View Full Version : B737 gross over-run accident. Time overdue to fix the simulator syllabus.

24th Dec 2018, 07:29
Report on the overrun of this B733. Seems like an exact copy of the Garuda B734 fatal overrun in Jogjakarta. A very interesting read. These sort of accidents will continue unless there is a radical change to the simulator training syllabus. Rather than just learn stabilization criteria by rote, there is a strong case to include a physical demonstration by the instructor of the dangers of ignoring stabilized approach criteria.

Set up a deliberate unstable approach (for example high sink rate, low or high airspeed from five miles out leading to hard landing or excessive float and a gross over-run. Easily demonstrated by a competent instructor . A picture is worth a thousand words. Then if it happens on line, the pilot will recognise whats happening from previous experience in the simulator and take appropriate corrective action
Incident: Solitaire B733 at Aqaba on Sep 17th 2017, runway excursion on landing (http://avherald.com/h?article=4ae7b1c1&opt=0)

24th Dec 2018, 10:05
The appropriate action in the event of an unstablised approach, at some point - according to slightly differing SOPs is to initiate a Go Around. Action that was not taken in the two examples cited.
That is what needs drilling in to skulls in the sim.

24th Dec 2018, 12:46
The Chinese found an effective way in managing that.
If you continue from an unstable approach, you're fired.
Never seen anyone ever getting even close there.

24th Dec 2018, 14:00
If the student manages to recover the unstable approach into a nice landing
That is not the purpose of the demonstration. It is the simulator instructor that demonstrates the dangers of the unstabilised approach. Not the candidate to see if he can make it to a smooth landing.

Mach E Avelli
25th Dec 2018, 02:13
No matter how skilled the instructor, I think it is a bad idea to ever demonstrate something unsafe. We discuss it at great lengths during training. We may even let the trainee do it once, but follow up with appropriate debriefing as to why it was a bad choice. Never allow it twice - at the point where it is clearly unstable the instructor should call "go around" until such time as crews recognise it and act on it instinctively.
A valuable lesson if they land long is to fail the brakes, not having briefed this as part of that training session.
On any check, landing from an unstabilised situation or beyond the accepted touch down zone should be as much 'fail' items as an altitude bust or sustained out of tolerance, regardless of whether the candidate pulls it off without crashing.
In my pre-test brief I list unstabilised and long landings as failure items, so there can be no excuses.
I think some instructor/examiners tend to let poor landing decisions go rather than fill in the failure paperwork.

25th Dec 2018, 03:50
For those readers that lack the time and inclination to read the link to the full report at Post 1 of the Solitaire 737 accident, the following extracts from the summary should only take a few seconds to read.

"At 35 miles from AQB VOR, ATC contacted the crew and asked them if they were able to maintain their speed to continue landing on runway 19. The tailwind was 16 knots. The crew accepted and continued to a straight in approach.
During final and after dropping the landing gear, the aircraft was not configured to the correct landing configuration, the flaps were set to Flaps One at height of 650 feet AGL, multiple GPWS warnings were triggered but disregarded by the crew (First Officer was PF throughout), the high speed was not corrected by the crew efficiently and the aircraft continued to the landing runway.

As the aircraft descended below 500 feet RA the computed airspeed was 218 knots (VREF+85) and reducing. The aircraft was configured at flaps one with speed brakes extended.and calculated sink rate approximately 1200fpm.
The aircraft passed the threshold at 115 ft RA at flaps five and continued along the runway to the point when the flaps were selected to Flaps 30 at 90 feet RA. The captain took over and managed to put the aircraft on the runway at 7400 feet beyond the runway threshold. The airplane came to a stop 10,600 feet beyond the landing threshold (600 feet inside the soft area)
The roles in the cockpit were not followed correctly. The copilot who was flying the aircraft was not assertive against the captains intentions and communication calls even after the GPWS warnings that gave them a clear indication of the unstabilized situation they were encountering. The situation had clearly indicated ineffective CRM skills and resulted in the incorrect decision of the eventful landing."

That debacle says much about the airline culture.. Thankfully there were no casualties. Not so for the unfortunates in the eerily similar final approach circumstances of the Garuda 737 over-run at Jogjakarta several years earlier.

25th Dec 2018, 06:03
I'm likely to get yelled down here, by here goes anyway.

Imo, I think that crews trained in the last decade or so seem to not have an understanding what an unstable approach is, other than a written criteria from their SOPs. By that I mean there seems to be no "feel" other than be able to note an exceedance without understanding the likely implications.

The Garuda B737 flight, if my memory is correct, passed over the threshold F5 and 190 kts. How the hell could ANYONE not understand that unless you were landing on a 3500+m you are going off the end. Even if you were landing on say 4000m you are so far behind the 8 ball in situational awareness, you should be getting out and doing it again.

