PDA

View Full Version : Thai low EDDF


gearlever
27th Mar 2020, 23:00
Thai A359 at Frankfurt on Jan 1st 2020, about 800 feet AGL about 7nm from touch down (http://avherald.com/h?article=4d1e782d&opt=0)

Four pilots on the flight deck.

668 ft AGL

DaveReidUK
27th Mar 2020, 23:55
BFU Interim Report:

Kit Sanbumps KG
28th Mar 2020, 07:30
A very poor report indeed, illustrating again, if we needed it, that the habit of publishing ‘facts’ (and in this case very selective facts - ‘It has to be noted that in the background on the channels of the PIC and the co-pilot the on-board entertainment music can quietly be heard’) without any analysis is harmful. The BFU should also know better than to use ‘minimum altitude’ as they did, to confuse ‘descent’ with ‘descend’ and that ‘head towards 340(deg)’ and the crucial ‘speed up the descent’ (their words, not a quote) is highly ambiguous and has no place in a State publication. If their English can’t be better (and I know it can), they should stick to German.

Kit Sanbumps KG
28th Mar 2020, 09:13
However, perhaps the most worrying thing is that this is another case of a major European airport’s ATC vectoring an aircraft towards the localiser, well above the glideslope.

Clearly, the lessons of TK1951 and others have not been learned.

Let’s hope that the BFU are not setting the crew up for a fall here, as seems to be indicated. The final report needs to focus extensively on ATC aspects.

DaveReidUK
28th Mar 2020, 12:16
A very poor report indeed, illustrating again, if we needed it, that the habit of publishing ‘facts’ (and in this case very selective facts - ‘It has to be noted that in the background on the channels of the PIC and the co-pilot the on-board entertainment music can quietly be heard’) without any analysis is harmful. The BFU should also know better than to use ‘minimum altitude’ as they did, to confuse ‘descent’ with ‘descend’ and that ‘head towards 340(deg)’ and the crucial ‘speed up the descent’ (their words, not a quote) is highly ambiguous and has no place in a State publication. If their English can’t be better (and I know it can), they should stick to German.

An Interim Report typically contains neither analysis nor conclusions/probable cause (hence the "interim"). This one is no exception.

VH DSJ
28th Mar 2020, 12:22
An Interim Report typically contains neither analysis nor conclusions/probable cause (hence the "interim"). This one is no exception.

Not to worry, though. PPRuNers will have the cause of this incident worked out by this time tomorrow. ;-)

DIBO
28th Mar 2020, 12:41
to confuse ‘descent’ with ‘descend’Could someone enlighten a non-native English speaker, what is wrong with "instruction to ... and descend to 3,000 ft", "to speed up the descent", "the rate of descent", which are the only cases (1x verb / 2x noun) of descent/d I can find??

gearlever
28th Mar 2020, 13:05
Could someone enlighten a non-native English speaker, what is wrong with "instruction to ... and descend to 3,000 ft", "to speed up the descent", "the rate of descent", which are the only cases (1x verb / 2x noun) of descent/d I can find??

Look here

Focus on: Descent & Descend (https://isabellaferraroblog.wordpress.com/2012/09/20/focus-on-descent-descend/)

Kit Sanbumps KG
28th Mar 2020, 14:31
An Interim Report typically contains neither analysis nor conclusions/probable cause (hence the "interim"). This one is no exception.

Yes Dave, I know. (I've dealt with enough of them).

I'm saying that, in 2020, it should not be so, and this is a particularly bad example. There is enough there to begin an analysis and share it, rather than waiting months on end; there is enough to focus on areas of interest, which are plain from the report to an expert eye but should be exposed. 'The investigation continues, examining in particular:...' would be the very least one might hope for.

The 'safety investigation' world is locked into behaviour which is not relevant in this modern age, but it never changes. It is not the excellent and respectable community that many take it for, and I believe we should all get to grips with its failings.

meleagertoo
28th Mar 2020, 14:54
I'm saying that, in 2020, it should not be so, and this is a particularly bad example. .

