Perimeter Sanny accident report

[Heliian]

Sani can be a brutal place. THERE IS NOTHING FLAT ABOUT IT. Sure, there isn't any mountains but since one end of the runway ends in rocks, sea and village and the other ends in rocks, it's not to be reckoned with. It's short and gravel, this is not a place you push weather. These clowns (whether it be the company or the PIC who decided to continue with a second approach) are completely at fault, they could have gone to an alternate, just like many other trips in there. I spent time there, I saw many types struggle with it but never saw a metro go in there, why would they even try to get that crap a/c in there?

[pelmet]

These clowns (whether it be the company or the PIC who decided to continue with a second approach) are completely at fault, they could have gone to an alternate, just like many other trips in there.

So you would have gone to the alternate then? It was 1/2 mile in snow at YGW. Here is the weather they got at top of descent.

CYSK 2200 UTC METAR: wind 010°T at 17 knots gusting 25 knots, visibility 2 sm in light snow and blowing snow, cloud ceiling overcast 1200 feet agl, temperature −5°C, dewpoint −6°C, altimeter setting 29.28 in. Hg.

CYGW 2200 UTC METAR: wind 030°T at 8 knots, visibility ½ sm in moderate snow, vertical visibility 400 feet, temperature −4°C, dewpoint −5°C, altimeter setting 29.18 in. Hg.

CYGL 2200 [UTC] METAR: wind 290°T at 10 knots gusting 17 knots, visibility 1½ sm in light snow and drifting snow, vertical visibility 1300 feet, temperature −7°C, dew point −9°C, altimeter setting 29.17 in. Hg. In remarks, visibility was noted as variable from 1 to 2 sm.

I say the clown would be the one to go to their alternate which was YGW, although I see in the report that the pilot planned to do so if the approaches at YSK didn't work out(I am assuming that YGW only has a non-precision approach). At 1200' and 2 miles vis, the weather at YSK is above the minimums of 600 feet and 1.5 miles. Your destination just became your alternate in my opinion. Easy decision(remember, YGL is too far away according to the report).

If caught in that situation, I would have continued with another approach. But first of all, they circled to the north with a tailwind on base leg. That is why it failed both times. It is quite possible that they lost sight of the runway strictly because they were so far off course in a turn well south of the runway having blown well through final approach. Look how far south of the runway the diagram of the circling track shows them to be. They were never close to being in a position to land. If the plan is to circle, then why not circle to the south and have the wind in your favour.

Second, if the circling does not work and it was because of low cloud or whatever obscuring visibility then work on plan B. Look at the track where they circled. It is over water and the water extends for quite a few miles. Why not just set yourself up on final at about ten miles back over the water and be aligned with the runway, corrected for wind, using the GPS with the NDB showing on course and fly inbound. Stay at circling minimums if you like and see what you see. If it doesn't work, try lower on a 3 to 1. Trust me, if you go to YGW at 1/2 mile in snow, you won't get in from the MDA if only a non-precision approach is available and will have to do the same in even worse weather.

The bottom line is that regardless of whatever mistakes were made to get into a situation where the destination and alternate are below minimums, once in that situation, you are in an emergency. Now that you have an emergency, the PIC has the legal right to break whatever regulation needs to be broken as long as it can be justified as being required to be broken to get in. It appears that that this accident happened due to an attempt to stay legal in terms of approach track flown.

I spent time there, I saw many types struggle with it but never saw a metro go in there, why would they even try to get that crap a/c in there?

Bearskin operates Metros on a bunch of 3500 foot runways in northern Ontario. This one is 3800, so why is there any problem with going into YSK?

Sani can be a brutal place. THERE IS NOTHING FLAT ABOUT IT. Sure, there isn't any mountains but since one end of the runway ends in rocks, sea and village and the other ends in rocks, it's not to be reckoned with.

There are lots of airports that end in rocks, sea or drop offs that are relatively short. That doesn't seem to stop operations. It was winter so it was likely hard pack snow, not gravel. Here is a picture. Slopes down on final so a 3 to 1 or slightly higher is the best profile.

http://www.panoramio.com/photo/83770756

Last, I would like to point out that I believe that Nav Canada should be considered in this accident. I sent a letter to them over 15 years ago complaining about the lack of approaches to airports just like this creating unsafe conditions due to circling only approaches on some runways. It appears that this situation still exists. There should be a straight in approach to 09, preferably an RNAV one. It baffles me why the TSB doesn't consider this to be a risk.

