Air Canada A320 accident at Halifax
Thanks... I had that, but under Safety Action by Air Canada on P 78, 4.1.1 (a) "The “lights only” call has been removed from standard operating procedures." Hence my query. See also RAT 5's post 391, previous page.
Last edited by slast; 9th Jul 2017 at 10:04.
Join Date: Jan 2007
Location: Johannesburg
Posts: 37
Likes: 0
Received 0 Likes
on
0 Posts
I am surprised that there has been no mention of the monitored approach concept. Obviously it is not AC SOP but do other North American operators apply it?
Marginal Wx or any non-precision approach - FO flies, Captain monitors and acquires requisite visual cues approaching minimums, while PF (FO) remains on instruments. At minima, FO/PF calls "decide" and either the Capt has the required visual cues in which case he says "continue" or "landing" and assumes flying control or he says "go around" and the FO executes the go-around.
The benefit is that when landing the Capt already has the visual cues and the PF does not now have to start assessing visual cues after minimums. Conversely if the visual reference is not there, the FO is already on instruments and in a better position to execute the go-around.
It seems from the report that the PM called "lights only" at minimums, which was then followed by further assessment by both pilots as to what they could or could not see - all below minimums. I'm sorry but the decision to continue or go-around must be made at minimums and based on solid visual reference. There should be no decision-making and discussion process happening under minimums.
I also note that no additive to MDA was mentioned. It is fairly standard to add say 50 feet to MDA on a constant descent final approach. The value of 50' is actually in the Boeing documentation for Boeing operators. This allows a go-around to be executed at minimums (MDA + 50') without breaching the MDA - which is a "hard floor".
Lastly, was there perhaps a culture during ILS approaches of continuing despite adequate reference knowing that the airplane will deliver the aircraft correctly to the runway and you will eventually see what you need to see?
Obviously with that mindset during a non-precision approach you will come short (no pun intended) due to the aircraft not "knowing" where it is relative to the glidepath.
Marginal Wx or any non-precision approach - FO flies, Captain monitors and acquires requisite visual cues approaching minimums, while PF (FO) remains on instruments. At minima, FO/PF calls "decide" and either the Capt has the required visual cues in which case he says "continue" or "landing" and assumes flying control or he says "go around" and the FO executes the go-around.
The benefit is that when landing the Capt already has the visual cues and the PF does not now have to start assessing visual cues after minimums. Conversely if the visual reference is not there, the FO is already on instruments and in a better position to execute the go-around.
It seems from the report that the PM called "lights only" at minimums, which was then followed by further assessment by both pilots as to what they could or could not see - all below minimums. I'm sorry but the decision to continue or go-around must be made at minimums and based on solid visual reference. There should be no decision-making and discussion process happening under minimums.
I also note that no additive to MDA was mentioned. It is fairly standard to add say 50 feet to MDA on a constant descent final approach. The value of 50' is actually in the Boeing documentation for Boeing operators. This allows a go-around to be executed at minimums (MDA + 50') without breaching the MDA - which is a "hard floor".
Lastly, was there perhaps a culture during ILS approaches of continuing despite adequate reference knowing that the airplane will deliver the aircraft correctly to the runway and you will eventually see what you need to see?
Obviously with that mindset during a non-precision approach you will come short (no pun intended) due to the aircraft not "knowing" where it is relative to the glidepath.
You're right, Arrow Air was a civil aircraft charter, but all passengers were US military personnel, so I think this would have caused greatest Canadian loss of life. However you look at it, a major catastrophe was only avoided by luck.
PMA and height loss
JC07,
I thoroughly endorse your comments, but have kept away from that subject so far. Some readers will be well aware of my views on it. However, FYI there was an AC accident in similar conditions (snow / low vis / night / minimal lighting ) in which the TSB picked up that point. See pages page 47 and 48 of the report, which is accessible at
1997 CRJ Go-around LoC on rejected landing Fredericton Canada | PicMA
Some years later Transport Canada published an Advisory Circular on the subject, and modified its Approach Ban criteria to allow operators using PMA procedures to have lower visibility limits.
http://www.picma.info/sites/default/...0239%20PMA.pdf
Remarkably, given the references it does make to other accident reports, this report makes no mention of either of these documents, nor of the Canadian Airlines B767 accident on the same runway, also at night in snow and poor visibility, where the runway slope was a significant factor.
As far as the 50' height loss is concerned, that was taken into account. Report page 2: "The MDA was calculated at 813 feet ASL, based on the published MDA of 740 feet ASL plus a cold temperature correction of 23 feet, plus 50 feet added to the corrected MDA, as required by Air Canada’s Flight Operations Manual (FOM)."
