ATSB report on very low flying Thai Airways B777 at Melbourne.
 

The latest ATSB report is worth reading. It took nearly two years to produce. I recall media reports at the time where residents were startled to literally feel a low flying wide-body barely 500 feet above their houses.

Investigation: AO-2011-086 - Operational non-compliance involving Boeing 777, HS-TKD, 15 km south Melbourne Airport, Vic, 24 July 2011

To have a Thai Airways Boeing 777 so low on final at such a long way out during an attempt at Melbourne runway 34 VOR/DME approach is frightening. This is a basic instrument approach which any general aviation pilot could do without drama. It proves what most professional pilots already suspect and that is just because you fly a bloody great jet transport does not necessarily mean you are a good pilot. The above report included reference to another Thai Airways 777 which flew dangerously low on final while trying to fly the Melbourne 16 NDB. Nothing wrong with the automatics but something seriously amiss with the competency of both captains. Of course, that view does not appear in the official reports by ATSB.:=

The real question should be! Why does YMML not have ILS approaches on alp runways? Having flown longhaul you can go years without flying a VOR approach except the odd one in the sim. Is that an excuse... No, but surely OZ needs to start playing an active role in SAFETY and do all they can to provide the most robust approach systems at international airports! OOL also springs to mind.

I am not an expert by any means, but........

To borrow a phrase, reading this report reminds me of drowning to death in maple syrup.

Could I be forgiven for thinking that being a Thousand feet below minimum safe altitude for a particular point on an approach over a populated city is effing a critical incident?

To put that in context, the minimum safe altitude was 1950 ft and this genius got down to 984 ft, in other words 1000 ft below where he was supposed to be.

To put that another way, this aircraft was off altitude by FIFTY PERCENT!

...And the ATSB blandly reports:

This is a masterpeice of understatement. Tautological nonsense. "The aircraft was lower than required" - that should have read "The aircraft was lower than commanded because the pilot was unable to comply with the instructions of the tower." , given that the elevation at the point of lowest approach was about 130 ft, this bloke shoved a jet to within 850 feet of a Melbourne residential suburb with no safety action apart from "don't do it again"? Well I suppose the ATSB wouldn't want to be seen as anything like judgemental would it?

God help us all. The ATSB has been completely and utterly neutered.

That little rant out of the way, I agree with the call for an ILS installation at every runway serving jet ops. Say what you like about skill levels, tired old guys make mistakes. Non-precision approaches are never as safe as an ILS...and the world is full of craters to prove it.

A major airport without a full suite of precision approaches is like a country with two lane national highways. Oh wait....

Thai
 

Putting aside precision vs non-precision approaches, it makes you wonder, with the frequency that Thai appear in ATSB reports for non-precision approaches into Melbourne, if thats a statistical blip or symptomatic of thier ability (or otherwise) anywhere in the world with a non-precision approach.

Not sure I'm game to fly Thai based on this and the previous NDB incident onto 16 about 4 years ago.

Comments regarding the ATSB and government shall be kept in the numerous threads already running.

Let's just concentrate on the incident, thanks.

Don't be so modest Sunfish (refer senate hearing thread)
The report should have read
" Pilot struggles to regain control as Aircraft plunges within metres of school and suburbs, screaming passengers declared that they thought their days were over. ATSB police department have arrested the pilot and charged him with an offence under the plunging aircraft regulation."

Repeated incidents like this will see you get banned from operating in some countries, that is countries that take safety seriously. And it must be said Tids that if a safety agency has a list of these incidents on file & god forbid, 'you know what' happens. Maybe this is what's going to have to happen to get rid of these idiot, PC, incompent authorities?

And 50 seconds to action the go around instruction. That alone should have them denied access to to Australian airports.

The handling of the approach sounds incompetent, but what part of the 'Go Around' instruction did they not understand,that it took 50 secs to comply.

Diverging:-)
 

Whilst such events as this one is of concern I am of the belief that humans are their worst own enemies when it comes to automation.

