SQ006 Investigation Report Commentary




Singapore Airways B747-400 Wrong Runway Take-off in Taipei


You will see that Taiwan does not entirely escape blame in the ASC report.

The reports by ASC and MOT are very thorough and leave no stones unturned in the above areas; lots of lessons noted and actions have been taken by both CKS and SIA to rectify the shortcomings .

Both approaches in investigations are correct .

But what alarmed me was
the formatting of the Final Analysis of ASC which classified the Causal Factors as:

"(1) “findings related to probable causes” which identify elements that have been shown to operate or almost certainly have operated in
the accident; ( please note the operative word: operated.)

In short
the investigation of the causal factors relating to the accident, will be entirely focused on the Pilot and the weather, and deficiencies of airport system and other contributing factors were relegated to Risks.

As a result,
the concluding "Probable causes" will, for all intents and purposes, always be on the weather and the pilot even before the ASC started its analysis of the Probable causes "

Even though Dr Yong Kay (
the chairman) has qualified subsequently that Risk Factors play an equal and critical role as causal factors in the accident, whether we like it or not, it is only natural for both sides to adopt a "self-protection" mode in this investigation.

Not sure how long 05R had been inactive as a runway. I had heard rumours it was as much as 12 years! Is there any record of when the last time 05R was actually used for takeoffs and/or landings? How long have the Taiwanese maintained this hazard to operations, a taxiway marked as an active runway? Might as well put VASI's on Mt. Fuji.

1. The Pilots failed to determine that
the runway condition should have been classified as ‘contaminated’ due to the inevitable buildup of standing water under cyclone conditions. This determination would have downgraded the cross-wind limit for takeoff from 45 knots (wet runway) to 15 knots (contaminated runway) and therefore prevented the scheduled departure.

2. The Pilots' low-visibility taxiing performance was deficient during
the final stage of taxi, and their navigation procedure was inadequate and did not utilize runway charts, signage, markings or cockpit instruments to guide them to the correct runway. The attention of the co-pilot Latiff and the relief pilot Ng was ‘inside’ the cockpit focused on the pre-takeoff checklist and calculating the cross-wind component. Only Captain Foong’s attention was ‘outside’ the cockpit focused on taxiing slowly in the slippery conditions and following the green taxiway lights to the runway.

3. The reason Captain Foong made a premature turn onto runway 05R was because:

i) he had
the false impression that ATC could see the aircraft. He was under that impression because ATC issued his takeoff clearance just as he reached the end of the taxiway and was beginning his turn onto the runway. Co-pilot Latiff also confirmed that the timing of the takeoff clearance gave him the impression that everything was in order.

The captain did not use the runway chart or the compass to guide him to the correct runway. Instead he stated that he “followed the green taxiway lights” despite the fact that this technique is not as per Heathrow and could/should not be used at CKS (where visual navigation is required).

iii) He also failed to notice signs, markings and lights which indicated he was turning onto
the wrong runway. These included the green taxiway lights to 05L (which were more widely spaced than normal), a clearly illuminated 05R sign as he turned into the runway, and the 05R marking painted on the runway in front of him as he lined up for take-off. But the first link in the chain was that the FIRST taxiway green centre-line light leading across 05R threshold to the 05L runway was unserviceable. And the second light (on from the u/s one) was DIM. The first serviceable centre-line light was distantly 116m away (further on and across the far side of the 05R threshold). Yellow taxi-way centre-line markings did not continue across the 05R threshold (between the 05R piano-keys) – as required by ICAO. Only 4/16 centre-line (lead-on to 05L) lights, as required by ICAO, were provided (and two of those were u/s). So what Cpt Foong had instead, in his clearly-lit foreground, was a very seductive set of 16 closely spaced Centre-line green lights leading him onto 05R.


4. As they held for takeoff, all 3 pilots thought they saw a ‘normal runway scene’ and failed to recognize crucial indicators that they were on the wrong runway, including:

i) that the center runway lights were green and not white (a printout record from calibration and maintenance equipment showed that the center runway lights were on 4 minutes before the crash)  

ii) that it did not have touch-down lights or high-intensity runway edge lights (lack of evidence from security and passenger video, together with eyewitness testimony from the pilot of a plane on the taxiway 110m from the 05R threshold indicated that the runway edge lights were almost certainly off).

iii) that the runway was 15m narrower than it should have been

iv) that the PVD instrument had not unshuttered.