"The tailwind was 16 knots. The crew accepted and continued to a straight in approach.
During final and after dropping the landing gear, the aircraft was not configured to the correct landing configuration, the flaps were set to Flaps One at height of 650 feet AGL, multiple GPWS warnings were triggered but disregarded by the crew (First Officer was PF throughout), the high speed was not corrected by the crew efficiently and the aircraft continued to the landing runway."

Again, how could any crew not see this config at this point as having no option BUT to GA? Forget stabilised app criteria, forget the GPWS barking in your ear, you are not even in the ball park to land!

In my early career, before unstable approaches started to be tackled seriously I saw a number that landed safely, usually a higher speed, but still configured correctly and over the threshold at the right speed, or a higher sink below 1000' but never > 1500fpm and corrected by 300' and never on limiting runways (I was an FO). I'm not calling for any return, just a comment that crews seems to have lost an understanding of what a dangerous approach even looks like. I don't wait 'til 500' to give away an unstable approach, you should be able to judge the likelihood of being stable and I usually have given it away by 1000'. A GA from 500' just unnecessarily raises pax anxiousness for no reason.

My Co SA height is 500' vmc, mine is 1000'

I can't remember an unstable IMC approach.

25th Dec 2018, 07:43
Anybody who: Flies an approach at Vref +65kts with a 16kt tail wind, (tailwind limit for a 737 is 15kts?); Continues the approach while multiple GPWS alerts are sounding; Flies over the threshold with only flap 5 - AND CONTINUES THE LANDING - (in a normal fully serviceable aircraft), is clearly not a Captain, or even a pilot. I mean, this beggars belief !!!!!!!!!!!

I think serious questions need to be asked about the Captain’s log book and licence, and the training organisation he trained under. I would not be at all surprised if it was faked, or there were fraudulent elements.

One occasionally hears about practicing medical Doctors who are exposed as fakes, with made-up qualifications and invented experience, sometimes after having been treating patients for years, and the same thing obviously happens in aviation. In the company I fly for, there was a chap who tried to join as a direct entry Captain, but luckily the trainers noted how crap he was in the SIM and unceremoniously showed him the door.

Questions need to be asked about the relevant CAA in this case and the checking procedures they employ. I smell corruption somewhere.

Mach E Avelli
25th Dec 2018, 09:39
Corruption indeed, though not always of the paper bag slipped under the desk variety. Certain countries have very poor oversight from the licencing authorities. Some places have a situation where the examiner has a vested interest in the airline - maybe via a shareholding, or a favorite son has a job there. One outfit I know, the civil authority examiner also moonlights for his assigned airline. Sometimes the examiner is also the Chief Pilot of the airline and possibly has a financial interest beyond his pay packet. Father and son pairings on sim checks occur. Nepotism is big in some cultures.
I have occasionally run the simulator panel and watched really poor performances which are still awarded a pass. Then the crew and examiner rush off to the nearest casino or bordello to make the most of their overseas allowances.
Whether this should be reported to ICAO and whether ICAO would take any action on the word of a simulator panel operator is debatable. If ICAO is anything like the U.N. it is toothless.

Big Pistons Forever
25th Dec 2018, 17:44
You can't train company culture. Vref + 65 and a tailwind and yet the crew continued and tried to land. That sort of Shyte just does not happen at airlines with decent flight ops management.

That being said a well crafted LOFT scenario could be constructed so that the crew gets suckered into a hot high, flight path just on the edge of the limit with the aim of producing a go around decision.rather than trying to force a result. I have not seen such a thing myself but others may have beat me to it.

26th Dec 2018, 04:51
You can't train company culture. Vref + 65 and a tailwind and yet the crew continued and tried to land.

It is rare to see any FCOM or FCTM recommend what actions are available to a first officer to force a go-around if it becomes obvious a captain is intent on landing regardless of stabilisation criteria. (the “real men don’t go around syndrome” so common in some ethnic cultures). If all “official company SOP” warnings are ignored, it is imperative the co-pilot take immediate decisive action to prevent the inevitable accident. In most cases, airlines and regulators tend to leave those actions to the judgement of the co-pilot without effective guidance on what actions are available.

Mouthing into the CVR “Captain you must go around” is usually offered as sage advice, yet judging from accident reports, totally ineffective in some cultures. That said, it is in these cultures where pragmatic advice to a first officer is most needed. In these type of situations, my advice to candidates during type rating training is to simply say “Captain - Go-Around – landing gear coming up” and select gear up without further ado. No captain will deliberately land wheels up to make a point. Sort out the tea and bikkies in front of the chief pilot later. At least you will be alive to do so..

