Halifax crash report coming Thursday

[RatherBeFlying]

My big problem with FPA in this aircraft is that it provides a putative glideslope that a significant headwind can drift a couple hundred yards short of the threshold.

Also that pilot training omits to make that clear.

The illusion of a precision approach without precision.

[rookiepilot]

I have an outdated(2013) CAP 7 Atlantic and there is a published GPS for runway 05 with LPV portion down to 720ft(257) and visibility of 1 mile. I assume the AC A320 fleet do not have the avionics to do these approaches that's why they did the LOC, but I stand corrected on the A320 equipment.

All AC fleets can do RNAV approaches to LNAV/VNAV minimums except for the non-GPS 320's. The company is in the process of upgrading the 320 fleet now. Some fleets can do RNP (AR) approaches, but none can do LPV approaches because no AC aircraft are equipped with WAAS.

Why not?

Must AC have safety equipment regulated before it's placed on aircraft?

One thinks if it wasn't mandated planes would have 1 exit door and one FA. Maybe.

Seems to me this is once again all about the almighty dollar. Nothing more. Always pushing to cut corners.

[BTD]

My big problem with FPA in this aircraft is that it provides a putative glideslope that a significant headwind can drift a couple hundred yards short of the threshold.

Also that pilot training omits to make that clear.

The illusion of a precision approach without precision.

I cannot speak to the Airbus although the report describes the system and it appears to work the same as the aircraft I fly:

The FPA uses the air data computers to calculate the flight path angle. Part of that function includes winds etc. A significant headwind or tailwind will not affect the FPA. The aircraft takes this into account and adjusts to compensate. A 3 degree FPA would look the same on a profile view if the aircraft was in a headwind or a tailwind.

The problem is when you have continuously varying gusts. In those gusts the aircraft's inertia and the delay while the ADCs figure out the change in wind, will allow the real FPA to change. There would be no indication on the flight deck. If there was, the airplane would have to recognize it and might as well just compensate for it.

On my aircraft we have a vertical profile view which allows us to see how the FPA is working, the Airbus does not, and it would be challenging to figureout how it is working when you are descending on cold weather corrections and the correction is decreasing at a variable rate as the ground gets closer.

Ps. I don't like these approaches very much.

[Jimmy2]

and we are supposed to fly with crews that can not figure out if they can see the runway or are imagining they can see it..

Good defense of that accident though, they were not sure so they just touched down to see if it was a runway or not.

Exactly. The approach in this case may not have been perfect but lets all remember the purpose of this approach is to put you in a position to acquire enough visual reference to continue to land visually. As others have mentioned, and as the US regulations explicitly point out, just seeing an approach light doesn't mean you can just continue descending into the ground hoping the runway will appear beneath you.

There is still some major issues with human factors that despite airlines having 'no fault go around' policies there are still pilots trying to 'get in' even at the major airline level. The report mentions one human factor that seemed quite relevant:

Plan continuation, a form of confirmation bias, is a "deep-rooted tendency of individuals to continue their original plan of action even when changing circumstances require a new plan."Footnote 94 Once a plan is made and committed to, it becomes more difficult for stimuli or conditions in the environment to be recognized as indicating change than if a plan had not been made. For a pilot to recognize and act on a reason to change the plan in a timely manner, a condition or stimulus needs to be perceived as sufficiently salient to require immediate action.

As is the case with most accidents we as pilots try to search, often in vain, for reasons for the accident other than pilots. 'It's the airplane's fault', 'Sue Airbus'. Nobody ever seems to want to admit that once again a perfectly good airplane was flown into the ground because the pilots couldn't see where they were going.

I'm disturbed by the response Cat Driver got, disturbed by those saying the pilots were just following the rules of the approach ban and required visual reference, and disturbed by comments like "deficiencies all round in which this flight crew got caught up in through no real fault of their own". It's like you guys are robots that will just follow the rules and sop's to a T expecting them to keep you alive.