Mankind invented the lighter than air machine, essentially the aeroplane many many years ago now & has improved out of sight it's capabilities from the days where Wilbur & Orville stood there on Kill-Devil hills with a machine in it's infancy but mankind himself has not evolved one inch in some ways.
Sure we have advanced training via Sims etc, higher education abilities thru knowledge gained over the years & multiple bad mistakes in which to learn from but still we humans take a perfectly serviceable plane complete with hapless pax straight into the side of a hill!
Even though the human has almost been taken out of the equation here in the way of automation (pilot-less planes are already here & technically we are not really needed up at the pointy end other than redundancy) we still do what humans do best, make mistakes. You can make 100 airframes identical to perform exactly the same way every time but you will NOT make a 100 pilots do likewise, there in lies the "oil & water" interface that will never be bred or automated out.
Checklists, training, Psycho testing crap at interview level all the stuff that's meant to make the human the best he can amounts to little at the end of the day, the above 'process' weeds out a few but that's all.
CRM is a modern day "feel good" terminology & it's an industry in itself but whether it's making or going to make a noticeable impact on why humans make mistakes is arguable I believe. Back before they invented the hairy fairy CRM thingy Capt's of flying machines when faced with adversary would have enlisted the help of everything & everyone so save if nothing else his own butt so CRM has essentially been around well before it was dreamed up!
Australia has some pretty ordinary recorded events from our own pilots so Thia are not the only Airline under question here I believe.

The case in question here the 777 event is the perfect Eg of humans in action when man & machine are interfaced. The weather was fine, the crew where trained (well meant to be) the machine was one of the finest available but the human was/is the weakest link in the chain of events here as in any incident/accident.
I know I've looked back on some silly things I've done (especially in the Sim) & wondered why the hell did I do that when I know better?
The human mind, we know very little about it which shows every now & then & not just thru aviation either:-)

We can improve, we can be seen to be doing the right thing to reduce but we simply have to accept that incidents such as this one will continue to happen as long as mankind has the desire to fly:-)

Just some food for thought not meant to be anything else here:-)


Wmk2


Wmk2

I will abide by Tids request, however it is very difficult not to mention government agencies when it appears they themselves could be part of the problem or be a contributing factor. And what I mean by that is areas such as airspace, ground infrastructure, authority oversight, procedures, regulations and investigations are just a couple of those points.

Be that as it may, back to the airline in question...

Once again, the ATSB raises more questions than it answers. But you have to wonder that if it takes 18 months to issue a relatively straight forward report - how long does it take for a complex one.

The ATSB are pretty good at using units of measurement that confuse rather than illuminate. For instance, nearly all the altitude reports are done against time, but the glideslope is defined by distance. A diligent report would have used distance so that the report was transparent someone reviewing it could plot it against the chart. The ATSB have denied us this ability.

The ATSB are also a bit prone to using emotive language / distortion. I'm not sure that the delay to go-around was quite as it was represented. And of course there is no transcript of radio calls which once again denies independent review. Another favorite ATSB trick. Other international agencies not only publish the relevant transcript, but also publish any corrections - including those requested by the participants after they were given the opportunity to review it. Another thing the ATSB doesn't seem to do.

The aeroplane was too low. There is no getting around that and the captain did admit that, but a point missed by Sunfish and glossed over by the ATSB is that they were cleared for visual approach. The controllers comments of confirming that the aircraft was low by visual identification confirms that the conditions were visual. Others will correct me if I'm wrong, but I'm pretty sure the command "cleared for visual approach" waives the approach altitude requirements and allows the pilot to descend at his / her discretion to make the landing. On that basis the pilot was guilty of poor airmanship, but not of breaching the approach procedure. It is entirely likely at the time of the instruction to go around, the aircraft was back on glideslope. According to the pilots description (confirmed by the ATSB report) he an initiated recovery before the ATC alert.

The ATSB listing of the sequence of events is not fully clear, but I would question whether the aircraft was as far below the glideslope as suggested. The report lists the minimum altitude that the aircraft reached, but is not clear where this was relative to the glideslope. It was inside the 6.5nm point at its lowest, so the comparison of 984 ft to 1950 ft is not valid. If the ATSB used distance from the aid rather than time, we could have a go at working it out. The MDA inside 6.5 mile is 760 ft, so the aircraft was ABOVE the published minima. Once again, the pilot was guilty of poor airmanship, but not necessarily in breach of the procedure.

