Air Canada A320 accident at Halifax

[peekay4]

GPS/WAAS equipment cost is just a fraction of the total cost. What's expensive is the actual cost to retrofit each and every individual aircraft. It's not like you one simply unplug old equipment and swap in new ones.

You need a team of engineers a full week per aircraft at a minimum. If you stick to the OEM providers (Honeywell, etc), just the labor costs alone would be in the $ hundreds of thousands per aircraft. Then you have to add the cost of flight testing, cost to upgrade the simulators, cost to re-train all the pilots (and update all the training materials, manuals, job aids, etc), cost to train the mechanics, cost to provision spares, project management costs...

Plus it might not make sense to just upgrade to GPS. You'd probably package the upgrade to also add ADS-B, maybe CPDLC, even a HUD. Perhaps even more exhaustive upgrades are justified for older aircraft (older A320s still have CRTs).

Pretty soon you're easily past $1 million per aircraft...

[ASRAAM]

I might have missed it but no one seems to have mentioned the flight directors. I would assume that during the instrument approach the autopilot was in use, either as a fully managed approach or in nav/FPA. Once visual the FDs should have been turned off. It's worth giving some thought as to what would have happened if they had been left on. For example if an FPA of 3 degrees was set but not followed because the crew were now visual and aiming at the runway. The situation would be somewhat similar to the Korean 777, except of course for all those airbus protections..... Until 100RA. Thoughts anyone?

[RESA]

Halifax A320 / Localiser bem/burm

I suspect that there was a major error in the execution of the approach . . . I just wonder why nobody questions why the entire undercarriage was left behind at the “frangible localiser array”? I don’t recall a snow-bank ever eviscerating an aircraft . . . totally removing the under carriage?

What’s under that snow . . . not just dirt ?

[Mudman]

If you look at the google street view of the area the ground slopes quite steeply below the array. The array sits on a concrete pad. I'd expect that below the dirt and gravel is a substantial concrete foundation. Given most footings and foundation walls around here have to be at least 3'- 4' deep to be below the frost line to prevent upheaval I imagine that is what took the gear off. Also being winter the ground was frozen solid almost as hard as concrete.

https://www.google.ca/maps/@44.86370...NV5w636Ziw!2e0

This Transport Canada Photo shows how far in front of the array the aircraft made first contact. Well below the level of the array. Anyone who has backed into a snowbank or slid of the road into one knows that snow and ice can be surprisingly unyielding.

[L39 Guy]

The WAAS upgrade is a matter of upgrading the chip in the receiver (to do a 5 Hz update rate rather than a 1 Hz), the antenna (using the same cabling) and the software. All that is being done is the GPS sensor in a multi-sensor MMR is being upgraded to incorporate the WAAS correction which, by the way, is on the same frequency as the USAF/GPS satellites; Galileo, incidentally, uses the frequency and is interoperable with GPS/Glonass/WAAS.

Doing upgrades for other items such as CPDLC, etc is beside the point. It is important to note that in order to be ADS-B compliant after 2020 in the US, a WAAS position will be required so, unless the air carriers snivel for an exemption, they will be upgrading to WAAS anyways.

With WAAS grade receivers, precision approaches to virtually every runway in North America becomes possible and with Galileo every runway in the world.

Compare this alternative with upgrading MMR's for Local Area Augmenation System (LAAS) which will require VHF receivers, software, certification, etc and at least $3 million per airport for the LAAS box to get the same results as WAAS/Galileo for airline only type aircraft and the choice is clear.

[scud]

GPS/WAAS equipment cost is just a fraction of the total cost. What's expensive is the actual cost to retrofit each and every individual aircraft. It's not like you one simply unplug old equipment and swap in new ones.

You have to include the costs to the airport infrastructure, such as High intensity runway and approach lighting, centre-line lighting, not counting imbedding all these in-pavement and such issues such as CAT II and CAT III hold short lines, especially if taxiways have to be redesigned for that. Plus, there has to be increased lighting for taxiways, such as the 'greens' centreline taxiways. And for airports that have never had CAT II/III systems, an huge stand-by generator to power the whole thing. Then, once an airport has imbedded runway and taxiway lighting, snowblough blades have to be converted from steel to plastic, so as to not rip them out during snow clearing operations.

