G550 closes Stansted

[Tony Mabelis]

Well, I dont think a Gulfstream 550 is capable of closing Stanstead without human interference!
Tony

[Scottso]

Dearth of info corrected

Tubby, thanks for the weather info. Puts a lot of flesh on the bones. Quite a rapidly deteriorating situation indeed.


10 W. thanks for your significant input. Full credit to the engineers and the calibrators. Well done and reflects a thoroughly professional resolution to the problem.


It will remain interesting to find out why the aircraft was more than 50' low at the threshold. Over to the AAIB.

[Gulfstreamaviator]

550m is CAT 1 Min for many operators

The aircraft was I believe crewed by ex Saudi Military crews.


550m: is a fully legal, under many ops manuals for CAT 1 approach. It is the lowest RVR permitted, 600m being the regular Min RVR.


HUD: on the G550, it is possible to "fly" a HUD approach, and see the lights before a non HUD approach would see the lights.
It is normal procedure to input to the HUD, the runway elevation as well as the desired glideslope, this provides a secondary visual, as to the ideal approach path. The GS pointer being the primary guidance cue.


It is also possible that the crew had trained to CAT 2, in the simulator, and felt that they were qualified to make a CAT 2 approach, if their operations manual was so approved.
Training to CAT 2 is common place, in the recurrent world, even if non in operations manual as an approved procedure. It is trained to provide confidence in the systems, in he event of an emergency requirement.


Stansted ATC, would normally ask the crew, assuming the actual weather at the time of the approach was below 55m, that they were CAT 2 capable. Perhaps the word "capable" was misunderstood.


All approaches in UK below Cat 1 minima are reported by ATC to the relevant authority, who pass the approach conditions to the operators controlling authority.


I also assume that they were destined to land at Luton, and diverted to Stansted.


I also (trust) assume that fuel criticality was not a factor.


Glf

[tubby linton]

Here is the Luton weather.

METAR EGGW 100550Z 19004KT 140V230 0100 R26/0275 FG OVC001
03/02 Q1029=
SA 10/12/2013 05:20->
METAR EGGW 100520Z 20005KT 160V230 0100 R26/0325 FG OVC001
03/03 Q1029=
SA 10/12/2013 04:50->
METAR EGGW 100450Z AUTO 19006KT 150V240 0050 R26/0275 FG
OVC001 03/03 Q1029=
SA 10/12/2013 04:20->
METAR EGGW 100420Z 19006KT 150V220 0100 R26/0275 FG OVC001
03/03 Q1029=
SA 10/12/2013 03:50->
METAR EGGW 100350Z 20005KT 0100 R26/0325 FG OVC001 03/03
Q1029=
SA 10/12/2013 03:20->
METAR EGGW 100320Z 19005KT 160V230 0100 R26/0275 FG OVC001
04/03 Q1029=
SA 10/12/2013 02:50->
METAR EGGW 100250Z AUTO 20007KT 0050 R26/0250 FG OVC001 04/03
Q1029=
SA 10/12/2013 02:20->
METAR EGGW 100220Z 20006KT 0150 R26/0450 FG OVC001 04/04
Q1029=
SA 10/12/2013 01:50->
METAR EGGW 100150Z 19006KT 1300 R26/1400 BR SCT002 03/03
Q1029=
SA 10/12/2013 01:20->
METAR EGGW 100120Z 18006KT 1000 R26/P1500 BR SCT001 04/03
Q1029=

I wonder if the decision to commence an approach at EGSS was based on MEt vis and not the RVR.

[Scottso]

G550 Incident - AFDAS

Has anyone seen any further update on this incident??


Secondly, what is AFDAS(Approach Funnel Deviation System??. How does it work? How is it used? And where can I read about it?


Thank you.


Scottso




Refers to Post 14

[silverknapper]

Here You go.

Look a bit of a textbook CRM case for the next couple of years.

[Jet Jockey A4]

Why do these things still happen in 2013.

They should have performed a G/A at the first sign of an unstable approach.

This fooling around with the A/P and A/T so far into an approach is ridiculous.