[Crusty]

The shortest strip in Bearskin's network is Fort Frances which is just under 4500 feet long.

[BTD]

As said above Bearskin does not operate into gravel or that short of strips. They are governed by accelerate stop distances etc, which are not a metro 3/23 s best quality.

Former Bear pilot.

[pelmet]

Ok, I thought they used to go into a bunch of 3500 foot strips prior to the routes being taken over by Wasaya.

[fish4life]

Perimeter goes into a ton of 3-4000' stripes all over MB, runway length wasn't the issue in this crash. As to whoever said the metro was a crap aircraft I would have to say it' held up extremely well in this crash I'm not sure a king air would have kept the passengers in the same shape.

[CID]

I am pretty sure there would have been no accident if there had been a proper LNAV approach to both ends of the runway. This would have allowed an into the wind CDA procedure to be carried out. There is no question that this is a much safer procedure than a NDB to circling limits.

In this day and age there is no excuse for public use airports to only have an NDB approach

So...the lack of an LNAV approach caused the accident? If they didn't bring charts for the NDB approach what makes you think they'd have charts for any LNAV approach? Besides, if they crashed by screwing up an LNAV approach would you then be stating that they wouldn't have crashed if they had a CAT III ILS?

The point is that you need to use what is available and you must respect the associated limitations. If you don't then you can't cite the lack of "extra" that doesn't exist.

[PositiveRate27]

I am pretty sure there would have been no accident if there had been a proper LNAV approach to both ends of the runway. This would have allowed an into the wind CDA procedure to be carried out. There is no question that this is a much safer procedure than a NDB to circling limits.

In this day and age there is no excuse for public use airports to only have an NDB approach

So...the lack of an LNAV approach caused the accident? If they didn't bring charts for the NDB approach what makes you think they'd have charts for any LNAV approach? Besides, if they crashed by screwing up an LNAV approach would you then be stating that they wouldn't have crashed if they had a CAT III ILS?

The point is that you need to use what is available and you must respect the associated limitations. If you don't then you can't cite the lack of "extra" that doesn't exist.

Bingo.

The accident wouldn't have happened if the runway was made out of cotton candy either... You gotta play the game with the field that's available.

[Big Pistons Forever]

I am pretty sure there would have been no accident if there had been a proper LNAV approach to both ends of the runway. This would have allowed an into the wind CDA procedure to be carried out. There is no question that this is a much safer procedure than a NDB to circling limits.

In this day and age there is no excuse for public use airports to only have an NDB approach

So...the lack of an LNAV approach caused the accident? If they didn't bring charts for the NDB approach what makes you think they'd have charts for any LNAV approach? Besides, if they crashed by screwing up an LNAV approach would you then be stating that they wouldn't have crashed if they had a CAT III ILS?

The point is that you need to use what is available and you must respect the associated limitations. If you don't then you can't cite the lack of "extra" that doesn't exist.

The accident statistics are clear. In order of probability of a most to least fatal outcomes. (Note: for all approach types circling at night is the most likely to result in a fatal accidents, followed by circling during the day, followed by night straight in followed by day straight in.)


1) NDB (no distance information)
2) NDB (with distance information)
3) Other Non Precision (LOC or VOR with no DME)
4) Other Non Precision (LOC or VOR with DME, LNAV)
5) Other Non Precision (LNAV with VNAV)
6)ILS or LPV or RNP.

So the aircraft was performing what is the most dangerous possible approach, a night circling NDB. I get that there was a ton of bad decisions made on the accident flight, but there is no doubt in my mind that the accident would have been less likely to have occurred if they could have flown a straight in, into the wind LNAV approach, and the only reason that this option was not available was because of a generalized unwillingness on the part of governments at all levels, and industry, to make the investments necessary to make flying safer.

I am discouraged that all the reporting on this event seems to be about baby seats in the cabin, not about addressing a systemic lack of approach aids in the North.

[126.75]

I would like some mature conversation on whether pilots who duck things up this poorly should be charged criminally. Continuing the flight without IFR approach charts? Sure their alternate crapped out, but they did not call YWG or YXU radio for updates? In my opinion these guys could have a case argued against them for negligence causing death and bodily harm.

Would knowingly flying past airports like YRL where they could duck in to print off the approach plates, be the same as waking up after a bender, still drunk and going flying? Or a pilot taking a bong hit before check in?