In the case of Halifax, the cost to the airport authority for the improved infrastructure to accommodate a CAT I for 06 was similar to the cost of the ILS itself to Nav Canada.

[aterpster]

L39 Guy:

With WAAS grade receivers, precision approaches to virtually every runway in North America becomes possible and with Galileo every runway in the world.

Not exactly. That's why the U.S. has a lot of LP approaches now, and more all the time. Some runways don't have the clear zones for a vertically guided approach (APV in FAA-speak).

And, ICAO has yet to anoint LPV as a precision approach.

[aterpster]

scud:

The U.S. now has many LPV approaches down to 200 and 1 with only edge lights (which are usually there anyway) and precision runway markings.

[peekay4]

L39 Guy:

The WAAS upgrade is a matter of upgrading the chip in the receiver (to do a 5 Hz update rate rather than a 1 Hz), the antenna (using the same cabling) and the software.

That's assuming WAAS upgrade is the only thing that's required. Which is the wrong assumption in most cases.

In reality, a large number of A320s (and even A330s/A340s) do not have even vanilla GPS installed as part of the navigation system! No GPS units. No MMR. These A320s depend solely on radio updating (VOR/DME) or manual IRU alignment before takeoff.

(I believe the Air Canada aircraft involved in this accident was not equipped with GPS.)

Also, some of these aircraft have dual FMS which are not GPS compatible. Total upgrade price == very costly.

[L39 Guy]

My, this is a tough audience. Allow me to clarify a few points in order of responses:

1. Scud: for a non-precision approach (i.e. minimums as low as 250 ft) there is no requirement of sophisticated lighting, hold short lines, etc. For 99% of the runways out there particularly at airports with an existing CAT II or CAT III ILS, LPV can provide approach limits to 250 ft with the only improved airport infracture is insuring the Glideslope Qualification Surface (GQS) - the region between DA and the runway is free of obstacles above a protected surface (I won't burden you with the calculation of this surface but it is roughly a 2 degree slope for a 3 degree Flight Path Angle (FPA). In Halifax's case, all four runway ends have LPV approaches with limits between 250 and 259 ft (DA value rounded up to the next highest 10 ft - TDZE). Neither Runway 05 or 32 have fancy lighting. The problem in this case is that the AC A320's don't have GPS and even if they did, they and virtually every other transport category aircraft (Airbus, Boeing, etc) don't have WAAS making these procedures unable.
2. aterpter: agreed about the strict ICAO definition precludes LPV from being called that. Notwithstanding that, however, for all intents and purposes LPV is a precision approach for the operator - limits as low as 200 ft, 6 sec time to alert, "glideslope-like" vertical guidance (better than the ILS in my opinion). For the pilot it looks and feels just like an ILS despite what ICAO might say. As far as the clear zones, see my comment above about GQS and I agree - we run into this issue in Canada too.
3. peekay4: the origin of my comments was to state that RNP/GLS is an inferior option from an economic, operational and procedure design perspective compared to RNAV with WAAS/LPV/APV. It's great business for Honeywell who would dearly love to install a GLS box at every airport and charge a grazillion dollars for avionics upgrades. You are absolutely correct that many FMS's in B737-300/400/500 and Airbus 319/320/330/330/340 do not have GPS hence cannot even do a simple RNAV approach. My point remains that a better option for the reasons above is to have these aircraft equipped with GPS/WAAS rather than GLS, assuming they are going to be upgraded. ADS-B in the US in 2020 is going to drive it for any aircraft that will be flying in that airspace. Having said all of that, it is shameful that a major international airport like Halifax with its location in a valley, with the maritime weather they get and the winds they get do not have a proper ILS to all four runway ends; 250 ft minimums if the proper approach lighting is a far cry better than having to do a LOC/DME approach

[JammedStab]

Originally Posted by peekay4

L39 Guy:

That's assuming WAAS upgrade is the only thing that's required. Which is the wrong assumption in most cases.

In reality, a large number of A320s (and even A330s/A340s) do not have even vanilla GPS installed as part of the navigation system! No GPS units. No MMR. These A320s depend solely on radio updating (VOR/DME) or manual IRU alignment before takeoff.

(I believe the Air Canada aircraft involved in this accident was not equipped with GPS.)

Also, some of these aircraft have dual FMS which are not GPS compatible. Total upgrade price == very costly.