5. Several factors were speculated to have led to these navigational deficiencies and observational failures including that:

the pilots were accustomed to using runway 06 and had not used 05L for 2-3 years and were therefore unfamiliar with it.

ii) Although having read a NOTAM concerning
the construction on 05R, the pilots had not read the SIA INTAM that contained important information on the status of 05R’s center and edge lights

iii) Captain Foong had not been officially trained in low-visibility taxiing

iv) The pilots all had an incomplete understanding of why
the PVD did not unshutter and incomplete training on the operational context of its usage.

v) The green taxiway lighting that led to 05L was abnormally spaced and instead of being 7.5m apart, each light was 23m apart

vi)  All 3 pilots were fixated on taking-off quickly before
the weather conditions deteriorated further and prevented takeoff (although there was no evidence of ‘organizational pressure’ on the pilots to depart on schedule that night).

6. The CKS airport authority was derelict in
their maintenance and safety responsibilities. By not placing mandatory construction warnings at the entrance of 05R, they failed to address the risk that aircraft might inadvertently attempt to take off from a partially closed runway. They also failed to adequately maintain taxiway lights and marking to accepted international standards.

Maybe I am missing something, but this seems a pretty unbiased statement of facts. If you look at
the NTSB comments in the appendix, they seem to think so too. Don’t be fooled by the spin doctors.


i.e. I think the wrong turn into 05R is not the BIG issue here. One can make a wrong turn (and I think this not as uncommon as one would have thought?) without the same disastrous outcome. i.e.:

a.  realize
the mistake and turn out again, or

b. don’t realize
the mistake but still take off from the wrong but operational runway ..... (safely!?).

The really big issue is being able to attempt take-off from a CLOSED (and wholly obstructed) runway. Some would insist that it was a PARTIALLY closed runway and
therefore not technically 'CLOSED'. How is one supposed to make that distinction if the difference is only 'apparent' as you are passing 'V1'? Some would counter that the centre lights were not white and there were no touch-down lights and that the edge lights "were almost certainly off" that this was not a runway but a taxiway. Some will say that the PVD "had not unshuttered" hence this was a wrong runway. Then how come "all 3 pilots thought they saw a ‘normal runway scene’ " and thought they were on the correct runway? Look at this explanation here.


"Co-pilot Latiff also confirmed that the timing of the takeoff clearance gave him the impression that everything was in order."

However, if you look at
the report, the elapsed time between clearance and start of the roll is relatively considerable… in the order of 80 seconds. Partly explained by the 5 kts taxi speed, I know.... but if their "ready" call was premature (clearly debatable), then ATC’s natural response to offer clearance (no traffic-and they can’t see them anyway) whilst still on taxiway NP might have induced the pilot to more readily accept the "lead-on" picture he saw at N1/05R intersection, given the time elapsed between the two. The longer the time-elapse is, the more willing he becomes to accept the notion that he is indeed on the correct runway.
the takeoff clearance had only just been issued (later, during the turn into N1), the pressure to accept those compelling green lights might have been less.

Turning onto the wrong runway was the cause of the accident and the main issue here!

But equally, because humans are naturally fallible and indulge in secretive wishful thinking,
the why and how that led the pilots into turning onto the wrong runway is what needs to be scrutinized.

From where I sit,
the following issues are the main areas/factors:

The inclement wea

The night environment,

Signage, lightings, markings of
the taxiway and failure to obstruct entry into the 05R Runway

Air Traffic control : Low Visibility Procedures (LVP) taxy monitoring and issue of takeoff clearance.

Take off procedures of SIA: for confirmation of correct runway.

The TW report avoids unveiling the "golden nugget" and instead opts for a mish-mash of comments and diatribe on several elements that make up the whole.  Yes, an accident is usually the result of a chain of events but the "probable cause" theory is still the best way to go, in my opinion. What was the last chain in the link that made the accident inevitable? Turning on to the wrong runway is about as clear an error as you can get. However this obviously wasn’t the last link in the chain, although if the NTSB had been at the helm, you can bet your bottom dollar that this would have been their conclusion. I see that their recommendations centre almost entirely on developing cockpit moving-map nav displays for ground operations.




     The Psychology of What they did - AND WHY


     How Dreadle Might have Stopped SQ006


     Large Scale Map of Taipei Runway Arrangement (large format jpeg [right-click/save as - to view off line])


     PowerPoint File of What Pilot Saw as He approached Runway Entry Points [save and view off-line]


Click to download free PowerPoint Viewer


Weather, Airport Status, Crew Errors Combined To Doom Singapore Flight, Report Says

By Sean Broderick/Aviation Daily

26-Apr-2002 1:40 PM U.S. EDT

A combination of horrendous weather, poor airport markings, and a lack of adequate precautions by both the pilots and air traffic controllers set the stage for the crash of a Singapore Airlines jumbo jet at Taipei's Chiang Kai-shek International Airport (CKS) 18 months ago, a report released today said.

The report, issued by Taiwan's Aviation Safety Council (ASC), does not include a probable cause. It lists three sets of findings, including eight "Elements Related To Probable Causes" that either contributed or "almost certainly" contributed to the accident.