I used to work at Perimeter and there was not this type of pressure, and it was most of the same managers at the time. Leads me to believe it was more pilot induced get there itis than say from the company (like Keystone).

Scary Shit

As for taking the metro in and out of 3500 strips. It is easy as long as you respected the airplane, you could not treat it much like a Navajo. Perhaps that is not "easy" but I'd consider it to be, does not mean every approach worked out. Had to be on the numbers, on slope and put the airplane down in the first few hundred feet. If it wasnt looking like you would go around and try again.

[CID]

and the only reason that this option was not available was because of a generalized unwillingness on the part of governments at all levels, and industry, to make the investments necessary to make flying safer.

I think that's unfair. There are a LOT of airports out there and the government has a set level of resources to implement these approaches. It's far too simple to say that it's an "unwillingness" especially when you consider the airport in question now has an LNAV approach to both runways.

[JigglyBus]

The accident statistics are clear. In order of probability of a most to least fatal outcomes. (Note: for all approach types circling at night is the most likely to result in a fatal accidents, followed by circling during the day, followed by night straight in followed by day straight in.)

1) NDB (no distance information)
2) NDB (with distance information)
3) Other Non Precision (LOC or VOR with no DME)
4) Other Non Precision (LOC or VOR with DME, LNAV)
5) Other Non Precision (LNAV with VNAV)
6)ILS or LPV or RNP.

Out of curiosity, where are you getting these statistics?

I'm not saying that they are incorrect, just wondering where you came across them.

[PROC_HDG]

I am pretty sure there would have been no accident if there had been a proper LNAV approach to both ends of the runway. This would have allowed an into the wind CDA procedure to be carried out. There is no question that this is a much safer procedure than a NDB to circling limits.

In this day and age there is no excuse for public use airports to only have an NDB approach

So...the lack of an LNAV approach caused the accident? If they didn't bring charts for the NDB approach what makes you think they'd have charts for any LNAV approach? Besides, if they crashed by screwing up an LNAV approach would you then be stating that they wouldn't have crashed if they had a CAT III ILS?

The point is that you need to use what is available and you must respect the associated limitations. If you don't then you can't cite the lack of "extra" that doesn't exist.

Bingo.

The accident wouldn't have happened if the runway was made out of cotton candy either... You gotta play the game with the field that's available.

Accidents never have only one cause. Having an RNAV at CYSK may have been the malaligned hole in the swiss cheese model required to prevent a lot of the other major errors causing a crash. Whether they had charts or not, having a t-bar approach with lower minimums to both runway ends almost certainly would have made a difference in this case.

PROC_HDG

[Crusty]

I think a secondary interesting discussion from an ethics discussion is this:

Your buddy calls you on the radio 10 minutes after he takes off and asks you to read him an approach plate for his destination so he can make notes because he forgot his plates. Do you do it?

[pelmet]

I am pretty sure there would have been no accident if there had been a proper LNAV approach to both ends of the runway. This would have allowed an into the wind CDA procedure to be carried out. There is no question that this is a much safer procedure than a NDB to circling limits.

In this day and age there is no excuse for public use airports to only have an NDB approach

So...the lack of an LNAV approach caused the accident? If they didn't bring charts for the NDB approach what makes you think they'd have charts for any LNAV approach? Besides, if they crashed by screwing up an LNAV approach would you then be stating that they wouldn't have crashed if they had a CAT III ILS?

The point is that you need to use what is available and you must respect the associated limitations. If you don't then you can't cite the lack of "extra" that doesn't exist.

Bingo.

The accident wouldn't have happened if the runway was made out of cotton candy either... You gotta play the game with the field that's available.

What makes me think that they would have had the information for an RNAV approach despite having forgotten their charts is that they had the information for an NDB approach despite having forgotten their charts. They would have gotten their info from the same source.

It is silly to bring in the question of a CAT III ILS into the conversation but it is not unreasonable at all to expect a straight in RNAV approach at an airport that is the most qualified type to get one. An airport with no terrain of significance and no straight-in approach available with traditional navaids to a runway that is frequently into wind.

We hear from so many posters after an accident on how the pilots should not be the only ones taking blame, that we need to look at the higher echelons such as management and regulating authorities and their role. Well how about the role of NavCanada who(now that I see YSK has two RNAV approaches) waited who knows how many years to finally put in RNAV approaches.