They are very lucky to be alive.

[His dudeness]

Category:Commercial Air Transport - Fixed Wing

Now Jet Jockey, can´t say I´m impressed with your lame approach of this accident. Where is your broad brush you like use so often these days ?

Like:

Please don't even get me started on the idiots that fly some corporate jets.

Come on now, get your blood boiling, just think this was a corporate aircraft, non-comm, we then at least could read something more aggressive from you.

[BizJetJock]

They shouldn't have started the approach since it was below their minima (yes, I know tehnically you can go to 1000' but why waste everyone's time). And selecting flap 20 knots above the speed for that setting?
Unfortunately in the sim I have seen far too many crews like this, and the most worrying thing is that many do not see that i is a problem.

[CL300]

If the guys were from a FAA part 91 background, with the pressure they may have forgotten , that on this side of the pond you have approach bans...

All things being considered, why do you want to take the AP OFF while shooting at minimums ? ( besides Dassault planes on a HUD CAT 3B ?)

[mutt]

To me it appeared that the pressed the TOGA button rather than the A/T disconnect hence the reason he lost the EGPWS protection, they were very luck to get away with it!

[Jet Jockey A4]

@ His dudeness...

Now Jet Jockey, can´t say I´m impressed with your lame approach of this accident. Where is your broad brush you like use so often these days ?

I'm so sorry if I disappointed you. Next time I'll ramp it up a few notches. :-)

This said, whether this was a charter operation or a private operation, it was still like a corporate type operation so they probably fall into the same boatload.

[BizJetJock]

If the guys were from a FAA part 91 background, with the pressure they may have forgotten , that on this side of the pond you have approach bans...

I can just imagine the reception if anyone "forgot" that FAR's apply in the US. No excuse, I'm afraid, just sheer unprofessionalism.

[CL300]

indeed,

What can prevent a crew to go around ? Shooting the approach at or already below minimas on manual flying with an autopilot that tripped twice, a press on it is arrival in the last minutes.. Just try to grasp the "fundamentals" of their decision.

Mind you , if there was no mistakes there will be virtually no "accidents"...

[bbear737]

No mention of EVS

Strange the AAIB make no mention of EVS. I'm not aware of a 550 without it and the open defects are mentioned. Doesn't excuse continuing approach but it would have added to their SA and could have influenced some of the flawed decision making.

[noneya]

UPDATE to G550 Stansted Inciden

Business & Commercial Aviation
Below Minimums and Unstabilized
A Gulfstream crew lucks out after dragging the wheels thorough the approach lights at Stansted
Apr 1, 2015 Richard N. Aarons | Business & Commercial Aviation
EMAIL
TWEET

COMMENTS 0
This month we’ll take a look at a serious incident at London Stansted Airport in which a Gulfstream 550 hit the ILS aerials mounted on the Runway 22 approach light structures and then plopped down some 520 meters (1,706 ft.) short of the touchdown zone. Despite those dire circumstances, the incident resulted in no injures among the four crewmembers and three passengers. Still, the sequence of events demonstrates how quickly an unstabilized approach can put even an experienced crew into harm’s way.

The information that follows comes from the U.K. Air Accident Investigation Branch’s (AAIB) report on the Dec. 10, 2013, incident. Investigators concluded “the accident occurred as a result of the approach becoming destabilized and the pilots attempting to regain the correct glidepath at a late stage rather than performing a go-around. Descending below the glidepath at such a late stage caused the aircraft to collide with the ILS antennas. The RVR was below the 550-meter [1,804 ft.] minima required for the crew to commence a CAT I approach.”

The pilot in command (PIC) of the flight was a 53-year-old ATP with 8,685 hr. of flight time, of which 1,311 were in type. He had flown 70 hr. in the previous 90 days, but 0 hr. in the previous 26 days. The copilot was appropriately licensed and current.

The PIC was the pilot flying (PF) and the copilot was the pilot monitoring (PM).