The crew of Singapore Flight 006, Boeing 747-400 9V-SPK (pictured), attempted to take off in strong winds and heavy rain brought in by the approaching Typhoon Xangsane on Oct. 31, 2000. Flight 006's crew was cleared to taxi to CKS's Runway 05 Left, but instead turned off early and lined up on Runway 05 Right, which was closed for construction.

"The flight crew did not review the taxi route in a manner sufficient to ensure they all understood that the route to Runway 05L included the need" to pass Runway 05R, the report said.

The crew had CKS taxi charts, but "none of the flight crewmembers verified the taxi route," which included a 90-degree turn from one taxiway to a second taxiway, followed by another 90-degree turn onto Runway 05L.

Instead, the crew made a 180-degree turn from the first taxiway directly onto Runway 05R. "None of the flight crew members confirmed orally which runway they had entered," the report said. "The flight crew lost situational awareness and commenced takeoff from the wrong runway."

The 747 struck concrete barriers and construction equipment about 3,300 feet down the runway. The plane broke apart and caught fire. Eighty-three of the 179 people on board were killed.

The ASC report cited several problems at CKS as "findings related to risk." Although Runway 05R was closed, its ends were being used as taxiways. The closed runway had green centerline lights that designated it as a taxiway, but investigators said there was a "lack of adequate warnings" at the runway's entrance that would have provided "a potential last defense" to a crew mistaking the strip as an active runway.

Investigators also found several problems with the lighting on and around Runway 05R, including inoperative bulbs and an inadequate number of lights. These and other signage and marking problems likely contributed to the crew's loss of awareness, the report said.

The local controller "did not issue progressive taxi/ground movement instructions and did not use the low-visibility taxi phraseology to inform the flight crew to slow down during taxi," the report says. Investigators point out that the crew could have requested the step-by-step taxi instructions, but did not.

CKS did not have Airport Surface Detection Equipment installed. It is not clear whether the aircraft movement-tracking system would have helped prevent the accident, because heavy precipitation can hinder the radar presentation, investigators said.

ASC made 60 recommendations to nine organizations as part of the report, including 14 to Taiwan's Civil Aeronautics Administration, 10 to the airline, and eight to the International Civil Aviation Organization (ICAO). Many of them focused on improving closed-runway procedures and adding ground radar. Several recommendations to Boeing, the U.S. FAA and Europe's JAA touched on evacuation standards and equipment design.

Singapore Airlines said it "generally" accepted the report's findings, but took exception to the emphasis on its crew's role. "It appears that the ASC report does not give due weight to the deficiencies found at CKS Airport," the airline said in a statement. "We believe these deficiencies misled the pilots into taking off from the wrong runway."

The airline also said the ATC procedures leading up to the crash deserve more scrutiny. "The air traffic controllers cleared the flight for takeoff at the critical moment that it was taxiing towards Runway 05R," the airline said. "This reinforced the crew's belief that they were entering the correct runway. In low-visibility conditions, air traffic controllers at CKS Airport are required to determine the position of the aircraft before issuing takeoff clearances. They did not follow this procedure."

The airline said it has changed several procedures and beefed up training based on lessons learned from the investigation as well as a parallel one done by Singapore's Ministry of Transport. "We will study the recommendations in the ASC and the MOT reports in more detail, and we will implement any recommendations that would further enhance safety," the airline said.

See Also:

ASC Singapore 006 Final Report (.pdf, 508 pages)
Singapore 006 Coverage On AviationNow.com


Poor Markings, Missed Clues Set Up Singapore Crash

By Sean Broderick

23-Feb-2001 3:58 PM U.S. EST

Pilots of the Singapore Airlines jumbo jet that crashed last October missed clues that showed they were taking off on a closed runway, and significant airport marking deficiencies helped fool the pilots into thinking they were on the correct strip, factual reports on the crash investigation released Friday indicate.

Flight 006, flown by the Boeing 747-400 carrying registration No. 9V-SPK (pictured), was cleared to depart on Chiang Kai-Shek International Airport's (CKS) Runway 05L last Oct. 31. Instead, the pilots attempted to take off on a parallel runway, 05R, which was closed to operations at the time due to construction in the center of the runway but was being used as a taxiway at each end. The plane was destroyed after hitting construction equipment sitting on 05R. Eighty-three of the 179 people on board were killed.

Weather at the time of the crash, which happened at 11:17 p.m. local time Oct. 31, was rainy and windy due to a typhoon bearing down on Taipei. Visibility was about 500 meters.