It will require some work to try and reconstruct, but I suspect that the most serious breach may have been while the aircraft was flying the 11 DME arc and under ATC direction (ie before it was cleared for approach). I suspect that this is the only part where the aircraft breached the LSALT. It was above MDA when it was at 984 ft which is where the attention is focused (and its easier to say it was the pilots fault exclusively).

The big question - which has already been raised - is why we don't have ILS on all runways. The crew was from Thailand. They may not be used to operating in third world countries like Australia that do not have the funds to invest in basic safety infrastructure. Has anyone noticed that the YWE VOR is still U/S? Together with the procedure having been removed from CWS - exactly where does one train for VOR approaches in Melbourne at the moment?

The other question I have is whether the controller should be expected to pick up a diversion from glideslope before it becomes such a large incident? Should the pilot expect a warning before being "waved off"the approach? This is a genuine question for the ATC among us and not a barbed comment.

Wondering where the pilot derived the 970' which was set in the MCP?

It's not a number which would make sense under my SoPs, does anyone know how it may have been calculated?

I'd be a bit careful about this. We have no transcript to be able to confirm this. It makes me as mad as hell that the ATSB don't do this and I suspect that it is done consciously to avoid scrutiny. Try reading a US NTSB or UK AAIB report to see it done nicely.

In the body copy (not under the go-around sun head) its pretty clear that the pilot initially mis-understood the ATC instruction. At the time of the go-around instruction the aircraft had recovered altitude and was probably at glideslope (hard to tell the way the ATSB present data). I think the ATSB claim of 50 seconds delay is mischievous. I suspect they have taken the time from the ATC first keying the mic to a response in altitude change (ie including engine spool-up time). The report states that the controller issues a second instruction after 35 sec. If it was so critical, why did the controller wait 35 sec for a second call? I question if the pilot heard the first call as an altitude warning and only acted on the second call. If we had a transcript, we could make our own judgements, but we do not. The Mojave Bankstown ATSB report has discrepancies between the draft and final reports which raises questions about whether the ATSB change transcripts to suit the report.

If the second call was the only one the pilot heard as an instruction, then his response time was under 15 seconds, which doesn't deserve the vitriol of the ATSB report.

Wal, there is CRM & there is plain incompetence

I believe it is a `culture` problem. First of all it is probably "it can't be happening to me" or even the captain thinking "WTF is ATC on about- I can see the runway even though it looks a bit flat" Quickly followed by "real men don't go around" otherwise it's loss of face.

Akro, in Australia an ATC is guilty until proven innocent & when the ATC is found innocent he is still guilty.

If an ATC sees something that 'doesn't seem right' regardless of whether a pilot is doing the right thing/wrong thing they are obliged to check.

As quoted in Footnote 4 on Page 1:

ATC, its workload, clarity and wording of the call are all absent from the report. Thailand uses 5 letter registrations (like Australia). Were there any aircraft in the Melb TCA with similar sounding 3 letter call signs? Did the mic clip one of the call sign letters which led to confusion on the part of the captain?

These are valid issues that may or may not have contributed that the ATSB should have considered, but are absent from the report.

With the lack of clarity of the ATSB report, I'm prepared to give the pilot some benefit of the doubt and argue that his response time may have been 15 seconds, not the 50 seconds that is emotively used in the ATSB report.

If there was any hint of confusion about the calls, then both sides of the coin need to be examined, not just the PIC.

I think that it is indefensible that the ATSB do not append transcripts to their reports as is common in other countries.

I think this is a case study of another sub standard ATSB report.

The 970' might well be an irrelevant lower altitude in the FO's mind to simply prevent ALT CAPTURE when cleared the visual approach. That said I wouldn't be doing that when cleared a visual approach at night in mist... And certainly not outside the circling area. I must feel for these long haul guys when given their STAR... To think a RWY such as RWY34 has no ILS and the amount it's actually used is VERY third world.

Actually, no. They were cleared for a VOR approach, and subsequently a visual approach once established on the PAPI and inside the circling area (5.28NM final).

