BE1900 IMC CFIT in Alaska

[aterpster]

NTSB Identification: ANC13FA030
Nonscheduled 14 CFR Part 135: Air Taxi & Commuter
Accident occurred Friday, March 08, 2013 in Aleknagik, AK
Aircraft: BEECH 1900C, registration: N116AX
Injuries: 2 Fatal.

This is preliminary information, subject to change, and may contain errors. Any errors in this report will be corrected when the final report has been completed. NTSB investigators either traveled in support of this investigation or conducted a significant amount of investigative work without any travel, and used data obtained from various sources to prepare this aircraft accident report.

On March 8, 2013, about 0814 Alaska standard time, a twin-engine turboprop Beech 1900C airplane, N116AX, was destroyed when it impacted rising terrain about 10 miles east of Aleknagik, Alaska. The airplane was operated as Flight 51, by Alaska Central Express, Inc., Anchorage, Alaska, as an on-demand cargo flight under the provisions of 14 Code of Federal Regulations (CFR) Part 135. The airline transport certificated captain and the commercial certificated first officer sustained fatal injuries. Instrument meteorological conditions were reported in the area at the time of the accident, and the airplane was operating on an instrument flight rules (IFR) flight plan. The flight had originally departed Anchorage about 0544, and made a scheduled stop at King Salmon, Alaska, before continuing on to the next scheduled stop, Dillingham, Alaska.

According to Federal Aviation Administration (FAA) personnel, as the airplane approached Dillingham, the flight crew requested the RNAV GPS 19 instrument approach to the Dillingham Airport about 0757 from controllers at the Anchorage Air Route Traffic Control Center (ARTCC). The ARTCC specialist on duty subsequently granted the request by issuing the clearance, with instructions to proceed direct to the Initial Approach Fix (IAF) to begin the approach, and to maintain an altitude of 2,000 feet or above. A short time later the flight crew requested to enter a holding pattern at the IAF so that they could contact the Flight Service Station (FSS) for a runway conditions report, and the ARTCC specialist granted that request. The ARTCC specialist then made several attempts to contact the aircraft, but was unsuccessful and subsequently lost radar track on the aircraft.

When the airplane failed to arrive at the Dillingham Airport, ARTCC personnel initiated a radio search to see if the airplane had diverted to another airport. Unable to locate the airplane, the FAA issued an alert notice (ALNOT) at 0835. Search personnel from the Alaska State Troopers, Alaska Air National Guard, and the U.S. Coast Guard, along with several volunteer pilots, were dispatched to conduct an extensive search effort.

Rescue personnel aboard an Air National Guard C-130 airplane tracked 406 MHz emergency locater transmitter (ELT) signal to an area of mountainous terrain about 20 miles north of Dillingham, but poor weather prohibited searchers from reaching the site until the next morning. Once the crew of a HH-60G helicopter from the Air National Guard's 210th Air Rescue Squadron, Anchorage, Alaska, reached the steep, snow and ice-covered site, they confirmed that both pilots sustained fatal injuries.

The closest official weather observation station is at the Dillingham Airport. At 0745, an aviation routine weather report (METAR) reported, in part: Wind from 100 degrees (true) at 17 knots with gusts to 30 knots; visibility, 7 statute miles in light rain; clouds and sky condition, 1,500 feet overcast; temperature, 34 degrees F; dew point, 34 degrees F; altimeter, 29.09 inHg.

On March 9, the National Transportation Safety Board (NTSB) investigator-in-charge, along with an additional NTSB air safety investigator, and an FAA operations inspector from the Anchorage Flight Standards District Office (FSDO), examined the airplane wreckage at the accident site. A comprehensive wreckage examination and layout is pending following recovery efforts.

RNAV 19 chart: http://aeronav.faa.gov/d-tpp/1303/05166R19.PDF

[oceancrosser]

According to Federal Aviation Administration (FAA) personnel, as the airplane approached Dillingham, the flight crew requested the RNAV GPS 19 instrument approach to the Dillingham Airport about 0757 from controllers at the Anchorage Air Route Traffic Control Center (ARTCC). The ARTCC specialist on duty subsequently granted the request by issuing the clearance, with instructions to proceed direct to the Initial Approach Fix (IAF) to begin the approach, and to maintain an altitude of 2,000 feet or above. A short time later the flight crew requested to enter a holding pattern at the IAF so that they could contact the Flight Service Station (FSS) for a runway conditions report, and the ARTCC specialist granted that request. The ARTCC specialist then made several attempts to contact the aircraft, but was unsuccessful and subsequently lost radar track on the aircraft.