Depending on what we see in the final report, there is a chance that we will get to see the price paid for not making this "costly" upgrade.

The last operator that crashed a jet in Canada were saving themselves money by waiting to upgrade their GPWS to TAWS standard. Several lives lost.

[peekay4]

With their extensive insurance policies, I would guess that AC's out-of-pocket expenses for this incident will be much less than the cost to upgrade their sizable fleet of non-GPS aircraft.

[aterpster]

L39 Guy:

2.aterpter: agreed about the strict ICAO definition precludes LPV from being called that. Notwithstanding that, however, for all intents and purposes LPV is a precision approach for the operator - limits as low as 200 ft, 6 sec time to alert, "glideslope-like" vertical guidance (better than the ILS in my opinion). For the pilot it looks and feels just like an ILS despite what ICAO might say. As far as the clear zones, see my comment above about GQS and I agree - we run into this issue in Canada too.

No doubt that LPV is better than many CAT I ILSs. The LPV G/S is extremely stable, unlike some CAT 1 ILs. But, LPV lacks the integrity to go below 200 feet. Technically I agree. Practically, I think it is all B.S.

Ah, ICAO and the FAA and their PBN stuff.

[Big Pistons Forever]

Seem to me the hull value of an A 320 would pay for a fleet wide GPS WASS sensor upgrade.

What would have been the cost if the jet was just a few feet lower and smoked the berm head on instead of bouncing over it, an event that almost certainly will produce lots of casualties ? Hundreds of Millions of dollars ? Not to mention the damage to reputation and forward bookings.

Sadly stinting on upgrades which will provide an obvious increase in approach precision and thus safety because it requires an upfront investment, has become ops normal for the bean counter, short term profit focused airline management at pretty much every airline in the world.

[peekay4]

Again, Air Canada will not bear the cost of hull replacement. Their insurance will cover that.

American Airlines paid $400 million to upgrade their 757/767 fleet with GPS. That's over $2.5 million per aircraft (120 of them), plus another $100 million in additional costs.

Rightly or wrongly, I'm sure AC's bean counters have thoroughly gone through the cost vs. benefit analysis of an upgrade.

[underfire]

Incidentally the blanket exemption in Canada allowing the use of MDA as DA (without adding 50ft) will expire at the end of this month -- April 30, 2015.

Thankfully.
Most have no concept of the difference, nor how far above MDA their actual DA is (it sure as hell isnt 50 feet, but at least the 50 feet + ROC gives some hope, and hopefully most obstacles are in the missed)

[oblivia]

With their extensive insurance policies, I would guess that AC's out-of-pocket expenses for this incident will be much less than the cost to upgrade their sizable fleet of non-GPS aircraft.

Indeed, AC (and whoever else owns the plane) has probably made a substantial profit on this hull and will have already received the cheque! Airlines insure planes at their full value at the time of purchase, so old birds like this one tend to be worth far less than the policy covering them. And insurers pay out in full within 72 hours of a loss, with policies that cover any and all risks.

BA got paid immediately and in full, for example, when a perfectly serviceable 747 was stranded in Kuwait after the Iraqi invasion. It was the oldest hull in the fleet and earned the company a nice profit. The Iraqis later burned it, long after the cheque had cleared.

The reason for this amazing level of insurance is, of course, the industry's remarkable safety record...

[fox niner]

Reading the statements made above, I would almost start to believe that airlines would actually welcome a hull loss once in a while to boost their profitability.

[oblivia]

In the case of losses where people die, it can actually a bit embarrassing for airlines to report a windfall profit while the victims wait years for the minimum possible payout the insurer can get away with. The MH370 victims' families will probably get far less (in total) than the airline was paid within three days of its disappearance—and they'll likely wait the best part of a decade to find out.

Even so, I don't think any airlines set out to have accidents for insurance purposes, but I'd be happy to hear stories from pilots who've been asked to write off an airframe. How does that conversation go...?

[YRP]

I find it hard to belief that insurance companies would insure for significantly more than the current value. At best they might use a year or two out of date values. Paying a claim for more than the value is betterment and is very undesirable.

Look at it from the insurance cost point of view. If AC insures a plane at twice its value, it has to pay twice the premiums (for hull insurance rather than liability). AC isn't going to want that any more than the insurance company.

Gambling on loss rates is exactly the opposite of the point of insurance.