Halifax crash report coming Thursday

[Eric Janson]

This is exactly it. FPA is a lot like using V/S in terms of "roughing it" down to MDA. You still have to pay attention to MDA (or DDA), MAP fix distance, and VDP. I have a heavy type with a FPA mode, and it's a sore spot for crews because some tend to rely on it blindly. Ours was based mostly upon groundspeed, which essentially makes it a glorified V/S. Our AOM called for setting FPA 0.3 NM prior to FAF or FAWP. Our company had safety reports of crews trying to fly it in completely void of a legitimate profile and references to continue down. The SCDA thing going on down in the States (some call it CDFA) is promoting this behaviour. The approaches are NPA but set up as a precision profile. Most operators have the missed altitude preset, so now there is nothing to catch you except callouts. Plus, on planes with RA only and no BARO computer callouts, you have to ignore the "minimums" automated call. Bad stuff...just to dispatch an aircraft.

Using FPA is a huge improvement over V/S mode imho.

Almost half the approaches I do are NPAs. These are done at night with a large 4 Engine jet.

If flown properly FPA allows for flying a very accurate vertical profile. The problem may be that people don't see NPAs outside the Simulator and don't practice them in VFR conditions to gain confidence in the system.

All NPAs are becoming continuous descent approaches - flying level at MDA is prohibited at my company. My company adds 40' to all MDAs to cover the go-around.

Continuous descent on the correct profile is far safer than making large corrections close to the ground imho.

If NPAs are not being flown correctly then that is a training issue imho. The manufacturers procedures are there for a reason.

We can set our minimums referenced to Barometric altitude or Radio Altimeter (Cat 2/3 only).

You should always be crosschecking your vertical position on every approach you fly.

[Ki-ll]

Normally, I'd agree, but the crew crashed a plane. MDA is MDA, and it seems that they broke it. Having the context of the audio is powerful and should shed a lot of light on exactly how this went down. They know it would be damning (as it should be), hence the resistance. I support my fellow airline pilots, but if you're out there busting mins in an airliner full of people, it should be game over for you.

Have you read the report? Do you have doubts in what TSB found to be the cause?

[Cliff Jumper]

You should always be crosschecking your vertical position on every approach you fly.

I completely agree with your point, but it might not be valid here.

Have a look at the alt/distance chart in figure 1 in the report, and then compare it to the cross-check altitudes on the plate in the appendix. If they had in fact been cross-checking, at which point would they have released that they were low?

Not until they were 0.7nm final.

Too late, in my opinion. Below MDA, and much later than I'd want the PNF cross checking an altitude chart.

[Eric Janson]

You should always be crosschecking your vertical position on every approach you fly.

I completely agree with your point, but it might not be valid here.

Have a look at the alt/distance chart in figure 1 in the report, and then compare it to the cross-check altitudes on the plate in the appendix. If they had in fact been cross-checking, at which point would they have released that they were low?

Not until they were 0.7nm final.

Too late, in my opinion. Below MDA, and much later than I'd want the PNF cross checking an altitude chart.

I see 5 different points where altitude vs DME can be checked after passing the FAF (the last one being MDA). Crosschecking these would have shown the aircraft was below the profile. It wasn't part of the SOP according to the report.

[Cliff Jumper]

Agreed, there are 5 different points. Look at the altitudes and distances on figure 1. The aircraft was above the slope on the first four points, and didn't fall below until the last one, which as you mention, is past the MDA.

So, if they had cross checked (which they didn't) they would have realized that they were a little high, but correcting. Until they got to 0.7 nm final, at about 200 feet. The idea that you would cross check that last point, once you're already past the MDA, and inside of a mile final, is a little absurd in my opinion.

Maybe that's just me however? Is your airline cross checking altitudes on NPA's, inside of a mile final?

[Eric Janson]

Agreed, there are 5 different points. Look at the altitudes and distances on figure 1. The aircraft was above the slope on the first four points, and didn't fall below until the last one, which as you mention, is past the MDA.

So, if they had cross checked (which they didn't) they would have realized that they were a little high, but correcting. Until they got to 0.7 nm final, at about 200 feet. The idea that you would cross check that last point, once you're already past the MDA, and inside of a mile final, is a little absurd in my opinion.

Maybe that's just me however? Is your airline cross checking altitudes on NPA's, inside of a mile final?

Not sure what you are looking at - the aircraft was low passing the FAF and remained low (diverging from the correct profile). That's page 5 of the report.