The passengers and crew were unaware of the impact with the antennas and it was not until the aircraft was taxied to the parking area and the after-flight inspection was carried out that the damage was seen. Credit: AAIB

The Gulfstream 550 (HZ-A6) was conducting a charter flight from Riyadh, Saudi Arabia, to London. The flight crew reported for duty at 1730 and carried out normal flight-planning activities including a review of the weather. The TAF for Stansted covering the aircraft’s ETA gave a 40% probability of fog between 0300 and 1000, with a surface visibility of 500 meters (1,640 ft.) and cloud overcast at 100 ft. Weather at Manchester Airport, the selected alternate, was forecast to be 10 km (6.2 mi.) visibility with scattered clouds at 3,500 ft. for the same period.

The log showed one deferred maintenance item concerning the captain’s audio control panel. The mask/boom selector button was stuck in the mask position, but he could use the hand-held microphone so the defect was accepted.

The aircraft departed at 2001 and arrived in the London area about 0300 after an uneventful en route segment. The Stansted ATIS reported: Information Bravo, Runway 22 at time 0220, wind 160 deg. at 4 kt., Runway 22, Runway Visual Range (RVR) 250 meters (820 ft.) in fog with broken cloud at 100 ft., temperature 2C with a dew point of 1C and a QNH of 1030 hPa. The crew descended and set up for a CAT II approach.

The ATIS was updated at 0250 with Information Charlie, which was essentially the same but with the RVR increased to 300 meters (984 ft.) and the temperature and dew point both at 1C. The crew told investigators later that they carried out the normal and missed approach briefings for the ILS DME approach for Runway 22 at Stansted with Manchester as their alternate.

The aircraft was radar vectored for the ILS and cleared by ATC for the approach. Landing weight was 63,000 lb., with a Vref of 126 KIAS to which 5 kt. had to be added, giving an approach speed of 131 kt. The pilots armed the autopilot’s localizer and glideslope upper modes and engaged the autothrust system. The aircraft intercepted the localizer at 10.84 nm and the glideslope at 8.41 nm from the runway threshold. The aircraft successfully captured the localizer and descended with the glideslope.

Distance to Runway 22 (nm) Radio Height (feet agl) IAS (kt) UTC Event
4.99 1612 178 3:22:56 AP disengaged
4.1 1387 167 3:23:14 AP briefly engaged
3.58 1388 165 3:23:25 AT disengaged
3.07 1234 161 3:23:39 AP briefly engaged
2.43 1000 167 3:23:52 1000'
1.79 690 152 3:24:11 FD-vert: Glideslope to GoAround FD-lat: AppLOC to HdgHold
1.66 647 150 3:24:13 FD-lat: HdgHold to Lnav(FMS)
1.02 300 129 3:24:31 300'
0.9 179 130 3:24:40 APPROACHING MINIMUMS'
0.64 141 128 3:24:45 MINIMUMS'
0.26 20 116 3:24:56 20'
0.25 19 115 3:24:57 Collision with localiser
0.13 10 112 3:24:59 10'
0 3 108 3:25:04 Mainwheel touchdown
0 0 103 3:25:06 Nosewheel touchdown

Data Source: AAIB

The crew changed to the Tower radio frequency, established radio contact at 6 nm, and were issued the following landing clearance: “The surface wind 170 [at] 5 kt.; you’re cleared to land; RVR 325, 400 300.” (For a CAT I approach the RVR must not be less than 550 meters [1,640 ft.].)

The clearance was acknowledged by the copilot transmitting: “Cleared to land HZ-A6.” At 5 nm and a height of 1,625 ft., the flaps were selected to fully down. The speed was 181 KIAS, which is above the flap limiting speed of 170 KIAS, and an overspeed audio warning was generated. The flap travel stopped at 20 deg. Shortly thereafter, the autopilot was disengaged. The flaps were reselected to the fully down (landing) position at 4.3 nm.

At 4 nm the autopilot was re-engaged, but shortly thereafter, at a height of 1,212 ft., the autopilot was disengaged, and at 3.6 nm and a height of 1,388 ft. and 165 kt. the autothrust was disengaged.