[Cat Driver]

I'm disturbed by the response Cat Driver got,

I'm not disturbed by the responses I get because it is that mentality that wreck mechanically perfect airplanes.

Remember that machine was wrecked way " before " the start of the runway and the normal touch down area is a thousand feet " after " the start of the runway for that category of airplane.

Hard landing.....yeh sure it was because there was all those approach lights to get through.

[BTD]

and we are supposed to fly with crews that can not figure out if they can see the runway or are imagining they can see it..

Good defense of that accident though, they were not sure so they just touched down to see if it was a runway or not.

Exactly. The approach in this case may not have been perfect but lets all remember the purpose of this approach is to put you in a position to acquire enough visual reference to continue to land visually. As others have mentioned, and as the US regulations explicitly point out, just seeing an approach light doesn't mean you can just continue descending into the ground hoping the runway will appear beneath you.

There is still some major issues with human factors that despite airlines having 'no fault go around' policies there are still pilots trying to 'get in' even at the major airline level. The report mentions one human factor that seemed quite relevant:

Plan continuation, a form of confirmation bias, is a "deep-rooted tendency of individuals to continue their original plan of action even when changing circumstances require a new plan."Footnote 94 Once a plan is made and committed to, it becomes more difficult for stimuli or conditions in the environment to be recognized as indicating change than if a plan had not been made. For a pilot to recognize and act on a reason to change the plan in a timely manner, a condition or stimulus needs to be perceived as sufficiently salient to require immediate action.

As is the case with most accidents we as pilots try to search, often in vain, for reasons for the accident other than pilots. 'It's the airplane's fault', 'Sue Airbus'. Nobody ever seems to want to admit that once again a perfectly good airplane was flown into the ground because the pilots couldn't see where they were going.

I'm disturbed by the response Cat Driver got, disturbed by those saying the pilots were just following the rules of the approach ban and required visual reference, and disturbed by comments like "deficiencies all round in which this flight crew got caught up in through no real fault of their own". It's like you guys are robots that will just follow the rules and sop's to a T expecting them to keep you alive.

I'll assume that this is primarily directed at my response.

The reason cat driver got that response is based on his sarcastic response to my honest
lack of clarity on my first read of the report. It is hard to tell if he has even read any of the report.

I already acknowledged that this crew made mistakes, indeed ones that had major consequences, but the system sure didn't help trap the errors they made. And it almost seems as though it was setup to let those errors compound.

The pilots weren't "just" following sops. They were following the sops and guidance closely which is one example that demonstrates this crew cared about what they were doing. Your assumption that they continued descending hoping things would work out is not demonstrated in the report. In fact the report points out the opposite. They had what they thought was good visual reference. They discussed this after the original lights in sight call.

You imply this crew just wanted to "get in". Again this is not demonstrated in the report. They had discussed diversions, fuel, and back up plans in the event they couldn't or didn't get in. The facts seem to show they were prepared and willing to go to an alternate.

Just to reiterate. They made errors, costly ones on a challenging night. But don't paint them as a couple of Cowboys willing to risk lives for the sake of a YHZ layover.

[brooks]

Sounds like the Airbus 320 is a piece of sh!t and AC's SOPs are inadequate.

[complexintentions]

Of course they wanted to get in, it's the goal of every approach, is it not? Yet I don't think anyone has remotely implied the pilots were "cowboys".

Ultimately the responsibility for the accident has to rest on the crew. Note I use the word "responsibility", not "blame" or "fault". I'm not interested in pointing fingers at what I am absolutely sure are highly capable colleagues, but I think Cat and a couple others are right to feel uneasy about the perceived attempts to smother this accident in accident report jargonese.

It could be said that in simplest terms a pilot's main task is ALWAYS being completely aware of his aircraft's position in time and space. Whether that is achieved by using advanced technology, or ADF bearings and a damn stopwatch, is secondary. In this case SA was not maintained, with the predictable outcome. They thought they were somewhere they weren't. Why? It's fine to make a laundry list of supporting factors like lighting, type of approach, FPA, training, procedures, AC's shitty old 320 fleet, etc etc and so on forever. But in the end it's up to the pilots to use the tools at their disposal.