At MDA (the last listed point on the vertical profile published on the approach plate) you are supposed to be visual with the runway and continue visually. You use the PAPI to keep the correct vertical profile.

My company uses charted visibility as the minimum to be able to fly this approach - as I've stated earlier we would not have been able to fly this approach on the night in question.

[C.W.E.]

At MDA (the last listed point on the vertical profile published on the approach plate) you are supposed to be visual with the runway and continue visually. You use the PAPI to keep the correct vertical profile.

And has been thus forever.

[ReserveTank]

Using FPA is a huge improvement over V/S mode imho.

Almost half the approaches I do are NPAs. These are done at night with a large 4 Engine jet.

If flown properly FPA allows for flying a very accurate vertical profile. The problem may be that people don't see NPAs outside the Simulator and don't practice them in VFR conditions to gain confidence in the system.

All NPAs are becoming continuous descent approaches - flying level at MDA is prohibited at my company. My company adds 40' to all MDAs to cover the go-around.

Continuous descent on the correct profile is far safer than making large corrections close to the ground imho.

If NPAs are not being flown correctly then that is a training issue imho. The manufacturers procedures are there for a reason.

We can set our minimums referenced to Barometric altitude or Radio Altimeter (Cat 2/3 only).

You should always be crosschecking your vertical position on every approach you fly.

I agree. Flown correctly, FPA can produce a good profile. However, I'm speaking from the point of view of a notorious 3-holer and not an Airbus. The data acquisition may be different, IDK. But if it is flown poorly, you'll hit MDA (or derived decision altitude, depending on if your company adds a go-around factor) at the wrong position along the approach. In the case of not having the required runway references in sight, the crew must abandon the approach. While it could be true that there isn't a crosscheck in the SOP language, it's a red herring issue because there was a descent below MDA without the required references having been visually acquired. This is basic IFR air law.

[GoinVertical]

Some people need to go read the TSB report.

They hit MDA 0.3 NM before where they should have on slope. That's 1800 feet or so short. Keeping on that slope has them aiming for 800' before the threshold.

They had, or thought they had, the runway environment (lights) in sight at MDA. They continued the approach.

There's nothing in the CARs saying you need to have the PAPIs in sight to descend below MDA. (But maybe there should be if you don't have any other sort of vertical guidance.)

Just ODALs at that visibility and their approach speed didn't give them enough time to judge just how low they were. Maybe if they were on full strength they would have been - but they weren't.

Staying on slope would have probably resulted in a successful landing. (It surprised me that the Airbus can't give an advisory slope on a non-precision approach...)

Brighter and better approach lighting system may have given the ability to judge how low they were earlier enough to correct and have a successful landing or at least a successful go - around.

An approach ban and advisory visibility that took into account the approach lighting system installed would have not had them landing in the first place.

Air Canada and its Express carriers have all modified their approach bans, and it takes the type of ALS into account. As far as I'm concerned, TC should implement something similar for everyone.

[C.W.E.]

They had, or thought they had, the runway environment (lights) in sight at MDA. They continued the approach.

If they were continuing visual why did they fly into the ground so far short of the runway?

There's nothing in the CARs saying you need to have the PAPIs in sight to descend below MDA. (But maybe there should be if you don't have any other sort of vertical guidance.)

No, however you should have the runway or runway lights in sight clearly enough to fly the remainder of the approach safely at all times at and below MDA.

They obviously didn't.

That is just basic flying skills.

The fact there were two of them supposedly flying it makes it really difficult to excuse the accident in my opinion.

[rookiepilot]

Some people need to go read the TSB report.

They hit MDA 0.3 NM before where they should have on slope. That's 1800 feet or so short. Keeping on that slope has them aiming for 800' before the threshold.

They had, or thought they had, the runway environment (lights) in sight at MDA. They continued the approach.

There's nothing in the CARs saying you need to have the PAPIs in sight to descend below MDA. (But maybe there should be if you don't have any other sort of vertical guidance.)

Just ODALs at that visibility and their approach speed didn't give them enough time to judge just how low they were. Maybe if they were on full strength they would have been - but they weren't.

Staying on slope would have probably resulted in a successful landing. (It surprised me that the Airbus can't give an advisory slope on a non-precision approach...)

Brighter and better approach lighting system may have given the ability to judge how low they were earlier enough to correct and have a successful landing or at least a successful go - around.

An approach ban and advisory visibility that took into account the approach lighting system installed would have not had them landing in the first place.