26th Dec 2018, 06:54
a well crafted LOFT scenario could be constructed so that the crew gets suckered into a hot high, flight path just on the edge of the limit with the aim of producing a go around decision.rather than trying to force a result.
in a well respected Airline recently we did just about the opposite in a recent LOFT. Positioned at 10;000ft 20nm from touch down and told to ‘manage’ the aircraft to either a landing or a go around. I suggested it was negative training and that I would simply ask for track miles but was told to ‘have a go anyway’. I have been flying long enough to know how to do it but it does show how some ego’s/ weird ideas make their way into examining roles in otherwise good outfits.

26th Dec 2018, 09:13
Sheppy #1, et al, I share the frustration of this type of accident, but it is a mistaken belief that ’the pilot will recognise whats happening from previous experience in the simulator and take appropriate corrective action’.

Deep breath. We have to acknowledge, collectively and individually, that these events represent an extreme aspect of ‘normal’ human behaviour, a limitation. The example quoted has mind-catching salience, but is contextually similar to other events or more general scenarios.

We judge behaviour on outcome, whereas there are many ‘routine’ operations which illustrate the same aspects of thought and action, but without notice.
Unstable approaches do not ‘cause’ accidents; there might be an association, but so too with many safe landings - the outcome depending on (unknown) circumstance.
How many of our landings, with a safe outcome, meet the required safety margins. How often do we trade safety margin to roll to-the-end, using less braking; or we accept that a slightly long or fast landing would be acceptable for a limiting landing. Do we know what is limiting?
What we might not know is when we are actually operating safely, not just based on outcome, but with factual knowledge.
A skill in landing is being able to identify those aspects which are not known, and managing the uncertainty in these operations - where will the aircraft stop. Do we teach that in simulators?

26th Dec 2018, 10:32
Unstable approaches do not ‘cause’ accidents
With all due respect PEI 3721, your waffle has lost this contributor. It might be worth your while reading an extract from the link below.

Unstable Approaches: Risk Mitigation Policies, Procedures & Best Practices.Many studies have been conducted with respect to unstable approaches and approach-and
landing accidents using different data sources, definitions and analytical logic. However, the findings of these studies consistently conclude that unstable approaches have been and continue to be a significant factor in commercial aircraft accidents. Without improvements in the rate of stable approaches flown and stable approach policy compliance, unstable approaches that continue to a landing will continue to occur, with the attendant risk of an approach and -landing accidents.
unstabilised approach Archives - Aviation Accident Database (http://www.aviation-accidents.net/tag/unstabilised-approach/)

Journey Man
27th Dec 2018, 09:36

It isn’t being claimed that stabilised approaches aren’t a contributory factor in landing over run incidents and accidents. If one lands without an overrun from an unstable approach, that erodes your position that an unstable approach causes landing over run incidents.

Your own link clearly states that unstable approaches are a factor. I think your position might have more in common with what PEI_3721 is stating than you realise.

29th Dec 2018, 08:58
When flying a Cessna 172 or a Seneca 3 or even a 19 seat turboprop into commercial airports, the runways are usually so much longer than required, one can get away with a (too) fast approach and still land and stop safely, (as I occasionally discovered in my early training years, ahem :O).

However when one gets onto a medium to heavy jet, the runways become more potentially limiting and one simply cannot take liberties and get away with continuing an unstable approach.

A typical definition of a stable IMC approach is :
At no later than 1000’ AAL, you must be :
On LOC within 1/2 dot
On Glide within 1/2 dot
On final approach speed within -5 and +10 kts, (and trending back to Vapp)
Engines spooled up
Aircraft in correct landing configuration.
Landing checklist completed.

If we don’t meet these criteria, simply go-around. Most Chief pilots won’t even bother asking us about it because they will be happy that we took the safe option and did not bend their aeroplane. However, if one continues a grossly unstable approach and goes off the end like the herbert in this thread, you will most definitely speak with the Chief pilot who will say. “Give me your airside pass. Here is your P45. There is the door.”

So, if unstable at 1000’ IMC; the choice we have is very simple: Go-around, or Look for a new job.

PS, Crewing or duty hour problems are not our problem. Never fly dangerously because of them :ok:

29th Dec 2018, 10:38
I fly for a large airline that has been evolving its stable approach policy and training over the years, to the point where an unstable approach that continues to a landing is an extremely rare event, trending towards zero. How did we get here, considering where we started from: almost no policy at all and a significant proportion of approaches that would be categorised as unstable under today’s rules?

Some of the crucial factors I would list as:

* A well-resourced FOQA setup with non-jeopardy bidirectional feedback
* Simplifying the stable approach criteria so that they could be more easily applied in a high workload situation
* Continuously highlighting and praising examples of discontinued approaches
* Dedicated training modules exploring the effects of environmental conditions and airframe configurations on landing safety
* Instilling a culture where a go-around when in doubt is seen as positive and professional
* Sim practice for rejected landings

There’s obviously a lot more to it, it’s taken a quarter of a century to get there and the job’s not finished yet. I understand there was a defining moment in Flt Ops some time ago where it was realised that it was a matter of *when*, not *if* we were going to have a runway excursion and/or hull loss.