Hmmm, I hope I am reading this correctly, but looking at the chart that is linked to at the bottom, the minimum over (and northeast of) the IAF (ZEDAG) is 4300' within a 4nm radius, and 6300nm within a 30nm radius, and 5400' 30nm SW of the IAF. I hope the refence to 2000' feet is simply wrong, they would have had charts out with the higher altitudes I think, and 2000 feet is way to low.

[thermostat]

I see that the min altitude over the IAF, Zedag, is 4300 feet, yet they cleared down to 2000 ft. or above. What kind of clearance is that? Looks like they should have crossed Zedag at 4300 and then started descent on a 3 degree angle to the MDA/VDP. Another screw-up?

[aterpster]

I see that the min altitude over the IAF, Zedag, is 4300 feet, yet they cleared down to 2000 ft. or above. What kind of clearance is that? Looks like they should have crossed Zedag at 4300 and then started descent on a 3 degree angle to the MDA/VDP. Another screw-up?

Big time screw up. The minimum holding altitude is 4,300. They asked to hold while they went off frequency. Their last clearance from ATC was "at or above 2,000."

[Dufo]

Sound similar to Adria MD80 in Ajaccio, 1980.
Cleared below MSA by ATC and hit the ground in IMC. :uhoh:

[BOAC]

I accept all the above, even if I am not familiar with Alaskan ATCC procedures, but is it not reasonable to expect crews to:-

a) Know where they are on an RNAV approach
b) Read the chart
c) Maintain a safe altitude vis a vis position /holding etc?

Especially when not under 'Control'

[kotakota]

Sounds like the good old days in Zurich . Turned down many of their shortcuts .

[ATC Watcher]

Their last clearance from ATC was "at or above 2,000.
Let me check on this. Does not sound right. Also that they would descend to that altitude blindly with the APP Chart on their knees.
Something mising here.

[Annex14]

I agree with what was posted here before. However, some remarks from an ATC - Dinosaur.
The flight was IFR in IMC en route from King Salmon to Dillingham. That routing indicates the probability that the flight was about to enter or within the 30 miles arc around ZEDAG with an MSA of 5400 ft. Next the crew requested to enter the holding pattern connected to ZEDAG. That most probably leads to a tear drop entry into the 30 miles sector north of ZEDAG with an MSA of 6300 ft.
With this as a background and the published RNAV Entry Minima at no time a clearance that contains wording to "maintain at or above 2000 ft. altitude" could be issued by ATC. As I remember ATC is responsible for the terrain clearance of IFR flights in IMC!!
So en route to ZEDAG no clearance below 6300 ft or any other MSA to be obeyed. After establishment in the holding pattern descent clearance to 4300 ft. And only than with the approach clearance issued the request - "report Zedag inbound to the field and leaving 4300 ft "
That´s how we did it some time ago !

[ATC Watcher]

ANNEX14 : As I remember ATC is responsible for the terrain clearance of IFR flights in IMC!!
Your memory fails you , no, responsibility in IFR for terrain avoidance is always with PIC except when under radar vectoring . (ICAO DOC 4444)

[Annex14]

ATC Watcher

Yes, you´re correct, forgot about that "limitation" !!
Never the less, in my active years I wouldn´t even have gotten it in my mind to clear a flight below MSA. Probably part of the "responsibility drill" we went through in education.

[galaxy flyer]

Strangely, or not, the MHA at King Salmon's RANV (GPS) s 2,000'. Could there have been confusion on the IAF by either the crew or the ATCO or in the comm loop?