Air Canada and its Express carriers have all modified their approach bans, and it takes the type of ALS into account. As far as I'm concerned, TC should implement something similar for everyone.

Brutal spinning.......Trudeau could take lessons...

[C.W.E.]

Brutal spinning.......Trudeau could take lessons...

Yeh, with that mindset making excuses for airline pilots wrecking perfectly serviceable airplanes while flying them visually I would rather drive than fly with them.

[GoinVertical]

They had, or thought they had, the runway environment (lights) in sight at MDA. They continued the approach.

If they were continuing visual why did they fly into the ground so far short of the runway?

Because they didn't know they weren't on slope. They had some lights in sight, but the lights seen at minimums in that visibility weren't enough to provide adequate depth perception to allow them to realize they were too low until they did, at which point it was obviously too late.

That's what's in the report.

There's nothing in the CARs saying you need to have the PAPIs in sight to descend below MDA. (But maybe there should be if you don't have any other sort of vertical guidance.)

No, however you should have the runway or runway lights in sight clearly enough to fly the remainder of the approach safely at all times at and below MDA.

They obviously didn't.

That is just basic flying skills.

The fact there were two of them supposedly flying it makes it really difficult to excuse the accident in my opinion.

I'm not disagreeing with you, I'm just laying out what the TSB report says.

In the CAP Gen there is a whole slew of things you need one of to continue below minimums. Most of them alone won't make for a pretty approach if you don't get something else in sight rather quickly.

If you are on slope, you have some approach lights in sight, so you continue "visually". If you continue on that slope you will eventually get more and more things in sight and (probably) have a successful landing.

They weren't on slope. They thought they were. They didn't get anything else in sight for them to realize this until it was too late.

Maybe if they had decided to go around at 0.7NM final instead of apparently having a discussion and agreeing that they were seeing the approach lights things would have gone differently.

Again, I am not trying to "spin" anything or defend anyone. I'm just rephrasing what the TSB report says.

Continuation bias is a real thing. When you've been successfully doing non-precision approaches on the same airplane, the same way, for years you expect to see a normal sight picture when you continue below minimums with a couple approach lights in sight.

I bet if they had done a step down, they would not have descended below MDA with only a light or two in sight - because they wouldn't have had confidence what they were seeing, or if they were high or low.

Pilots of aircraft on
instrument approaches are prohibited from continuing the descent below DH, or descending below
MDA, as applicable, unless the required visual reference is established and maintained in order
to complete a safe landing.

They then go on to say that "the approach lights" (and apparently on their own) is required visual reference. If you see one ODAL at a time until you see the runway threshold lights, that's enough to be "legal", apparently.

Like you say, what you "should" have is a different story. Especially when it's an approach without vertical guidance. But again, pilots have the expectation of getting more as they go, so they continue.

When do you go around? What do you need to see? Is it different for a non-precision approach than an ILS? Should it be?

[C.W.E.]

When do you go around? What do you need to see?

I go around when I do not have sufficient visual contact to see the landing surface I want to land on clear enough to safely judge my position from.

I was doing non precision ADF approaches flying a transport category airplane for an airline in the late sixties frequently landing on off airport sites and our limits were 400 feet and one mile.

Did it for decades accident and incident free.

The secret is simple, don't fly outside of the approach limits, know your position at all times and think ahead of where you are.

[pelmet]

I believe the PAPIs were covered in snow. I suspect that the depth perception was minimal but that there were lights in sight so they continued with the required visual references in sight. But while the CAP lists the required visual references, I believe it may be hidden in the CARS where it talks about there being a requirement for depth perception as well(perhaps someone can find it).

On an ILS, one will typically continue down with minimal lighting in sight and just follow the GS. One has to wonder if this is what was done on the vertical indication of the FPA. Airbus mentions deep within the FCTM about the possibility of external perturbations moving the aircraft away from the original calculated vertical path and establishing a new path parallel to the original path. Unfortunately, there is no further discussion from the TSB in the report on whether the pilots were aware of this or what they were really using to make them feel that they were continuing on a proper descent path(eyesight, FPA path, both). Maybe the CVR transcript will help.

[Old fella]

I believe the PAPIs were covered in snow. I suspect that the depth perception was minimal but that there were lights in sight so they continued with the required visual references in sight. But while the CAP lists the required visual references, I believe it may be hidden in the CARS where it talks about there being a requirement for depth perception as well(perhaps someone can find it).