Because of the poor weather and nighttime conditions, the pilot and co-pilot elected to switch on the "para-visual display" (PVD), facts gathered by investigators and released by Taiwan's Aviation Safety Council (ASC) show. The PVD, a mechanical instrument mounted on a panel in front of each pilot position that helps the pilots line up and stay on a given runway's centerline, works with the plane's instrumentation to monitor a runway's Instrument Landing System (ILS) signal.

The PVD resembles a barber pole sitting on its side, with black stripes on a white background. It is not mandatory equipment, and carriers that use it only require it to be activated when visibility is much worse - 50 meters or below, in most cases - than the visibility the Flight 006 crew was faced with.

When the aircraft gets in range of the ILS runway signal that the plane is tuned to, a small shutter on the PVD opens, revealing the black-and-white pattern. The stripes remain stationary so long as the plane is on the runway centerline. When it moves left or right, the stripes move in the direction of the runway centerline, helping guide the pilots back to the middle of the strip.

Both the pilot and first officer of Flight 006 switched on their PVDs at the gate, investigators found. When the aircraft taxied into position at what the crewmembers thought was the end of 05L, all three pilots - including a relief pilot sitting in the cockpit - noticed the PVD had not activated. But since visibility was well above the level that requires PVD usage and they could all see centerline marking lights clearly, the pilots decided to proceed.

"The PVD hasn't lined up," the co-pilot said as the plane turned onto 05R, according to the cockpit voice recorder transcript.

"Never mind. We can see the runway," the captain responded. "Not so bad."

The visual takeoff may have caused the pilots to miss two other clues on their instruments that could have indicated a problem. When a 747's ILS is tuned for a specific takeoff runway, two indicators appear on the plane's primary flight display (PFD). A pink diamond shows the aircraft's position relative to the runway's ILS localizer, and a green trapezoid shows the runway, which should be centered and just below the PFD's horizon when the plane is aligned properly. When the plane is not aligned with the runway centerline, both indicators are well off to the display's side.

The centerline lights could have served as another clue to the pilots. The captain told investigators that he "followed the curved centerline lights" onto 05R, an ASC report said. "He commented that he was attracted to the bright centerline lights leading onto the runway."

The centerline lights that run all the way down 05R are green, designating it as a taxiway. Centerline lights on runways are white at the beginning and later change to red near the end. While a similar set of green taxiway lights leads from N1 onto 05L, the lights running down the middle of 05L are white.

Both 05L and 05R have bi-directional runway edge lights that appear white, yellow, or red, ASC said. The two sets of lights are identical. The CKS ground controller working on the night of the accident told investigators that the 05L edge lights were on, but the 05R edge lights were not. Soon after the accident, the captain told investigators that he was "80% sure" he saw edge lights along 05R, but in follow-up interviews, he was "less sure," ASC said.

While the pilots may have missed some clues regarding their wrong-runway mistake, they were almost surely hampered by airport surface marking deficiencies.

As they followed taxiway lights down NP and turned right onto N1, they did not see any centerline lights straight in front of them that would have led them to 05L. They did, however, clearly see the curving set of taxiway centerline lights, spaced about 7.5 meters (25 feet) apart, leading to 05R from N1's south end.

When investigators surveyed CKS four days after the crash, they found that the two taxiway lights designed to lead aircraft further down N1 past the inactive 5R to the active 5L runway were not working perfectly. One was not illuminated at all, and the other was "less intense than the other lights." The lights, spaced about 25m apart, run straight down N1's centerline and meet up with another set of curved, green taxiway lights that connect 05R's centerline with N1's north end.

As Flight 006's captain taxied down N1 and approached 05R, he was "focused on the image of the runway to his right, and he did not notice any further green lights ahead and along the extension of N1," he told investigators.

Investigators found several other lighting and marking problems. Some of the runway edge lights on both 05L and 05R were either broken or "aligned away from the direction of the runway length," ASC said.

Also, there was nothing over the 05R threshold markings that indicated the runway was closed.

Runway 05R had been closed since mid-September for needed pavement repairs. The plan was to convert it into a full-time taxiway on Nov. 1, but the timeline was pushed back before the Flight 006 crash. Before being closed, it was used for visual departures only.

The Flight 006 captain told investigators he was aware of 05R's status. He had used the runway in the past; his last departure on it was "two or three years" ago, ASC said. The captain's last flight to Taipei before Oct. 31 was sometime in early to mid-September, the agency's Human Factors report said.

Singapore Airlines most often used Runway 06, the parallel runway south of CKS's terminal, because it is "closer to the parking bays used by the company," the Flight 006 captain told ASC. But 06 is a Category I ILS strip, and the weather on Oct. 31 persuaded the pilot to request 05L, a Category II runway, because it is "longer and would therefore afford better margins for the prevailing wet runway conditions."

See Also:

Flight 006 Factual Reports (click on "New Releases")


to Hot Off the Press