface. The flight crew only realized of the runway excursion after noticing the aircraft was moving violently over the surface. He then immediately applied full rudder force to the left to return to the runway but the aircraft was already on the soft ground. Based on the FDR data and CVR recording, the aircraft departed the runway surface approximately 3 seconds after the first touchdown.

In theory, travelling at a ground speed of 146 kts with a drift angle of about 4o over 5 seconds of flare maneuver (as per FDR), would result in the aircraft touching down approximately 10 m to the right of centerline. This was evident from the tyre tracks on the runway which were measured and found the aircraft touched down at approximately 10 m from the centerline.

Without having considerable crosswind component from the right and/or significant left rudder input, the pilot induced bank angle would cause the aircraft heading to veer to the right and drift away from the centerline.

Simulation exercise carried out using the same parameters extracted from the incident produced similar outcome. With an induced roll of up to 6o to the right, the aircraft would end up in drift angle of about 4o from the runway alignment. The aircraft would touchdown at almost 1/3 to the right of the runway centerline in about 4-5 seconds. If this heading was maintained without immediate rudder input to correct the offset, the aircraft would go off the runway.

With respect to visibility the BKSU analysed:

Both crew mentioned that they could see the entire runway edge lights before they encountered the heavy downpour below 100 ft RA. The visibility then dropped significantly and they were only able to see a stream of lights on both sides of the runway. The wipers were set at high speed and both inboard and outboard landing lights were switched ON.

The intensity of the runway edge lights and PAPI was set to 100 % by the ATC controller.

The reduction in visibility below 100 ft RA was contributed by the sudden increase in the intensity of the rain which would have made it difficult to identify visual references. This was further exacerbated by the reflection of the landing lights from the water droplets, especially in the darkness of the night. Despite the reduction in visibility, the PF mentioned that he could still see the runway edge lights sufficiently well and was confident that he could safely continue with the landing.

...

... the investigators collectively agreed that the flight crew had limited visual reference available while descending below 20 ft RA in heavy rain. Below this height, the PF would be looking entirely outside of the aircraft to execute flare and landing maneuvers. With only the runway edge lights available as visual reference, it would be difficult for the PF to accurately judge or detect deviation from the runway centerline, including if there was any inadvertent or deliberate bank
angle induced.

With respect to center line lighting the BKSU analysed:

Although runway centerline lights is not a requirement as per ICAO Annex 14 Aerodromes Standards for Category 1 Airport, the availability of the runway centerline lights is certainly beneficial when operating in marginal visibility in heavy rain, mist, fog or haze (all of which are common types of precipitation in this region), especially at night.

Airports that have higher exposure to inclement weather conditions, based on the meteorological and risk factor studies of the regional weather phenomenon, should be given highest consideration to the installation of centerline lighting.

Additionally, the “black hole effect” is also prominent when approaching Sibu Airport at night due to lack of other lightings surrounding the airport. Only lighting visible are the approach lights, runway edge lights and PAPI that are brightly lit.

Related NOTAM:
D0334/17 - AIRPORT CLSD DUE TO DISABLED ACFT ON RWY. 08 APR 14:30 2017 UNTIL 09 APR 02:00 2017 ESTIMATED. CREATED: 08 APR 16:09 2017

Metars:
WBGS 081600Z 12001KT 9999 -RA FEW005 SCT018 OVC150 24/23 Q1010=
WBGS 081500Z 10002KT 7000 RA FEW005 SCT018 OVC150 24/23 Q1011=
WBGS 081400Z 32002KT 6000 RA FEW005 SCT018 OVC150 25/24 Q1011=
WBGS 081300Z 22002KT 9999 -RA SCT018 BKN150 27/25 Q1009=
WBGS 081200Z 34003KT 9999 SCT018 BKN150 27/24 Q1008=
WBGS 081100Z 24002KT 9999 FEW015CB BKN150 27/25 Q1007=
WBGS 081000Z 16003KT 9999 FEW015CB SCT018 BKN300 28/24 Q1006=