813 ft indicated altitude corresponds to an ISA altitude of 790 ft, vs. the true MDA of 740, which gives the required 50 ft height loss allowance . But according to my plot of the altitudes, the F/O call of "minimum, lights only" came after the aircraft had passed the calculated/indicated MDA where a decision should have made, and marginally below the true MDA. The callout takes approximately 1.5 second and the Captain's response of "Landing" was about 20 ft below true MDA, as he started assessing the cues. So there can be no doubt that had the Captain called for a go around instead of landing, it would still have been from well below MDA and obstacle clearance criteria would have been infringed.
I thoroughly endorse your comments, but have kept away from that subject so far. Some readers will be well aware of my views on it. However, FYI there was an AC accident in similar conditions (snow / low vis / night / minimal lighting ) in which the TSB picked up that point. See pages page 47 and 48 of the report, which is accessible at
1997 CRJ Go-around LoC on rejected landing Fredericton Canada | PicMA
Some years later Transport Canada published an Advisory Circular on the subject, and modified its Approach Ban criteria to allow operators using PMA procedures to have lower visibility limits.
http://www.picma.info/sites/default/...0239%20PMA.pdf
Remarkably, given the references it does make to other accident reports, this report makes no mention of either of these documents, nor of the Canadian Airlines B767 accident on the same runway, also at night in snow and poor visibility, where the runway slope was a significant factor.
As far as the 50' height loss is concerned, that was taken into account. Report page 2: "The MDA was calculated at 813 feet ASL, based on the published MDA of 740 feet ASL plus a cold temperature correction of 23 feet, plus 50 feet added to the corrected MDA, as required by Air Canada’s Flight Operations Manual (FOM)."
813 ft indicated altitude corresponds to an ISA altitude of 790 ft, vs. the true MDA of 740, which gives the required 50 ft height loss allowance . But according to my plot of the altitudes, the F/O call of "minimum, lights only" came after the aircraft had passed the calculated/indicated MDA where a decision should have made, and marginally below the true MDA. The callout takes approximately 1.5 second and the Captain's response of "Landing" was about 20 ft below true MDA, as he started assessing the cues. So there can be no doubt that had the Captain called for a go around instead of landing, it would still have been from well below MDA and obstacle clearance criteria would have been infringed.
Difference between CDA and ILS at Mimimums
On an ILS with the glideslope flown within tolerances, descent past DH is within terrain margins, indeed, all the way to TDZ.
On a CDA with substantial headwind, the aircraft can drift back into a zone where terrain clearance margins are not defined. But in this case, the aircraft ended below the runway. There may well have been an influence from the quite considerable downslope wind, but this is not addressed by TSB.
In the tailwind CDA case the aircraft is blown towards the threshold and airspace with designed in terrain margins.
On a CDA with substantial headwind, the aircraft can drift back into a zone where terrain clearance margins are not defined. But in this case, the aircraft ended below the runway. There may well have been an influence from the quite considerable downslope wind, but this is not addressed by TSB.
In the tailwind CDA case the aircraft is blown towards the threshold and airspace with designed in terrain margins.
Only half a speed-brake
It is CDFA. A stands for angle, so the wind effect should be negligable. Provided we check during the final descent and adjust the flightpath appropriately.
On an ILS with the glideslope flown within tolerances, descent past DH is within terrain margins, indeed, all the way to TDZ.
On a CDA with substantial headwind, the aircraft can drift back into a zone where terrain clearance margins are not defined. But in this case, the aircraft ended below the runway. There may well have been an influence from the quite considerable downslope wind, but this is not addressed by TSB.
In the tailwind CDA case the aircraft is blown towards the threshold and airspace with designed in terrain margins.
On a CDA with substantial headwind, the aircraft can drift back into a zone where terrain clearance margins are not defined. But in this case, the aircraft ended below the runway. There may well have been an influence from the quite considerable downslope wind, but this is not addressed by TSB.
In the tailwind CDA case the aircraft is blown towards the threshold and airspace with designed in terrain margins.
However the explanations given in the report for the developing deviations do not make sense. If you look closely at Fig 1 and Fig 15, the "perturbations" go in opposite directions.
We have a similar system our aeroplane and while it does a good job at maintaining 3°, if it is "bumped off" the nominal 3° path by whatever, it stays "bumped off". You have to put it back on the charted slope. If you don't, you'll end up going in short or long. You must closely monitor the profile and intervene with FPA adjustments to maintaining the charted slope, especially when limiting steps are involved.
My impression is that too many think that these "wonderful" CDFAs are simply "point 3° from the FAF and all will be OK". Not so.