[ohnutsiforgot]

Under stress, I took a quick glance at the approach chart. I saw 1900 feet proceeding down the glideslope, a 2550 foot obstacle and a buch of lower obstacles (1623, 1860 ) and a missed approach of 2600..plus an authority figure in my life just told me 2000 feet. I worried only about the 2550 foot obstacle and vowed to hang around 2600 to keep a cushion
and do the missed. I filtered out all the other information becasue I was under stress.

[aterpster]

ATC Watcher:

Your memory fails you , no, responsibility in IFR for terrain avoidance is always with PIC except when under radar vectoring . (ICAO DOC 4444)

Actually U.S. ATC procedures are much more restrictive in this regard than most of the world as a result of the 1974 TWA 514 CFIT accident.

ATC is responsible for assigning an altitude at, or above the MEA if the flight is on a published route. If the flight is cleared direct then en route ATC (which was the case here) must assign at altitude at or above the ATC center's MIA chart.

The crew also had a duty to challenge this improper clearance because the approach chart had a wealth of operational altitude information.

The ATC clearance per the U.S. FAA controller's manual should have been not less than 4,300 since the flight was cleared direct to ZEDAG.

The preferred clearance would have been, "Cleared direct ZEDAG, maintain 4,300 [or higher], expect clearance for the RNAV Runway 19 approach."

A second compliant clearance would have been, Cleared direct ZEDAG, cross ZEDAG at 4,300 [or 4,300 or above], cleared for the RNAV Runway 19 approach."

[aterpster]

Strangely, or not, the MHA at King Salmon's RANV (GPS) s 2,000'. Could there have been confusion on the IAF by either the crew or the ATCO or in the comm loop?

The Anchorage Center MIA is also 2,000 from King Salmon to fairly near PADL, where it becomes 4,000. Five west, or so, or PADL the MIA is 5,000.

[avionimc]

Under stress, I took a quick glance at the approach chart. I saw 1900 feet proceeding down the glideslope
Relax, there is no glideslope, it is a non precision.

[westhawk]

Wow. A wake-up call for sure. A reminder that under the wrong circumstances you don't have to be very far off your game...

Whatever the NTSB ends up saying, the lesson I'm reminded of when it comes to ATC clearances is best described by one of Ronald Reagan's most notable quotes: "Trust.... But verify."

And I was just telling a story yesterday about that: We were holding short of the runway at KEGE one morning, just finishing the checklist when the tower cleared us for takeoff. I put my hand on the parking brake handle as I looked up the approach path. What I saw made me put my hand back on my lap. I looked at the FO and he looked at me, then I pointed to myself, indicating I'd like to be the one to make the radio call. The FO nodded as I called tower and told them we'd need to hold short for a couple of minutes. The next radio call apparently came from the AA B-757 on a quarter mile final: "Thanks guys." The new voice on tower sounded a little stressed when he canceled our takeoff clearance. Lesson? Everyone makes mistakes, So trust... But verify.

[DownIn3Green]

As a 10 year ATC, and 23 year ATP, Arpster is correct, sort of....

It seems that this accident closely mirrors the TWA accident in Virginia...

However, ATC procedures were changed as a result of that accident...When not on a pulbished segment and subsequently being "cleared for approach", the correct phraseology is "Maintain xxx until ESTABLISHED on the approach, cleared for..."

In this case, the controller didn't " protect" the flight with the clearence to descend to 2000'...Then once they requested holding, the clearence should have been something along the lines of ..."Climb and maintain 4,300, cleared to hold..."...etc...

In any case, the PIC is responsible to insure the clearence is correct, as ATC is responsible to assist when the flight descends below the MVA...

In a non-radar enviornment this becomes critically important...

Seems someone here dropped the ball...from what I've seen posted here, there is enough blame to go around so everyone can have a piece of the pie...

[aterpster]

DownGreen:

As a 10 year ATC, and 23 year ATP, Arpster is correct, sort of

Why sort of? :)

If nothing else you make the case for MVAs and MIAs being widely distributed aviation public data instead of FAA state secrets.