I hate FPA mode. If I have to do a basic modes approach - which is exceedingly rare - I always use V/S so that I make the adjustments to the descent rate, not the aircraft. I'll pick an initial rate based on groundspeed and then make small adjustments as I descend. If my calculation is even close it usually doesn't even need to be touched. But that forces me to constantly consider groundspeed, height, and distance to run. Yes, it's more work, at a busy phase of the flight, but I consider it a worthwhile tradeoff to retain my own SA. That's just my preference, some people love FPA, fine. Maybe it's better on the Airbus, but on the Boeing I always found it crude and laggy. But either way I cannot conceive of flying a basic modes approach without distance/height crosschecks. SOP or no SOP, you have to know where you are. Hell I do that from habit on every single approach, starting from top of descent. Whether ILS, RNAV, or RNP-AR I'm doing my own mental math all the way in to crosscheck the magic. But I'm a suspicious sort. Some might say paranoid.

It was suggested that this crew put too much faith in FPA, considering it "quasi-precision" guidance. I hope that's not the case, it's a serious lack of systems knowledge if it is.

Lest this post be taken as "malicious", let it be known I wish the crew no ill will. But I would hate if too much earnest "swiss cheese modelling" also served to dilute the message that you have to always, always, always know where you are and what the aircraft is doing. And if the weather, aircraft and/or the airport just aren't good enough, go somewhere else.

Then you can bitch about AC and YHZ airport comfortably in the bar in YQM or wherever.

[Rockie]

My big problem with FPA in this aircraft is that it provides a putative glideslope that a significant headwind can drift a couple hundred yards short of the threshold.

Also that pilot training omits to make that clear.

The illusion of a precision approach without precision.

False. It is not a putative glideslope and no AC pilots think that it is. It is an angle through space unattached to anything.

This crew was operating entirely within regulations, Airbus operating procedures and company SOP's - which is the problem. Back then AC had three responses when reaching MDA, 1) "Landing", 2) "Go-around", or 3) "Lights Only". The "Lights Only" call allowed us to continue provided we could see some portion of the approach lights, which was and still is fully compliant with Canada's inadequate required visual references regulation.

Halifax has ODALS on RWY 05 which can easily be confused with cultural ground lighting in very poor visibility. Coupled with the continued legality in this country of flying a non-precision approach with only half the charted visibility, it's a wonder this kind of accident hasn't happened long before now. Maybe it has, but hasn't gotten the same exposure AC624 did.

Since then Air Canada has eliminated the "lights Only" call and implemented their own much more restrictive visibility limits than what regulations permit. They've also been very persistently vocal to the crews that if we don't like the conditions for landing we are not to land. It isn't just noise either - they mean it.

This accident shone a spotlight on Canada's unique and unsafe approach ban regulation as well as our inadequate "Required Visual Reference" regulation. It was my hope both the TSB and TC would acknowledge that and fix it. Looks like I was wrong.

[fish4life]

One important factor is the lack of approach capability still. Why isn't AC's entire fleet LPV capable? Money it's as simple as that, companies all over the north are investing in full glass LPV capable avionics yet AC doesn't want to spend a few bucks to make their fleet LPV capable.

[RatherBeFlying]

False. It is not a putative glideslope and no AC pilots think that it is. It is an angle through space unattached to anything

Then there needs to be head and tail wind limits when FPA is used.

Yes, I have a real problem with AB's engineering of FPA on this aircraft; so do AC's lawyers.

[Rockie]

Then there needs to be head and tail wind limits when FPA is used.

Yes, I have a real problem with AB's engineering of FPA on this aircraft; so do AC's lawyers.