To give some idea, when I joined and was being trained on the 737, I remember being encouraged to land half way down a 4,000m runway because we were vacating at the far end. The only constraint on landings was that it was frowned upon to do glide approaches to touchdown - power up at 100R was fine! Stopping distance was almost never calculated, just “eyeballed” and woe betide you if you touched the Lever of Shame (speedbrake). Funny/scary looking back, especially given the kind of hangovers you were normally sporting...

30th Dec 2018, 00:57
Really don't see how sim training will help prevent this. It was terrible airmanship and decision making the whole way.
The crew disregarded basic principles of flying an airliner. Of course bad things will happen.

30th Dec 2018, 08:21
This type of training can be negative, in last years ADT ( additional days training) we gave candidates various tailwinds/short vectors high speed approaches to illustrate the risks, they knew what was coming of course and most mitigated the risk, it wasn’t helped by using CPH either with 3.5km runways.

Yet i I still see the odd hot/high speed approach more often than not after getting a direct routing they didn’t expect or plan for and a failure to set personal height/speed gates in their own plan.

That we see few if any over runs in Western Europe is down to both ATC and airspace constraints, you’re unlikely to be hot and high after 3 trips around the lambourne hold, however the biggest factor is cultural differences, not only do F/o speak up when unstable approaches are being flown by Captains they’ll call the go around and western Captains in 99.9% of cases will go around and ask why later, this clearly is not the case in the Far East.

I regularly operate into snow covered runways in Northern Norway, the threats are so obvious, BA medium or less, 4 degree glide slopes, short runways, high terrian, tail winds on long/medium finals and very go around focused with option land.

its about mind set and training, but above all cultural both on the flight deck and more important pilot management ( they after all tend to screw things up more often than line pilots)

Here’s wishing you all a happy new year where ever you are and a safer 2019 than 2018 has been in Asia

30th Dec 2018, 09:55
Re ‘unstable approaches are significant factor’, i.e. not a cause. A Boeing study notes that approx 30% of overruns involve an unstable approach, but cite combinations of long, fast, and deceleration (braking action) as factors in the majority of overruns; *1.

The piloting task is to identify and judge situations involving these combinations and the importance of each factor beforehand.

The industry needs to take a radical view of the overrun problem; we must not expect that ‘more of the same’ will improve the situation. Airbus statistics compare the changes achieved with CFIT and LOC; *2

Considering mind-set, airmanship, etc; a simple and possibly effective view of expertise - ‘a checklist of things to think about’;- https://fs.blog/2012/03/daniel-kahneman-on-intuition/
‘The pre-landing briefing is a flight plan for the mind.’
‘The post landing debrief adds a new experience, every time.’
We might wish to consider how and when we brief / debrief; or have we lost the will to learn, - from every landing ?

*1 Boeing (page 7) http://www.boeing.com/commercial/aeromagazine/articles/2014_q2/pdf/AERO_2014q2.pdf
*2 Airbus (page 30) https://www.skybrary.aero/bookshelf/books/4342.pdf

20th Jan 2019, 07:57
I can comfortably say that I have never (knowingly) continued an unstable approach. Does that mean my airspeed has always been knot perfect or my rate of descent hasn’t briefly dipped over 1000fpm? Almost certainly not.
But there surely is a difference between a professional aviator making a brief and recognized error and the Garuda skipper who is a criminal and should never spend a free day roaming the planet again (the captain in this instance would likely be in the same category had the ending been as catastrophic).

20th Jan 2019, 08:40
At my airline, we just changed from 1000' IFR/500' VFR stable to 1000' stable. They also noted that many unstable approaches occurred because the aircraft was configured too late. When I was flying in Africa our "Chief Pilot" made the rule to go gear down at 8 miles out. I have been using that technique, and it seems to work quite well for me (now it is "Gear Down, Flaps 20").

We have "no-fault" go-arounds, and we are encouraged to go-around more than we have been. They would prefer a report (ASAP) for every go-around, but it is not required. We are told any operating crew can call for a go-around, which I don't agree with. If I am, or I have a jumpseater (or deadhead crewmember) in the cockpit, I hope that they would call for a go-around if necessary, and I expect that the flying pilot will go-around regardless of who calls for it.

I know there are/were airlines where the crews would get in trouble if they went around...NOT VERY NICE!

21st Jan 2019, 07:17
. One outfit I know, the civil authority examiner [on the CAA payroll, not an industry delegate] also moonlights for his assigned airline.

Whether this should be reported to ICAO and whether ICAO would take any action on the word of a simulator panel operator is debatable. If ICAO is anything like the U.N. it is toothless.

In my personal experience, not unknown in the UK in years gone by. In the day, add UK CAA to the above.
Tootle pip!!