[Annex14]

What I missed yesterday when reading your first post was the height of the crash site. Started a search for and found it in here:
Crash: Ace Air B190 near Dillingham on Mar 8th 2013, wreckage found (http://avherald.com/h?article=45ee9215&opt=0)

They obviously were indeed at the low altitude of 2000 ft which was part part of their clearance from ATCC Anchorage. Aside of what is apparently ATC responsibility the question remains: why for heavens sake the crew descended so low with all the information on the approach charts at hand ??
Hopefully not another case of over confidence in the magic words "identified, Radar contact" and the mentally wrong expectation of flying safe if only the instructions are followed.

[BOAC]

"identified, Radar contact" - ?? Radar contact? I thought this was a procedural/self-nav?

[Annex14]

Admit, to mention these words might be somewhat misleading. The assumption that they were flying on just flight plan data and reporting to ANC ATCC via a radio relay station might be correct. The opposite might be correct as well and via some detached RADAR station ANC can see them on RADAR ??? But the question still is unanswered: Based on which factual information the clearance for a direct course to a RNAV position connected with an instruction " to maintain at or above 2000ft" was transmitted to the crew ??
My basic assumption in this case is to believe that all of the person involved in this accident knew all about the weather conditions and the terrain situation much better than one at hindsight can determine.

[BOAC]

Yes. Barring some unknown unfortunate problem occurring, we appear to be in the 'old fashioned aviation' arena here:-

1) Know where you are
2) Know where to want to go
3) Know how to get there

[Annex14]

Yes, fully agree!!
Interesting reader comments under that Aviation Herald report!

[BOAC]

Well, it's either me or them, but why are people on AvH saying 'they should have requested to climb to 4300' when they were cleared NOT BELOW 2000' ?? Is it me or these posters not understanding the clearance?

[Annex14]

Probably the understanding of clearances on this side of the pond differs ??
As I see it, this clearance covered any level between 2000ft. and as high as they could fly. That is one of the reasons that make me shrug, a good clearance should not leave any space for individual interpretation.
Great guess, it must have to do with the special conditions in Arctic Flying.

[BOAC]

Hopefully the answer to why they held at ZEDAG below 4300' will come from the CVR.

[aterpster]

BOAC:

Hopefully the answer to why they held at ZEDAG below 4300' will come from the CVR.

The flip-side to that is why were they issued a clearance to fly as low as 2,000 when they were entering an area where the minimum off-route altitude (known as minimum instrument altitude "MIA" in FAA Air Route Traffic Control "ARTCC" domestic ATC parlance). In that part of Alaska flights typically communicate directly with the ARTCC and on the route they were flying are typically radar identified.

The FAA and NTSB already have the answer to the question, "Why were they cleared at or above 2,000, and how high did the 'above 2,000' go?"

[ATC Watcher]

Looking at the Alaskan VFR chart ( the only one I have ) the MSA from the East ( coming from ANC ) is 2.800 Ft. After passing the Chigmit Mountains the terrain is pretty flat until DLG, the highest obstacle in that peak at 2550ft near the IAF. ( Ironically a direct track ANC-DLG would keep you well clear of that Peak.)
Surely the guys had also a GPS moving map with terrain on it, and warnings probably. Must be something else...

[BOAC]

Copy all that aterp, but the question remains why they held at ZEDAG below 4300' - no one told them to.

Are you now saying they WERE probably radar identified?and on the route they were flying are typically radar identified.:confused:. Why not, then. control?

[aterpster]

BOAC:

The area along their route has radar coverage.

They went off frequency to obtain a runway conditions report. So, if or when the ARTCC's EMASW went off ATC couldn't contact them in the brief time before impact.

The remainder of your questions will have to await release of the radar, ATC communications and (hopefully) the CVR data.

[aterpster]

Someone on another board plotted their probable route of flight on SkyVector.com:

SkyVector: Flight Planning / Aeronautical Charts (http://skyvector.com/?ll=59.04817558961184,-157.8334170759333&chart=48&zoom=3&plan=A.PA.PAKN:V.PA.AKN:G.58.89450186670755,-157.86639404384033:F.PA.ZEDAG:G.59.32015874876182,-158.26560974206853)

[BOAC]

I am not aware of the different levels of ATC service in those parts, but have we established what they were on, or perhaps thought they were on?

Also why BOTH off frequency? Single radio?

[aterpster]

OK465:

Apart from altitude issues....