As an example of how poor construction can cause confusion I commend the above quote. Not having a dig, just illustrating how poor construction can render a sentence meaningless, or in this case potentially understood in two utterly different ways.

In this case does "bad example" mean that the example itself is a bad one (ie unsuitable or misleading) one, or that it is an example of bad practice giving cause for concern.
It could be taken either way.

Cave!
As they say in Latin!

Jwscud
28th Mar 2020, 20:18
It’s just a rubbish interim report. A graph of altitude versus time with the runway on it isn’t really very useful. Altitude versus DME? Position clearances were given?

This crew were clearly high and cocked it up. No hint is given as to how/why that occurred in this report.

gearlever
28th Mar 2020, 20:29
It’s just a rubbish interim report. A graph of altitude versus time with the runway on it isn’t really very useful. Altitude versus DME? Position clearances were given?

This crew were clearly high and cocked it up. No hint is given as to how/why that occurred in this report.
Page #5 of the report.


https://cimg0.ibsrv.net/gimg/pprune.org-vbulletin/900x539/bfu5_406d499da52fae0de02e20656df7e9888f2b60c5.jpg

DaveReidUK
28th Mar 2020, 21:58
Version at native resolution:

https://cimg6.ibsrv.net/gimg/pprune.org-vbulletin/1136x546/bfu_360b1741016f4a6a6d229d0067fb0b8135b3edf8.jpg

Kit Sanbumps KG
28th Mar 2020, 22:17
JW, those images show that the radar vectored approach was significantly above the glideslope, in contravention of ICAO etc. The BFU note the ‘on-board entertainment music’ and show us the dimensions of an A350, as if we needed them, and say nothing at all about the ATC aspects. I also found the ‘seizing’ of the recorders unnecessarily dramatic.

But to look at that and say the crew ‘cocked it up’ is to be as bad as the BFU.

Deeply incompetent, at best.

hans brinker
28th Mar 2020, 22:40
JW, those images show that the radar vectored approach was significantly above the glideslope, in contravention of ICAO etc. The BFU note the ‘on-board entertainment music’ and show us the dimensions of an A350, as if we needed them, and say nothing at all about the ATC aspects. I also found the ‘seizing’ of the recorders unnecessarily dramatic.

But to look at that and say the crew ‘cocked it up’ is to be as bad as the BFU.

Deeply incompetent, at best.

Yes, the aircraft was above the glideslope, but from the report it is unclear who is responsible for that. They were assigned 3000' and told to expedite descend, so maybe the crew had been slow to descend earlier or maybe the altitude assignment had been late. Impossible to tell with the info we have.

gearlever
29th Mar 2020, 00:51
Yes, the aircraft was above the glideslope, but from the report it is unclear who is responsible for that. They were assigned 3000' and told to expedite descend, so maybe the crew had been slow to descend earlier or maybe the altitude assignment had been late. Impossible to tell with the info we have.
Yep. Very important facts, I agree.
As already mentioned, there are completely irrelevant facts in the report, like dimensions of the plane, blah, blah, but this isn*t unusuale.

Chris2303
29th Mar 2020, 01:10
It’s just a rubbish interim report. A graph of altitude versus time with the runway on it isn’t really very useful. Altitude versus DME? Position clearances were given?

This crew were clearly high and cocked it up. No hint is given as to how/why that occurred in this report.

One can only wonder if this report was written to satisfy the ICAO demand for a preliminary report within 30 days, something about which some people on here are extremely pedantic

sonicbum
29th Mar 2020, 09:30
JW, those images show that the radar vectored approach was significantly above the glideslope, in contravention of ICAO etc. The BFU note the ‘on-board entertainment music’ and show us the dimensions of an A350, as if we needed them, and say nothing at all about the ATC aspects. I also found the ‘seizing’ of the recorders unnecessarily dramatic.

But to look at that and say the crew ‘cocked it up’ is to be as bad as the BFU.

Deeply incompetent, at best.

The aircraft ended up in an Undesirable Aircraft State - Mitigated by the crew by a go around. This is so far the only real "fact" we have.
ATC can definitely be a contributing factor, but it is very unlikely to be the root cause.