On an ILS, one will typically continue down with minimal lighting in sight and just follow the GS. One has to wonder if this is what was done on the vertical indication of the FPA. Airbus mentions deep within the FCTM about the possibility of external perturbations moving the aircraft away from the original calculated vertical path and establishing a new path parallel to the original path. Unfortunately, there is no further discussion from the TSB in the report on whether the pilots were aware of this or what they were really using to make them feel that they were continuing on a proper descent path(eyesight, FPA path, both). Maybe the CVR transcript will help.

Specific requests for CVR transcripts on this accident are for litigation purposes only and the tort lawyer(s) want to assign blame in order to reach a monetary settlement for their clients against AC and whoever else they deem responsible albeit unproven at this point. The TSB Final report is the official public document on the findings as to cause and contributing factors, nothing changes that.

[Cliff Jumper]

what they were really using to make them feel that they were continuing on a proper descent path(eyesight, FPA path, both)

Well, the A320 doesn't have a visual depiction of the select FPA, nor a depiction of your proximity to the originally selected path/angle. So, you have to cross check,,,Like Eric mentioned.

The point I was trying to make, still am, was that in this case, even if they were cross-checking every 3 seconds the whole way down, they wouldn't have discovered that the were below the VDA until they were already past the MDA.

Not sure what you are looking at - the aircraft was low passing the FAF and remained low (diverging from the correct profile)

Disagree. You should look again. Figure 1 is their actual profile vs their selected profile (3.5).... NOT the VDA (3.08). They were above the VDA almost the whole time until the end.

[Eric Janson]

what they were really using to make them feel that they were continuing on a proper descent path(eyesight, FPA path, both)

Well, the A320 doesn't have a visual depiction of the select FPA, nor a depiction of your proximity to the originally selected path/angle. So, you have to cross check,,,Like Eric mentioned.

The point I was trying to make, still am, was that in this case, even if they were cross-checking every 3 seconds the whole way down, they wouldn't have discovered that the were below the VDA until they were already past the MDA.

Not sure what you are looking at - the aircraft was low passing the FAF and remained low (diverging from the correct profile)

Disagree. You should look again. Figure 1 is their actual profile vs their selected profile (3.5).... NOT the VDA (3.08). They were above the VDA almost the whole time until the end.

The selected profile (-3.5) is the VDA corrected for low temperature.

From the report:-

The flight crew did not notice that the aircraft had drifted below and diverged from the planned VDA flight profile, nor were they aware that the aircraft had crossed the minimum descent altitude further back from the threshold.

Hopefully this clarifies things.

[pelmet]

what they were really using to make them feel that they were continuing on a proper descent path(eyesight, FPA path, both)

Well, the A320 doesn't have a visual depiction of the select FPA, nor a depiction of your proximity to the originally selected path/angle. So, you have to cross check,,,Like Eric mentioned.

The point I was trying to make, still am, was that in this case, even if they were cross-checking every 3 seconds the whole way down, they wouldn't have discovered that the were below the VDA until they were already past the MDA.

Not sure what you are looking at - the aircraft was low passing the FAF and remained low (diverging from the correct profile)

Disagree. You should look again. Figure 1 is their actual profile vs their selected profile (3.5).... NOT the VDA (3.08). They were above the VDA almost the whole time until the end.

The selected profile (-3.5) is the VDA corrected for low temperature.

From the report:-

The flight crew did not notice that the aircraft had drifted below and diverged from the planned VDA flight profile, nor were they aware that the aircraft had crossed the minimum descent altitude further back from the threshold.

Hopefully this clarifies things.

I am not sure it does. When you are on a visual segment, you may be using visual cues only, instrument cues only, or something in between to maintain what you feel is the appropriate descent path. Even on a clear night with an ILS, typically, one will use both the glideslope and the view out the window, even if it is just to ensure that there is no nastygram from the company at a later date asking why the approach was technically unstable due to glideslope deviation. On a VNAV approach, to use Boeing terminology, one will still typically follow the vertical path indication that takes you 50 feet above the threshold. But FPA seems to be subject to errors.

It appears that no distance vs. altitude measurement was done as per their procedures that were approved by Transport Canada. I don't think the report specifically says what they were using to confirm in their minds that they were on what they thought was the appropriate trajectory.

[Cliff Jumper]

The selected profile (-3.5) is the VDA corrected for low temperature.