I would guess that for RNAV equipment, a 'vintage' Be1900 would probably be equipped with a G530 or even 430, a fairly basic display of the profile, IIRC without displayed altitude constraint info.

In addition, when holding using this type of system, the steering guidance for NAV goes into a 'suspend' mode at the intended holding fix, requiring use of a HEADING SELECT type mode for the outbound leg of the holding pattern AND the turn inbound. So you're not really on a structured RNAV routing at this point. This is holding 101.

This requirement coupled with the existing 17 knots or more of right cross outbound at holding speeds probably ~150 certainly makes precision tracking as well as positional awareness challenging.

No doubt that holding with 400/500 is quite a challenge, having to use OBS mode and such. And, "suspend" for a HILPT is only good for one circuit, then OBS holding would be required.

Same for a 400W/500W although they supposedly get course guidance the first time around. In any case the evaluated and protected airspace for this hold is huge, because it was evaluated for climb-in-hold (310 KIAS) for the missed approach for the RNAV Rwy 1.

The terrain they hit is well within the 200 KIAS primary holding area.

[aterpster]

okc465:

This is reminiscent of the old 'cruise' clearances, i.e. "cruise 2000, cleared for an approach". Pilot adjusts as required, phone call on the ground....

Is that in play at PADL?

Shouldn't have been. You can descend on a cruise clearance, not climb.

[pattern_is_full]

What are the odds that a B1900 in cargo configuration (< 6 passenger seats) had a CVR? FAR 135.151 indicates one would not be required.

ACE flies a "combi" B1900 with passenger seating as well as 4 cargo planes - but it's not the accident aircraft.

[aterpster]

BOAC:

Also why BOTH off frequency? Single radio?

No doubt that will be a fact of great interest.

[Annex14]

I checked in FlightAware for a track of the flight and found this:Alaska Central Express (KO) #51 (http://de.flightaware.com/live/flight/AER51/history/20130308/1615Z/PAKN/PADL)

I´d like to say now there is even more than a mystery why they had to hit that mountain. The Speed / Altitude graph clearly shows they were at a safe altitude until few minutes before the crash.
As well, aside of all RNAV related questions it appears they initially where flying on an airway, most probably established on a radial of AKN VOR and having DLG ahead of them. Plus the RADAR coverage.
Big, big question WHY !!!

[BOAC]

If those speeds are reliable, they are fast for a hold and the rate of descent is high.

[Annex14]

Looks like they had a good tailwind>> check weather in AvH report!
Could that fairly steep descend indicate some ice problems ? Don´t know how good the ice provisions are on a Be 1900??
Was mentioned before by other poster, but to place that holding pattern to the side of the highest mountains and a non standard right hand pattern is something to wonder about at least.

[BOAC]

Not forgetting that there are no hills at 4300' - and the entry side has to be considered as well.

[aterpster]

Annex 14:

I´d like to say now there is even more than a mystery why they had to hit that mountain. The Speed / Altitude graph clearly shows they were at a safe altitude until few minutes before the crash.

Terrain is flat until you head for the course-reversal hold at ZEDAG. The 2550' mountain is smack in the center of the holding pattern. ATC should have climbed the flight to 5,400 before they entered the downwind TAA.

http://aeronav.faa.gov/d-tpp/1303/05166R19.PDF

[Annex14]

Fully agree!!
At the very moment they turned off AWY V 453 MEA no longer was valid for them, instead the mentioned MOA 5400 ft.

With that now no longer applicable " to maintain or above 2000 ft" part of their clearance in mind the crew might have expected to be safe under RADAR control and probably too pre-occupated with the weather at Dillingham to catch the flaw.
Would be also interesting to get to know what their filed FL on the flightplan was.
Anyway, according the Metars no chance to break clouds from that 2000 ft altitude. Guess that´s the reason why they went inbound ZEDAG for a normal RNAV approach.

[BOAC]

Still no-one able or willing to answer my query about ATC in those parts and what sort of service the a/c might have been having.

1) Before we start lambasting the controller, based on what we know the a/c was NOT 'descended' to <2550' by ATC. The crew either deliberately did it (in the wromg place) or some event caused it. To say "ATC should have climbed the flight to 5,400 before they entered the downwind TAA" implies the controller had descended the a/c - which I am not seeing, and also that he/she was aware of its altitude which is not stated.