Kit Sanbumps KG
29th Mar 2020, 09:33
Ah, the fallacy of the 'root cause'... https://www.safetydifferently.com/wp-content/uploads/2014/08/SDChemicalProcessing.pdf

sonicbum
29th Mar 2020, 09:48
Ah, the fallacy of the 'root cause'... https://www.safetydifferently.com/wp-content/uploads/2014/08/SDChemicalProcessing.pdf

There are some very interesting point in the article you have linked, but in the end I believe it all depends on how "deep" You go when looking for the root cause of an event. I 100% agree it is definitely too easy to hammer the crew as they are ultimately responsible for the safe conduct of the flight and this is where the real just culture should come in, trying to understand what led a competent crew to end up in that situation and fix it, and the root cause could definitely be within the system and not the crew.

ATC Watcher
29th Mar 2020, 10:50
sonicbum : ATC can definitely be a contributing factor, but it is very unlikely to be the root cause.
Well if there is a thing that I have learnt the last decades when investigating incidents is that "root causes" will largely depend on what end of the stick you are or where you want to shift the blame .
From an ATC point of view , in this case it could be the fact that it was a medical emergency, and someone wanted (or was asked to) give the a flight a priority , , and/or the desire of the F/O to show it could do it , etc, etc..
But it looks like an unstable approach due high speed with a high rate of descent with catching the glide from above. Those 3 factors rarely end well, unless you are in a single seater..
Happens relatively often but rarely end at 700ft AGL 7 NM out... hence the investigation , which am sure the BFU will provide a good one in due time .. they normally quite good at it .

I however agree that this interim is not their normal standard , maybe someone new working from home unsupervised...,not surprising in the recent days .. :rolleyes:

Centaurus
29th Mar 2020, 11:50
trying to understand what led a competent crew to end up in that situation

I recall a quote from the Boeing chief test pilot on the 787 project when briefing simulator instructors. He said: " We designed the 787 assuming it will be flown by incompetent pilots - hence all the protections in place."

While a pilot may have passed all required simulator competency tests required by a particular State Regulator, it does not necessarily follow he is automatically competent when flying in command in the real aeroplane. Often ethnic culture is not always compatible with flight safety.

lederhosen
29th Mar 2020, 11:50
This is an interim report and as has been pointed out there are a few other distractions going on right now. But you can get a pretty good feel for things that might have played a factor in what was a serious loss of situational awareness. As ATC Watcher says they had a medical emergency and the controllers were trying to be helpful by shortening their track miles. On the other hand why anyone should be surprised that they might have additional pilots for a thirteen hour flight ending in a period of circadian low has not really thought things through. The pilot flying was pretty experienced for what is a new aircraft type, but with 1500 hours on the aircraft he probably still had relatively few landings most of them ending in a sedate ILS. The captain had 400 hours on type so even less experience. Things got out of hand in a fairly short period of time. The aircraft was vectored very high and the co-pilot flying by hand allowed a high rate of descent to build and take him well below the glide. I would not be at all surprised if magenta line / flight director issues added to the confusion. The captain only called the tower as they went around so was obviously busy with other things. For the tabloids they were seconds from crashing. A more balanced view was that the crew performed poorly and were saved by an aircraft where in this case the warnings worked and you can pull full back stick and get out of a whole heap of trouble.

Kit Sanbumps KG
29th Mar 2020, 12:34
the co-pilot flying by hand allowed a high rate of descent to build and take him well below the glide

At 1927:59 hrs, the controller issued the air traffic control instruction to head towards 340° and descend to 3,000 ft AMSL

At 1930:42 hrs, the CVR had recorded the automatic altitude announcement Two Thousand Five Hundred. At the same time the autopilot was disengaged.

My underlines. It looks, from the text, as if the autopilot was engaged until after the aircraft descended through 3,000 ft.

But looking at the FDR plot (Fig 5), the autopilot was disengaged at about 1930:24, not 18 seconds later as the text states. That's very bad.

If this report was to serve any purpose, and it could have done, it needed to provide other parameters, which would enable much better understanding, including downlinked selected altitude information and autoflight modes.