The aircraft was significantly above the glideslope at this point, prompting ATC to advise the crew: “HZ-A6 indicating slightly high on the glidepath — confirm correcting,” to which the copilot responded: “Yes, we are correcting.” At 3 nm, the autopilot was again engaged but almost immediately disengaged and the captain hand flew the aircraft in a descent toward the glideslope.

The pilots told investigators they both had the approach and runway lights in sight throughout the approach, but did not see the PAPIs. At 1.7 nm, for reasons that could not be established, the autopilot mode, autothrust and vertical mode all changed to go-around, but the commander continued to fly the aircraft toward the runway.

Selecting the go-around mode changes the Primary Flight Display (PFD) from an ILS presentation to the horizontal and vertical go-around presentation. ILS guidance is no longer provided and the EGPWS Mode 5 “GLIDESLOPE” warning is no longer available. The captain continued the approach from that point by visual reference to the approach and runway lighting.

At 1.0 nm, the aircraft was at a height of 237 ft., 120 ft. below the glideslope, and it continued to descend to 30 ft. at 0.3 nm —130 ft. below the glideslope. At a height of 11 ft. and just under 0.2 nm from the runway threshold, the lower part of the fuselage and landing gear struck the Runway 22 ILS localizer monitor aerial and the Runway 04 localizer aerial array.

The aircraft continued in the flare at a height of 3 ft. at 0.1 nm from the threshold, before touching down at 108 kt. on the concrete surface of the runway undershoot area, 55 ft. below the glideslope and 109 ft. short of the runway threshold.

During the final approach there were no EGPWS warnings, apart from the normal radio altimeter height “callouts” and “APPROACHING MINIMUMS” alert, which were heard on the CVR. The passengers and crew were unaware of the impact with the antennas and it was not until the aircraft was taxied to the parking area and the after-flight inspection was carried out that the damage was seen.

It was apparent that the aircraft’s left landing gear had struck the monitor aerial and the localizer array. The left-wing underside panels ahead of the gear showed evidence of scratches and orange paint transfer. Most of the scratches were superficial, although one gash was 300 mm (1 ft.) in length and 5-mm (0.2 in.) deep. The skin had remained intact, with no fuel seepage. The leading edge was unmarked.

The left landing gear door showed significant impact damage on its leading edge; the appearance and dimensions of the damaged area suggested this had been made by one or more of the dipoles. The geometry of the main landing gear is such that it is probable the tires also made contact with the ILS equipment, although they bore no obvious marks. A hydraulic line also had been damaged.

After the on-site examination, the aircraft was cleared for the short flight to the aircraft manufacturer’s U.K. facility at Luton Airport, where temporary repairs were made. The aircraft was then flown to the manufacturer’s main base in Savannah, Georgia, for annual inspection and permanent repairs.

AAIB and ATC personnel determined that the aircraft had successively struck the Runway 22 ILS localizer monitor aerial and the Runway 4 localizer aerial array — structures that were located only 5 to 6 meters (16.4 to 19.7 ft.) apart.

The monitor aerial consisted of a tower approximately 5.5 meters (18 ft.) high, which was constructed from lightweight fiberglass material and supported a coaxial aerial cable. The 4.2-meter-high (13.8 ft.) ILS array comprised a series of 24 stanchions, each carrying 14 horizontally orientated dipoles made of aluminum alloy tubes covered with orange plastic sheathing. The stanchions were arranged equidistant from each other, with 12 on either side of the runway
centerline.

The aircraft had broken off the top of the monitor tower before striking the dipoles on stanchions 13 and 15, which placed the aircraft slightly right of the runway centerline. The latter impact had dislodged eight of the dipoles, which were scattered over the grass toward the runway.

Some tire marks were observed on an 85-meter (279 ft.) paved extension that preceded the “piano key” marks at the start of the runway. These indicated that touchdown had been made right landing gear first, left of the runway centerline and approximately 40 meters (131 ft.) from the start of the paved extension. This was approximately 520 meters (1,706 ft.) from the start of the touchdown zone.