2) It is to me an unusual 'clearance', but not one that would have caused me any difficulty, merely given me great flexibility in what was, I assume, a quiet piece of sky. 'Downin3' says it was an 'incorrect' clearance ("However, ATC procedures were changed as a result of that accident...When not on a pulbished segment and subsequently being "cleared for approach", the correct phraseology is "Maintain xxx until ESTABLISHED on the approach, cleared for...") - incorrect at all times or only when under ATC control?

3) What is the significance of the min 2000' clearance - does some sort of airfield zone start there?

So - for the 'locals' - what sort of 'control' was this? Would radar be used, a/c identified and full control be expected? Would an altitude squawk be available to whoever was 'controlling'? Can a crew elect in IFR to 'pass;' on a full service?

[autoflight]

It is not that rare that a very dangerous ATC clearance is in place. In my case, 3 times in 40 years of flight.

Crews have no business accepting clearances without at least an appropriate degree of safety confidence.

Crews might think that during radar vectors on a familiar route in day VMC and all good english speaking crews might not need the same type of situation awareness as more demanding circumstances. There is more risk with the latter, but I could have been just as dead with one instance of the former.

Some of us have been distressed to learn of the loss of a friend due CFIT. Avoiding CFIT starts a long time before take-off. Maintain SOPs, crew management, sober and rested and study unfamiliar routes will all help. Then when that slightly suspect ATC clearance appears, crews will be less diverted by narrow perceptions and lack of preparation. Just "being carefull" is not enough.

[Island-Flyer]

While I agree the pilots should have questioned the altitude clearance, if they were in a radar environment ATC could descend them below the published MEA/MSA down to the MVA (since the MVA is a more detailed charting of minimum safe altitudes). The pilots most likely would have had no available reference for MVA. Many pilots fall into the trap of trusting ATC too much and at worse it appears these pilots fell victim to controller error and complacency.

That was a very weird altitude clearance though, I believe standard practices require that ATC only descend an aircraft to the IAF or FAF crossing altitude and then clear them for the approach. I've never seen ATC arbitrarily select an altitude for descent.

[Zeffy]

Island-FlyerThe pilots most likely would have had no available reference for MVA.

User groups and safety organizations have been begging for cockpit MVA/MIA references, especially now that electronic moving maps could facilitate them.

This article from nearly ten years ago:

http://flightsafety.org/fsd/fsd_sept04.pdf

[BOAC]

We are nibbling at the edges here:-

It is being implicitly assumed (unproven) that the pilots were being 'vectored'.

I believe these pilots did NOT need MVA or any other trigraph except MSA/SSA which I assume is on the charts?

What possible use is MVA (or MEA) if you are not under radar control?

I'm still slightly puzzled by the speeds shown on Annex14's link and the rates of descent.

[Zeffy]

BOAC
What possible use is MVA (or MEA) if you are not under radar control?

An MIA chart could have enabled the crew to detect the erroneous altitude assignment by ATC.

When the approach clearance was issued, the airplane was in a 2000' MIA sector, but traveling toward rising MIA's.

[BOAC]

They were not 'assigned' any altitude.

[Zeffy]

OK, then -- "Clearance".

From the NTSB Preliminary:

The ARTCC specialist on duty subsequently granted the request by issuing the clearance, with instructions to proceed direct to the Initial Approach Fix (IAF) to begin the approach, and to maintain an altitude of 2,000 feet or above.

[Annex14]

Plse check aterpster # 44 link to the approach map. You will see there is no procedure altitude of 2000 ft at the IAF ZEDAG.

The 2000 ft (plus) clearance was only good on that awy V 453. Leaving that safe track automatically required MOA 5400 ft - depicted in that app map - or with a clearance of ATC for a RADAR vector towards ZEDAG and the minimum for the RNAV procedure at that point - 4300 ft.
Anything else in the weather and terrain situation at hand would have been guesswork. And for the latter there is very impressive phrase - Never base control on assumptions!!!!. This sentence was sadly proven to be correct from both parties involved - ATC and the two pilots.