However, I don't like these guessing games very much and we are heading straight into WYLFIWYF territory (https://www.erikhollnagel.com/onewebmedia/Arbitrariness_of_accident_analysis.pdf).

sonicbum
29th Mar 2020, 14:13
I recall a quote from the Boeing chief test pilot on the 787 project when briefing simulator instructors. He said: " We designed the 787 assuming it will be flown by incompetent pilots - hence all the protections in place."



I find this statement really odd. Incompetent pilots should not be flying, and it is according to ICAO, not according to me.

While a pilot may have passed all required simulator competency tests required by a particular State Regulator, it does not necessarily follow he is automatically competent when flying in command in the real aeroplane.

Legally, yes. Now the question : can a competent pilot end up in a situation where he/she needs retraining ? Yes. Why is that ? Root cause analysis.

Often ethnic culture is not always compatible with flight safety.

Interesting. Can You expand ?

lederhosen
29th Mar 2020, 15:35
The pilot flying expedited descent as instructed using an unspecified autopilot mode. The 2500 callout looks to me like it got his attention that things were getting out of control and triggered him to fly manually in order to try and recover the situation, which is where things ironically only got worse.

Kit you point out rightly that there is a discrepancy of 18 seconds in the report and say that this is very bad. I am not completely sure what you mean, but I think this may well be a 'Zahlendreher' e.g. 24 is 42 reversed (in German we say the numbers the other way around) and am unsure how relevant it is. Anyway to me it looks like the pilot flying got behind the aircraft. ATC were not particularly helpful, but at the end of the day the pilot flying carries the responsibility and needs to say if something is not going to work out.

Your profile says you fly the 737, as I did before flying the Airbus. The basic principles are pretty similar. A good strategy to recover from being way above the glide is to fully configure and slow right down (lower ground speed less vertical speed needed). This is not something ATC always seem to understand and is definitely not compatible with the instruction to maintain 170 kts.

This is a rumour network for professional pilots. For most of us it is 99% entertainment and maybe 1% really useful. But just sometimes there are real nuggets and can be very instructive way earlier than other sources, for example some of the original information on the Coronavirus. This one looks like a number of other situational awareness incidents I can remember for example the Ryanair visual in Memmingen some years ago. The final report will hopefully be more instructive.

Kit Sanbumps KG
29th Mar 2020, 17:08
The pilot flying expedited descent as instructed using an unspecified autopilot mode. The 2500 callout looks to me like it got his attention that things were getting out of control and triggered him to fly manually in order to try and recover the situation

WYLFIWYF, as I warned.

Kit you point out rightly that there is a discrepancy of 18 seconds in the report and say that this is very bad

It's very bad, not least because it has caused you to misunderstand the situation. According to the data, he disconnected before the 'twenty five hundred' call, not after (print out the FDR plot and draw a pencil line vertically up from the A/P disconnect event). That is what is 'very bad' about a report which contains an important factual inaccuracy. Accident investigation must be done with excruciating precision where facts are concerned. (I originally pondered going down the rabbit hole of considering the selected altitude might have been set to a lower value than 3,000, a frequent coping strategy on all types when altitude capture might interfere with glideslope intercept from above, and an open descent towards the lower value would serve as some explanation, if the glideslope were not captured - perhaps because manual flight began and the PF was manoeuvering below the FD cue to achieve a higher RoD while becoming task saturated and missed the glideslope capture - but this is all horrid, dangerous, WYLFIWYF conjecture).

A good strategy to recover from being way above the glide is to fully configure and slow right down (lower ground speed less vertical speed needed). This is not something ATC always seem to understand and is definitely not compatible with the instruction to maintain 170 kts.

That's elementary energy management, which anyone on line should know (but not all do - I've found myself with command upgrade candidates who don't get it). The 170 KIAS instruction is a focus for further investigation. Vectoring high and tight, with a minimum speed restriction... It's not too soon to learn from that aspect alone, and that should be one of the outcomes of a competent interim report, if the world can wake up to the fact that it's 2020.

Zum Glück spreche ich genug Deutsch, um Zahlendrehern zu verstehen. Ich habe auch Erfahrung mit der französischen elektrische Familie.