From the report:-

The flight crew did not notice that the aircraft had drifted below and diverged from the planned VDA flight profile, nor were they aware that the aircraft had crossed the minimum descent altitude further back from the threshold.

Hopefully this clarifies things.

I think you just supported my point. The crew planned 3.5, the chart was 3.08. Once 3.5 was planned, they had no way to cross-check on the way down, because the altitudes vs distances were totally different from those published on the chart.

Again, if you compare the alt vs dis on figure 1, to those published on the chart, you'll see they were above the published VDA the whole time, until 0.7nm final.

[Cliff Jumper]

The selected profile (-3.5) is the VDA corrected for low temperature.

From the report:-

The flight crew did not notice that the aircraft had drifted below and diverged from the planned VDA flight profile, nor were they aware that the aircraft had crossed the minimum descent altitude further back from the threshold.

Hopefully this clarifies things.

I think you just supported my point. The crew planned 3.5, the chart was 3.08. Once 3.5 was planned, they had no easy way to cross-check on the way down, because the altitudes vs distances were totally different from those published on the chart.

Again, if you compare the alt vs dis on figure 1, to those published on the chart, you'll see they were above the published VDA the whole time, until 0.7nm final.

[complexintentions]

I think you just supported my point. The crew planned 3.5, the chart was 3.08. Once 3.5 was planned, they had no easy way to cross-check on the way down, because the altitudes vs distances were totally different from those published on the chart.

Again, if you compare the alt vs dis on figure 1, to those published on the chart, you'll see they were above the published VDA the whole time, until 0.7nm final.

Totally missing the big picture.

Fine. So the crew had no way of making distance vs altitude checks on this approach. How exactly does this contradict Eric's point that these crosschecks are absolutely crucial on a NP approach? Hint: it doesn't, in fact it only underscores the vulnerability of the accident flight once they started their descent relying solely on a calculated path with no backup method of verifying their position. Your observation is actually more damning of the crew, not exculpatory.

I don't think the report specifically says what they were using to confirm in their minds that they were on what they thought was the appropriate trajectory.

I don't think they used anything whatsoever. The thought makes my palms sweaty actually.

Pushing a non-precision approach at an ill-equipped airport, in an ill-equipped aircraft, at night, between reports of heavy snow and low viz...the red threat flags would be going bezerk in my mind. Ultra high-alert mode, spring-loaded to GA and divert, and certainly not going to do an NP approach without some sort of backup check for the descent path. Since when do lack of formal company SOP's prohibit the use of practical airmanship?! FFS

A lot of words to say that ultimately, you have to be totally confident in the position of your aircraft before continuing to land.

From the report: "When the aircraft was about 0.7 nm from the threshold, the flight crew had a conversation in which both confirmed they could see some approach lights."

Did they?

I'd say not.

[goingnowherefast]

Doesn't the report say the crew likely misinterpreted some commercial lighting as the approach lights? So yes they probably thought they had approach lights in sight and waited for the PAPIs and runway to come into view. Perfectly legal, except the "approach lights" they saw weren't the actual approach lights and the runway never came into sight.

The approach ban rules are garage, the FPA non-precision procedure is garbage, continuation biases definitely played a part.

CVR and FDR are not to decide fault. The lawyers and the judge are undermining future investigations. I'll stop saying what I'm thinking, promotes safety and CRM. I'll only say what I want the lawyers to hear.

[Cliff Jumper]

Totally missing the big picture.

Fine. So the crew had no way of making distance vs altitude checks on this approach. How exactly does this contradict Eric's point that these crosschecks are absolutely crucial on a NP approach? Hint: it doesn't, in fact it only underscores the vulnerability of the accident flight once they started their descent relying solely on a calculated path with no backup method of verifying their position. Your observation is actually more damning of the crew, not exculpatory.

Complex, I never set out to damn or exculpate. I am simply pointing out a fact.

This crew had no method to cross check their altitude vs their planned FPA/VDA. That's my point, no more, no less. Zero comment on any decision or action they took or didn't take.

I don't feel like the fact that they didn't have this available was 'damning' to the crew. Or, are you suggesting that they should have pulled out the calculators and created a Alt/Dis crosscheck chart from scratch, using 3.5 degrees?

[ReserveTank]

Doesn't matter which angle or how far back, really. They blew through the MDA without required visual references. "Thought they saw approach lights" is not acquisition of the reference to continue. PAPIs covered in snow make no difference. You have it or you don't. They didn't have it, and they crushed a perfectly good airplane.