At night you cannot commence a visual approach until this point which the controller spelled out to them.

They were on an instrument approach and clearly breached the instrument approach requirements. Using FLCH on an instrument approach at night in rain without the relevant minimum altitude set in the MCP according to the approach plate would be a hanging offence in my airline. The only acceptable modes are V/S or VNAV.

Akro, very few fly on rego now - it's all flight number callsigns, except for lighties and the odd biz jet. The RTF would have been something like "Thai four seventy six".

Good point about flight number that I should have realized

Thanks Buckshot, I should have looked further.

In rain, at night especially, there can be a diffraction error in the visible picture. Certainly not of enough magnitude to explain this degree of vertical deviation, but it can be a contributory factor to being a couple of hundred feet low that far out.

Night visual approach requirements
 

The requirement to be inside the circling area at night may be unique to Australia. I cannot recall it being in force in any other country in which I am licensed. Here is a fair use copy from the ICAO regs in the Jepp WWT:

Visual Approach 6.5.3.1 Subject to the conditions in 6.5.3.3, clearance for an IFR flight to execute a visual approach may be requested by a flight crew or initiated by the controller. In the latter case, the concurrence of the flight crew shall be required. 6.5.3.3 An IFR flight may be cleared to execute a visual approach provided that the pilot can maintain visual reference to the terrain and;

a. the reported ceiling is at or above the level of the beginning of the initial approach segment for the aircraft so cleared; or

b. the pilot reports at the level of the beginning of the initial approach segment or at any time during the instrument approach procedure that the meteorological conditions are such that with reasonable assurance a visual approach and landing can be completed.

6.5.3.4 Separation shall be provided between an aircraft cleared to execute a visual approach and other arriving and departing aircraft. 6.5.3.5 For successive visual approaches, separation shall be maintained by the controller until the pilot of a succeeding aircraft reports having the preceding aircraft in sight. The aircraft shall then be instructed to follow and maintain own separation from the preceding aircraft. Etc...

Twin Beech, thanks, but once again I'm going to repeat that you'd expect that out of 30 pages the ATSB might devote 1 sentences to whether the Australian procedure differed from ICAO and might have been a factor. The ATSB reports repeatedly fail to identify and deal with all potentially relevant factors.

Going back to an earlier point of mine, the ATSB does not identify whether the captain acknowledged any or all of the ATC instructions. The report beats up the pilot for not responding to a go-around request but is mute on whether the pilot acknowledged or read back the instruction. If he did, then he's got no-where to hide. If he did not, then the controller needs to be asked why he waited a further 35 seconds before trying again.

At the end of the day (or night) once you pass 6.5DME I guess you could dive straight to the MDA, it would be technically legal but not SOP for an airliner. I agree with Akro. Give the transcripts and a graph with the profile of the approach, it would be readily available and much more useful, than the lawyer version of this so called ATSB investigation.

Australia certainly seems unusual to me in that they don't have co located DME's on all their ILS's and in some cases don't even have an ILS. That is not 'worlds best practice' or whatever they like to say.
Also, I have often flown with Englishmen and Americans who have trouble understanding the Australian accent when it is spoken quickly as it often is over the radio.
That said, I think that quality training probably is the biggest issue in this incident.

DME x 300 works really well for a 3 degree approach on to this runway because of the combination of elevation and DME distance from the threshold.

It's published on the plate anyway.

In short, there's no excuse for a competent crew not to be able to fly a VOR approach.

However, there's also really no excuse in this day and age for a major international airport not to have ILS approaches to all runways.

There is blame to be shared in a number of directions here.

Is it just me or does this event mirror Air Asia X's effort up at the Gold Coast?

Two Asian Airlines who separately turn inbound and dive down to the minima.

So the question to ask are NPA approaches in Asia written differently?

What on earth would make these guys think they can just intercept final and dive down to 4-500 feet AGL?

If it keeps happening like this someone is going to plough into a hill or building.

Why didn't either of them descend in VNAV or VS?

Why on earth does Australia have such lack of infrastructure at International airports?

There's nothing wrong with infrastructure at Australian shopping centres, it's just too bad you have to go through security these days.