Steady headwinds, tailwinds or crosswinds are not an issue. It's when the wind changes, or when there's turbulence or any other disruption that moves the aircraft momentarily off the selected FPA. When that happens the aircraft resumes the selected FPA but cannot return to the original flight path location. That's why we configure the aircraft before selecting and intercepting the final approach descent.

It is a well known characteristic of the system that is accounted for. There's nothing wrong with it at all provided it's used correctly, in fact it's a huge benefit that makes executing NPA's safer and more accurate.

[TailwheelPilot]

There is no difference between the two methods, they both account for a small inertia delay so that by the time the descent point is crossed the airplane is starting down on the correct path. Either way has been used for years by multiple companies with no problem.

The problem is that it is a non-precision approach with no vertical guidance. While it is unquestionably safer to fly it as a SCDA it denies the crew the time necessary to make a proper determination of aircraft position and trajectory at MDA. This forces them under the Canadian approach ban rules to make a snap assessment at minimums (with half the charted visibility) and either continue to land or go-around. It's a recipe for failure without the charted visibility.

Monitoring the altitude/range past the FAF could be useful, but don't forget this was in winter and the altimeter was subject to cold temperature errors making the charted altitudes wrong anyway. And even if the crew detected a difference there's nothing they could do about it because we cannot change the FPA beyond the FAF (for good reason). All it would do is alert the crew that they may not be in a position to land, which they have to be ready for regardless because it's a non-precision approach.

I see. It would have been clearer if the TSB said although the procedure was different, the result is the same for people not familiar with the Airbus.

The FCOM quoted by the TSB says "POSITION/FLIGHT PATH---MONITOR/ADJUST". When on an FPA approach, autopilot still on, how do you adjust the flight path if you cannot change the FPA?

[Flying Low]

http://www.avherald.com/h?article=484c306e&opt=1

It took a year and a half for the Japanese to get their report done on a similar accident. An interesting read.

[Say Altitude]

The ONLY reason GPS is being put in the baby Bus is because of Halifax. It's all about the cash - no problem outfitting the fleet with gogo wifi because it'll be another source of ancillary revenue, but spend money on safety?? C'mon, we all know where the priority lies.......

[Gilles Hudicourt]

A few quotes from the report :

The occurrence aircraft was equipped with the appropriate navigational aids to conduct a LOC non-precision approach, and these aids were serviceable at the time of the accident.

The Air Canada FCTM indicates that the flight crew should monitor the aircraft's position along the vertical flight path while conducting a non-precision approach, by referring to the distance indicated on the distance measuring equipment, the altimeter reading, and the time.

The SOPs state that the appropriate raw data on the navigational display should be selected for monitoring the tracking and the descent. As well, the flight crew should use the bird in the PFD as the flying reference.

From the Air Canada SOP :

At 0.3 NAUTICAL MILES prior to FAF, or descent point if other than FAF, if vertical guidance selected

FPA................................PF..............SET

Do not use the Vertical Deviation on the PFD.

Set charted FPA to follow the published descent profile, taking into account the minimum altitudes.

Use radio navaid raw data to monitor the lateral navigation during a conventional NPA.

Note: For RNAV (GPS) approaches the descent point may differ from the FAF (prior or after the FAF). Approach charts provide the appropriate descent location, altitude and flight path angle.

In the AC FCTM but not in the SOP :

‐ Crosscheck distances versus altitudes as published on the charts.

‐ If approaching on a conventional radio NAVAID procedure, monitor the lateral and vertical guidance using raw data.

What I noted from the accident report is that no mention is made about serviceable GPS.

If the aircraft did not have GPS, what accuracy did the crew have on their FMS when they began their approach ? The accuracy is supposed to be checked in such cases. The FCOM even has a table on how to set up the cockpit according to NAV accuracy checks.

Did they manually tune in the ZHZ (Split Crow) NDB in their NAV page ?

Did they tune in the IHZ DME in their NAV page in order to get DME readings ?

There are three ways to confirm passing Split Crow: Monitoring the NDB, 6.3 DME from IHZ, or using the FMS.