My impression is that too many think that these "wonderful" CDFAs are simply "point 3° from the FAF and all will be OK". Not so.
AC624 Report critique
There is a detailed critique of the accident report available from a link right at the bottom of this page:
2015 A320 snowstorm CFIT Halifax Canada | PicMA
2015 A320 snowstorm CFIT Halifax Canada | PicMA
Why did they then have an ils on 05 prior to the mk airlines accident? I may be mistaken it has been a number of years
We have a similar system our aeroplane and while it does a good job at maintaining 3°, if it is "bumped off" the nominal 3° path by whatever, it stays "bumped off". You have to put it back on the charted slope. If you don't, you'll end up going in short or long. You must closely monitor the profile and intervene with FPA adjustments to maintaining the charted slope, especially when limiting steps are involved.
My impression is that too many think that these "wonderful" CDFAs are simply "point 3° from the FAF and all will be OK". Not so.
My impression is that too many think that these "wonderful" CDFAs are simply "point 3° from the FAF and all will be OK". Not so.
Join Date: Jun 2000
Location: last time I looked I was still here.
Posts: 4,507
Likes: 0
Received 0 Likes
on
0 Posts
Except: it is height above TDZE. And, it is 320'/nm for 3 degrees, and it is GS/2X10+50 for V/S on 3 degrees. The DME v ALT tables on approach charts confirm this. One is a simple rule of thumb, and the other is accurate, but not too difficult.
Originally Posted by Rigpiggy
GS/2x10 ergo 5 nm at 1500 130 knots gs -650 fpm. Very very close easy math
Not so much, because you are supposed to level off at the MDA, and only descend from there with required visual reference.
RNP Y05 .30 is 455 above tdze and the LOC 05 is 357.
Yes you can be more exact, however reading the dme and a quick crosscheck should have shown the error.
RNP Y05 .30 is 455 above tdze and the LOC 05 is 357.
Yes you can be more exact, however reading the dme and a quick crosscheck should have shown the error.
Wrong instrument descent angle wasn't the cause.
An undetected incorrect instrument flight path angle was NOT the cause of the accident, despite the implications of the TSB report.
On any instrument approach the exact position of the aircraft relative to the desired (perfect) flight path will vary. Depending on the degree of precision it could be closer or further away from it.
Cat 3 = very close indeed, Cat 1 precision fairly close, non-precision LOC further away, non precision NDB further still. Minimum instrument flight height (DH/MDA) naturally goes up as accuracy decreases. A Continuous Descent Final Approach NPA as used by the AC crew (whether or not using the autopilot/FPA mode etc.) drastically reduces the position scatter compared to "dive and drive, but use the same MDA.
This accident occurred when the crew descended below that minimum safe height (MDA) without having completed a visual confirmation that the aircraft's actual position and velocity (rate of change of position) showing that they were NOT going in short of the runway.
Such a descent is prohibited under ICAO Annex 6 but IS allowed under Canadian rules because Canada has a much less stringent definition of required visual reference, as already described post 396
When the very idea of DH/MDA was being introduced into ICAO by the FAA, correspondence with the US NTSB made it clear that "We see nothing in the definition of Decision Height which permits the decision to be made after the aircraft has descended below the prescribed height while the Captain attempts to locate the runway." But that is exactly what the Air Canada crew were doing, quite legitimately, under CARs.
For full explanation see this
On any instrument approach the exact position of the aircraft relative to the desired (perfect) flight path will vary. Depending on the degree of precision it could be closer or further away from it.
Cat 3 = very close indeed, Cat 1 precision fairly close, non-precision LOC further away, non precision NDB further still. Minimum instrument flight height (DH/MDA) naturally goes up as accuracy decreases. A Continuous Descent Final Approach NPA as used by the AC crew (whether or not using the autopilot/FPA mode etc.) drastically reduces the position scatter compared to "dive and drive, but use the same MDA.
This accident occurred when the crew descended below that minimum safe height (MDA) without having completed a visual confirmation that the aircraft's actual position and velocity (rate of change of position) showing that they were NOT going in short of the runway.
Such a descent is prohibited under ICAO Annex 6 but IS allowed under Canadian rules because Canada has a much less stringent definition of required visual reference, as already described post 396
When the very idea of DH/MDA was being introduced into ICAO by the FAA, correspondence with the US NTSB made it clear that "We see nothing in the definition of Decision Height which permits the decision to be made after the aircraft has descended below the prescribed height while the Captain attempts to locate the runway." But that is exactly what the Air Canada crew were doing, quite legitimately, under CARs.
For full explanation see this