Bbbzbzbzbzbzbzbzbzbzbb

To me it's quite telling where the captain stated he didn't expect the AFDS to climb the aircraft to meet the path when VNAV PTH was announciated on the FMA.
I fly this approach into YMML fairly regularly but normally ask for the RNAV due to better minima being available.
Having flown with an ex Thai guy into YMML, he tried to do exactly the same thing on the VOR, i.e dive and drive, inspite of our company SOP of a constant 3 degree descent on a non-precision approach, so I'm wondering if its as much a company culture thing or perhaps the way were initially trained?
In any case there is really no excuse, the FMC if you check it against the approach plate allows for a beautiful arrival with no level segments, it works great on the 777.
The visual call to the tower says to me he might have used that as a get out of jail call....

VNAV Path is nice but during approaches a nightmare . Enough said, just aviate and it will be fine..

Dive and drive at MSA on non precision approaches and on arcs is still common in the parts of SE Asia I have flown in.

A lot of huffing and puffing about their lack of ability to fly a VOR approach. I doubt that is the case. So many experts!

However, as the report points out, it was more to do with automation.

Automation confusion lead to a lack of situational awareness.

If they planned that approach using Vertical Speed and a Lateral Nav Mode from the start, there would have been no problem.

They were caught out by the unexpected behaviour of VNAV PATH, and from that point were unable to recover from the automation surprise!

And that is a big problem these days.

Voice recorder?
 

If the ATSB did not have access to the Thai CVR, how are they able to state that the ATC transmissions were not blocked by other radio traffic as heard in the Thai flight deck? I can't count how many times I have heard some **** call ready in turn (when he's number ten) block a late landing clearance for someone else on short final for example.

Indeed haughtney the captain clearly didn't/doesn't understand 'on approach mode' in the 777 which will indeed climb to get back onto the path if you have been dive&driving in FLCH and then select VNAV.

I too have flown that approach many times in a 777 and if you simply put the approach & transition in and leave the aircraft in LNAV/VNAV it's a doddle.

The approach has a 3 degree path coded in and from 8dme you'd be hard pressed to tell the difference between a VOR APP and an ILS except for the higher minima and the FMAs would say /LNAV/VNAV PATH instead of /LOC/GS.

This incident is purely an example of a crew not understanding their aircraft - a training and standards issue. I'd be interested to hear what the captain was flying before the 777 and his time on type.

Why is anyone surprised the ATSB report is seriously deficient?

No doubt the appropriate Australian civil aviation authorities will have long ago examined the feasability of putting an ILS on Runway 34. Part of that study included looking at the number of times in a year the weather from the north was so bad that an ILS was needed. Northerlies in Melbourne rarely bring low vis unless a thunderstorm is passing through and in that case the aircraft have no business trying to land during a thunderstorm.
The tax payer eventually foots the bill for an ILS installation and on-going maintenance and I for one would object my hard earned money as a delivery van driver being spent on some foreign airliner crewed by two incompetents simply because they are out of their depth on basic instrument flying skills.

Some months ago I was talking to a former foreign student of mine who flew into Melbourne acting in command under supervision in an A330 from overseas. Never mind the name. I recall that day. Strong northerlies and ATC gave duty runway as 34. The captain of the A330 declined even though the runway was two miles long and a 25 knot HW component. He asked for 27 as it had an ILS and it was a fine day except for gusty northerlies.
The aircraft landed heavily as the ICUS pilot left autopilot disconnect very late and heavy braking was needed.

The ICUS pilot told me the reason why the captain did not want to use runway 34 was he was unsure of the VOR/DME arc needed for 34 and didn't like using PAPI - preferring to accept the the strong cross wind on the relatively shorter 27 runway as the aircraft could be auto-coupled right down to the flare. Make no mistake about it, there are overseas airlines operating to Melbourne every day crewed occasionally by incompetent pilots, who rely blindly on the automatic coupled ILS approach as the preferred method of landing. Melbourne does not need an ILS on runway 34 since the present VOR approach is perfectly safe and easy to fly and already has a low MDA plus PAPI on all runways.