ATC recorded the aircraft landing at 0325 UTC and at 0328 was advised that Runway 22 ILS had suffered a “technical fault.” A technician was dispatched to investigate and the ILS was downgraded from CAT III to CAT I; the airfield operations duty manager issued a NOTAM to that effect at 0345.

The AAIB Analysis

AAIB investigators issued the following analysis.

The aircraft struck the ILS antennas and touched down more than 500 meters (1,640 ft.) short of the touchdown zone as a result of descending below the glideslope. There were no EGPWS warnings, which initially posed questions as to the serviceability of the TAWS system. Although some messages in the fault history database (FHDB) could not be explained, it was concluded that the system had functioned correctly up to the point where the go-around mode was selected, which would have inhibited the glideslope alerts. This was reinforced by the genuine Mode 5 alert that was issued by the system during the subsequent flight from Stansted to Luton.

The crewmembers were properly licensed and qualified to conduct the flight. In their preflight briefing and planning they had identified the possibility of fog at Stansted and had nominated Manchester as a suitable alternate. The transit flight to Stansted was uneventful and the commander’s unserviceable radio selector panel was not relevant to the accident.

The RVRs in the ATIS reports and those passed to the crew with their landing clearance were below the CAT I minimum RVR of 550 meters (1,804 ft.). The approach should not have commenced as the U.K. ANO requirements did not allow the crew to descend below 1,000 ft. above airport level in these conditions and the company general operating manual procedures did not permit an approach to be made in such conditions.

During the descent toward Stansted Airport the crew briefed the arrival and missed approach should one be necessary. The autopilot captured the localizer and the glideslope for Runway 22, but the selection of full flap, above the flap limiting speed, appears to be the start of a chain of events that destabilized the approach.

The disengagement of the autopilot and autothrust led to the aircraft leveling and rising above the glidepath, which was notified to the crew by ATC. While correcting the flight path to regain the glideslope, two attempts were made to re-engage the autopilot; however, these were unsuccessful so the captain continued to hand fly the aircraft. The reason for the unsuccessful re-engagements was not determined.

The aircraft flew above the glideslope where it remained (mostly between one and two dots deviation) until about 300 ft. AGL and 1 nm from the runway, just as the airspeed slowed to 131 kt. At a height of 691 ft., the go-around mode was selected, but the captain continued visually toward the runway, passing through the glidepath at about 300 ft. at 1.0 nm. The reason for the change to go-around mode could not be determined, but it was significant in that glideslope deviation alerts would no longer have been provided. The aircraft continued to decelerate and then descended below the glideslope, reaching four dots deviation as the aircraft collided with the localizer antenna at 19 ft. AGL.

The glidepath deviations were outside the stabilized approach criteria and when combined with the reducing airspeed below VAPP, a go-around should have been flown. (See the accompanying table for timing of configuration changes.)

Although the approach and runway lights were visible to the pilots, the PAPIs were not and therefore the approach path angle was a matter of judgment. Apart from the normal advisory callouts, the EGPWS did not alert the crew to the deteriorating situation as the aircraft began to undershoot the runway because the go-around mode had been selected.

The radio altimeter height callouts, combined with the visual perspective of the runway lights, provided the cues to flare the aircraft.

The fact that the pilots could see the runway and approach lighting caused them to believe that, as long as they remained visual with these landing references, they would comply with their company procedures and thus could continue their approach.

Final Thoughts

So, the unstabilized approach led to an undershoot that could have been disastrous were it not for a few feet of vertical luck.

Stabilized approaches are always important for a reasonably good landing, but they are absolutely essential for a safe approach at minimums. If, for some reason, you are forced into a below minimums approach, then conducting a stabilized approach is vital in the true sense of the word.

Boeing studies determined two decades ago that pilots cannot depend on their eyes or the seats of their pants when landing swept-wing jets in low visibility, night approaches. Human eyes just aren’t equipped to do it. We have to depend on the systems and the SOPs (to say nothing of the regs) to tell us that it is time to call it off the moment the approach destabilizes or the airplane becomes misconfigured.

‘Nuff said. B&CA