It is nice to say that pilots play with what is available to them and there is something to be said about that but why are pilots being still given so few tools to play with about a quarter century after GPS approaches started. A massive amount has been spent by them on ADS and CPDLC by Nav Canada but I don't think it has prevented one death. It is a well known fact that circling approaches, especially into tight dark holes like this on old aircraft are much higher risk with many, many accidents over the years.

So coincidentally, two straight in RNAV approaches are installed after the accident. The equivalent of the so-called "Tombstone technology" where it took a significant amount of deaths to push for regulatory/technological change.

[Diadem]

For those who think a stabilised approach below mins would be an appropriate solution, I'd like to point out that the crew had the runway in sight on both approaches. They were also below mins on both. At one point on the second approach, the FO commented that the runway looked good. This wasn't a CFIT where they flew into the ground without seeing it; the captain flew a shitty approach at ref+30 with a tailwind, and touched down 3000' past the threshold. It wasn't a matter of not seeing the runway, and not having any other options, it was his poor decision making and inability to judge that he was fucking up. Taking off with 200 lbs of granny gas, not having the approach plates, and utilising an alternate that was marginal even on the best forecasts were indications that his decision-making was poor, but ultimately it came down to shitty flying.

[stef]

Ok. Point taken Diadem.

[Rookie50]

.

[pelmet]

For those who think a stabilised approach below mins would be an appropriate solution, I'd like to point out that the crew had the runway in sight on both approaches. They were also below mins on both.

Isn't that why they had the runway in sight, because of their further descent. They just did it in an improper manner

the captain flew a shitty approach at ref+30 with a tailwind, and touched down 3000' past the threshold. It wasn't a matter of not seeing the runway, and not having any other options, it was his poor decision making and inability to judge that he was fucking up. Taking off with 200 lbs of granny gas, not having the approach plates, and utilising an alternate that was marginal even on the best forecasts were indications that his decision-making was poor, but ultimately it came down to shitty flying.

I agree the flying was pretty bad. For the circling, all they had to do was circle to the south. While I haven't flown a Metro, my guess would be with runway in sight, a turn to the left for 20 to 30 seconds depending on the winds, past the threshold for 30 seconds or so depending on the winds, using an appropriate wind correction angle backed up by GPS track display and then a turn onto final. It is fairly simple and works well. The captain appears to have wanted to fly and be the one looking outside as he kept the airport to his left.

Why not just fly and let the FO help guide you around with a proper briefing on what calls to make. I find that calls in this situation can be helpful or not helpful depending on what is said. Instead of the PNF saying things on the turn inbound that may not be particularly useful, how about a call such as "20 degrees bank" and "500 fpm descent" instead of your a little bit tight and start descending" leaving you having to guess on how much correction to make. Exact commands can be very easy to follow promptly. They also had PAPI's on this runway which are very useful once they come into view. Once on the descent and the PF has the airport in view, the PNF can be really useful by calling out "ref plus ____ sink 800" as the PF is concentrating 100% looking outside.

To try to dive into a very short for type runway at 1800 fpm and ref plus 30 with a 15 knot tailwind(and according to the report, a significant number of seconds delayed prior to initiating this approach from almost the MAP) makes no sense.

Then there is the go-around, Why would you wait until being at the very end of the runway to go around at 50 or less feet when you know you would overrun even if you landed halfway down the runway. But a go-around should not be a big deal. If I remember correctly, the PNF made a call that they were 5 knots slow and according to the report, the captain likely pushed forward on the control column. I suspect quite a bit. Something like this typically happens in what are known as Somotogravic effect accidents where the acceleration gives the pilot the sense of pitching up excessively when in fact this is not the case. The best defense against this is a good scan. I notice that the captain had flown widebody jets at two carriers previously to this. Typically they have modern EFIS flight displays that are known for reducing pilots instrument scanning skills over the long term.

I'm surprised that the TSB didn't look further into this as a go-around even at low energy should not be a big deal, especially in cooler temperatures like that day with plenty of power available, assuming that plenty of power was applied of course.

[PositiveRate27]

I agree that having an RNAV approach would have reduced the chance of this accident from happening, but that is not the lesson that should be interpreted here. We as pilots can't just say "hey, the government didn't give us the best equipment available so it's out of our control."

The fact is these pilots demonstrated alarmingly poor decision making and we as a group need to look at this and ask ourself how we can avoid putting ourselves in these types of situations.