The report is mum on the subject........

How do they know when they passed the Split Crow NDB to begin their descent ? With FMS data only ?
Was that FMS data reliable ?

[rookiepilot]

One important factor is the lack of approach capability still. Why isn't AC's entire fleet LPV capable? Money it's as simple as that, companies all over the north are investing in full glass LPV capable avionics yet AC doesn't want to spend a few bucks to make their fleet LPV capable.

But the new paint looks awesome.

Yet in court they will dispute any liability and stick it on airbus. Didn't know it was Airbus's responsibility to make sure their commercial planes were properly equipped with what is now the standard in light Cessna's in GPS equipment.

[ZBBYLW]

A few quotes from the report :

The occurrence aircraft was equipped with the appropriate navigational aids to conduct a LOC non-precision approach, and these aids were serviceable at the time of the accident.

The Air Canada FCTM indicates that the flight crew should monitor the aircraft's position along the vertical flight path while conducting a non-precision approach, by referring to the distance indicated on the distance measuring equipment, the altimeter reading, and the time.

The SOPs state that the appropriate raw data on the navigational display should be selected for monitoring the tracking and the descent. As well, the flight crew should use the bird in the PFD as the flying reference.

From the Air Canada SOP :

At 0.3 NAUTICAL MILES prior to FAF, or descent point if other than FAF, if vertical guidance selected

FPA................................PF..............SET

Do not use the Vertical Deviation on the PFD.

Set charted FPA to follow the published descent profile, taking into account the minimum altitudes.

Use radio navaid raw data to monitor the lateral navigation during a conventional NPA.

Note: For RNAV (GPS) approaches the descent point may differ from the FAF (prior or after the FAF). Approach charts provide the appropriate descent location, altitude and flight path angle.

In the AC FCTM but not in the SOP :

‐ Crosscheck distances versus altitudes as published on the charts.

‐ If approaching on a conventional radio NAVAID procedure, monitor the lateral and vertical guidance using raw data.

What I noted from the accident report is that no mention is made about serviceable GPS.

If the aircraft did not have GPS, what accuracy did the crew have on their FMS when they began their approach ? The accuracy is supposed to be checked in such cases. The FCOM even has a table on how to set up the cockpit according to NAV accuracy checks.

Did they manually tune in the ZHZ (Split Crow) NDB in their NAV page ?

Did they tune in the IHZ DME in their NAV page in order to get DME readings ?

There are three ways to confirm passing Split Crow: Monitoring the NDB, 6.3 DME from IHZ, or using the FMS.

The report is mum on the subject........

How do they know when they passed the Split Crow NDB to begin their descent ? With FMS data only ?
Was that FMS data reliable ?

Without GPS they would have likely used the DME of IHZ. Indeed we start the decent 0.3 prior to the decent point (in this case 6.6 DME). This is to take into account the time it takes the airplane to change from FPA of 0.0 (which you would be approaching the decent point) to -3.5 in this case. It indeed is interesting that the profile view of the approach indicates that where the decent started it was already below the profile and remained below the entire way down with the several spots where the aircraft was disturbed onto a lower profile as well.

Prior to TOD they would have done a Nav Accuracy Check (for both FMS) in which they would have found a VOR and compared with actual radial/dme to what the aircraft FMS calculated position thinks would be the track and distance to that station. Sometimes when you're close (within 10 NM) and at high altitude the distance will be off a bit because the DME is obviously taking into account the altitude as well. At their in range checklist they would confirm that the FMS still has high nav accuracy. Seeing how this was not brought up in the report I will assume they did it correctly. Based on the report, sadly it does seem they followed the SOPs to the best of their abilities and still ended up where they did.

Why in Canada we fly NPA approaches to below the advisory vis is a big question. The amount of flow charts and stupid sh!t we have to do now to find if we are legal, based on what kind of lights, approach, level of service etc. is a treat itself. Should be time to fly approaches to the advisory vis and thats it.