Kit Sanbumps KG
29th Mar 2020, 20:42
It occurs to me to point out that the requirement for a 30 day interim report is not to satisfy some notional administrative deadline, it’s to let people know important things about what has happened.

In that respect, some SIAs are no longer competent.

BalusKaptan
29th Mar 2020, 21:14
EDDF RNAV arrivals contain many track miles to touchdown with heights and/or speeds at some designated waypoints. It's unusual to fly the entire arrival as published but you don't know which waypoint you maybe short tracked to by ATC. It's mainly dependant on traffic and you cannot always assess it accurately yourself as TFC will be entering the arrival procedure from different sectors and therefore from different frequencies however TCAS does help. The secret here is to program the FMC to be as low as possible and keep the speed back to the minimum the procedure allows modified by your assessment of the traffic. Some of the controllers will assist by telling you expected track miles to run and or to plan to be a particular point at a height &/or speed though not yet cleared to do so. The incidence of observing a/c being high appears to have dramatically increased since the introduction of RNAV arrivals at EDDF. This is not to say the design is not the best but that Pilots need a better understanding of the reason for the design, how the arrival process is used for max efficiency and how to operate their a/c in accordance to it.

DaveReidUK
29th Mar 2020, 22:30
It occurs to me to point out that the requirement for a 30 day interim report is not to satisfy some notional administrative deadline, it’s to let people know important things about what has happened.

The requirement for a 30-day interim report (described as a "Preliminary Report" in Annex 13) only applies to accidents, of which this was not one.

Kit Sanbumps KG
30th Mar 2020, 06:43
The requirement for a 30-day interim report (described as a "Preliminary Report" in Annex 13) only applies to accidents, of which this was not one.

Indeed, but here the BFU have chosen to publish one, whether because they do so habitually or felt moved to do so in this case, or because their State rules are more expansive than Annex 13, I don’t know.

Starbear
30th Mar 2020, 07:34
Yes Dave, I know. (I've dealt with enough of them).

I'm saying that, in 2020, it should not be so, and this is a particularly bad example. There is enough there to begin an analysis and share it, rather than waiting months on end; there is enough to focus on areas of interest, which are plain from the report to an expert eye but should be exposed. 'The investigation continues, examining in particular:...' would be the very least one might hope for.

The 'safety investigation' world is locked into behaviour which is not relevant in this modern age, but it never changes. It is not the excellent and respectable community that many take it for, and I believe we should all get to grips with its failings.
As you have dealt with many of them, surely you would be aware of the statement at the end of this one which says inter alia:

This document is a translation of the German Investigation Report. Although every effort
was made for the translation to be accurate, in the event of any discrepancies the original
German document is the authentic version.

Which from my much more limited experience, I believe is the norm in Germany, France and several other states/countries. This surely goes someway to explaining some of the,admittedly odd language examples, such "Head towards 340deg etc.

lederhosen
30th Mar 2020, 08:06
You make some interesting points Kit and there is certainly some conflicting information in the report. If we were analysing why the aircraft was in the wrong place to make the approach and subsequently needed to go-around then ATC appear to have played a part. But if the question is why did the aircraft have a high rate of descent below three thousand feet and come perilously close to hitting the ground (which is a lot more unusual than being vectored high) then I expect the report is more likely to focus on instrument flying skills.

Using a format we pilots are familiar with, I would ask the exam question: which is the most likely? (in time honoured fashion there is only one right answer, my test my rules.)

The A350 had an excessive sink rate at low altitude, flew well below the glide slope and nearly collided with the ground because:

A.) ATC were to blame
B.) Poor instrument flying skills of tired long haul pilots
C.) Hacking of the flight control system
D.) Poor quality of aircraft accident investigation

The internet is an imperfect medium for communication and humour does not always translate well. I am just as capable as the next person of getting the wrong end of the stick (as I have amply demonstrated on previous occasions), and I look forward to reading a hopefully instructive and comprehensive final report.

Instrument flying skills are a perishable commodity and if my hypothesis is correct perhaps those of us flying less in the current situation need to think about how we deal with this issue.