[Gilles Hudicourt]

This is a direct quote from the TSB report:

The SOPs state that the appropriate raw data on the navigational display should be selected for monitoring the tracking and the descent. As well, the flight crew should use the bird in the PFD as the flying reference.

So the AC SOPs at the time did require that the crew use DME (raw data) for monitoring the descent ? But no direct quote from that part of the SOPs is provided.

But then later, they state that in a specific part of the SOPs, which they quote, the AC SOPs differed from the AC FCTM and the FCOM, where this sentence had been deleted.....

If approaching on a conventional radio NAVAID procedure, monitor the lateral and vertical guidance using raw data.

[BTD]

Complex,

I agree with the majority of your post. By assuming the role of PIC you are granted certain authority. With that there is also the responsibility. That includes where the aircraft comes to rest at the end of the flight. There can be no doubt about that. We are all responsible as PIC.

As far as your method of doing a SCDA approach. This wasn't an option to this crew. It isn't included in the SOPs, and wasn't trained for at any time. In addition, there have been accidents while completing a SCDA using V/S. As you said it comes down to situational awareness. Regardless of what method is used.

Yet I don't think anyone has remotely implied the pilots were "cowboys".

My comment there was regarding the below quotes:

just seeing an approach light doesn't mean you can just continue descending into the ground hoping the runway will appear beneath you.

despite airlines having 'no fault go around' policies there are still pilots trying to 'get in' even at the major airline level

Good defense of that accident though, they were not sure so they just touched down to see if it was a runway or not.

This is either painting the crew as cowboys or is applying a sarcasm that distorts the original intent of the post.

[Cat Driver]

"swiss cheese modelling"

Which of course is just another way to try and avoid the fact the airplane was landed over a thousand feet short of the normal touch down point on the runway.

In other words it is word salading horseshit.

[tailgunner]

Cat,
How many different aircraft types have you been certified to fly?
You come across as an expert on every aircraft out there.
Maybe that is your intent, or maybe I am being hyper sensitive.
Cheers.

[rookiepilot]

Copied from PPrune. Hmmmmm....

"If an Asian carrier does it , poor standards, low hour pilots , not enough time in the bush. When AC does it , let's blame airbus for screwing up an NPA"

[Cat Driver]

Cat,
How many different aircraft types have you been certified to fly?
You come across as an expert on every aircraft out there.

How many aircraft types does one have to be certified on to be expert on to make a comment on someone flying a commercial jet into the ground during a landing approach?

I got my first type rating on a DC3 at Austin Airways in 1968 and flew that type for over five thousand hours about half that time in the far north and Arctic in every kind of weather known to man and never landed off the landing area or runway.

But to ease your stress a little bit I have over thirty thousand hours in over fifty different countries around the world flying how ever many types I flew, all accident and contravention of the law violations free.

Is that enough experience for me to make comments about safety in your opinion or should I get more experience to have enough credibility to satisfy you?

P.S.

I have flown thousands of IFR approaches to landings on the snow, ice , sand, gravel, pavement etc.. in airplanes a lot worse equipped than a Airbus using far less approach aids never ever missed the landing area....

.....OH I forgot, I was in the employ of Airbus for several years and was given flight training the A320 Sim at the factory in Toulouse by about half a dozen different Airbus instructors.

Does that help?

[ditar]

Does that help?

You should have taken a course on humility while you were at it. If pilots of your generation never succumbed to human error and CFIT was merely a recent phenomenon, perhaps you'd be justified in all your pontificating.

Nobody goes to work intending to have an accident, and I believe the majority make the best decisions they can with the information they have at the time. In spite of that, accidents still happen because people screw up. Just because you have been accident free for 50 years doesn't mean that you are immune to the possibility.

The laws of probability dictate that any event, no matter how improbable, will eventually occur given enough trials. If you claim different of yourself, or that the probability is zero, then your skills must transcend the limitations of the Universe itself.