[Zeffy]

Annex14Plse check aterpster # 44 link to the approach map. You will see there is no procedure altitude of 2000 ft at the IAF ZEDAG.

No argument there and the charts are familiar to me.

Only proposing that pilot knowledge of MIA might have been helpful.

Your point is a good one, though.

Since it appears that the crew ignored or failed to properly consult the approach chart -- and appeared to place sole reliance on the ATC clearance -- it might be a stretch to assert that they would have displayed any other available source for minimum off-route altitude.

However, it remains to be seen whether ATC radar was ever terminated. It sounds as though the flight was still receiving radar services at the time the erroneous ATC clearance was issued.

[Annex14]

The sad part of the story is - according that FlightAware record > see #40 link - untill few minutes before the crash they were at a safe altitude.
So, what made them descend so rapidly ??
- a yet unknown weather related problem ?
- an attempt to brake clear of clouds early and proceed VMC to Dillingham?
- a fatal trust on a clearance for an altitude that became invalid by the track they were sent to?
Many open questions and only the CVR - if installed and active - and ATC tapes of RADAR and Radio data may give an answer.

[ATC Watcher]

I was told the NTSB will start its work on Monday and that shortly after a new intermediate report will be issued.
Then we will know what kind of ATC service was provided by ATC and if the clearance issued was standard or not for that type of service. .

Now from a pilot point of view, I find this RNAV chart crazy, considering the RNAV/GPS nature of the procedure who on earth decided to put an IAF so close to the only peak in the whole area ?

Reminds me the old APP chart for Santa Maria in the Azores, it you were inbound on the wrong VOR ( SMR or SMA, I can't remember) coming from the East ,the APP track made you overfly the only bloody (high) peak in the whole island.
It was an accident waiting to happen, and sadly a US B707, did get caught with a wrong QNH one day, and all on board died as a result.

Very often guys designing these charts do that from an office, very far away from the actual location and with no real flying experience . Not saying this was the case here, but it smells like it.

[BOAC]

ATC Watcher - nothing 'crazy' there - why not ask why someone built a 2550' hill near the approach to R19? With the MSAs around there, a 13 mile IAF is about right - could not be closer. In any case, you need almost 60mb altimeter mis-setting to hit the hill!

[Annex14]

The RNAV procedure as published is safe, though it shows a weird holding pattern !!
However, that is only correct to say if the published altitudes are obeyed.

Moving the IAF closer to the rwy means lower altitude at IAF, most probably causing terrain clearance problems in procedure construction.

[BOAC]

The RNAV procedure as published is safe, though it shows a weird holding pattern !!
However, that is only correct to say if the published altitudes are obeyed. - errm - how else would you do it? What, as OK says, is wrong with a right hand hold?

[Annex14]

Because it is right hand and towards the only real obstacle in miles.
To have it mirrored and flown in standard turns can´t be that big of a problem.
Had the pattern been on that side it appears very likely they hadn´t hit the rocks.

[Annex14]

BOAC

you were too fast for me to enter that of course it doesn´t make a difference to fly right or left hand holding patterns. But in my - past world - there was a preference and that was called left hand standard pattern.

[galaxy flyer]

Looking at the chart, one could fly inbound from ZEDAG to the FAF at 1,900' which I'm guessing was the the reason for the "cross ZEDAG at 2,000 or above" clearance. The late request for the hold was what led to the CFIT. AFAWK, neither the controller, nor the crew took action to CLIMB to the 4,300' MHA. The terrain on th final approach course is fairly flat.

The course inbound is 13nm to accommodate the gradient for the intermediate segment which is dictated by the terrain in the course reversal hold. It is 316'/nm from the IF to the FAF, as it is.

I'm not so sure making it standard would have lowered the MHA. There is more terrain than depicted.

[BOAC]

Had the pattern been on that side it appears very likely they hadn´t hit the rocks. - don't forget the 'joining area'? At 2000ft had they come from the north-west.......

At 4300' there are no obstacles, left or right. There is only one reason they 'hit the rocks'.

[Annex14]

Yes, agreed, you know it, I know it and many others too. But fairness dictates to wait for the results of an independent evaluation and research.

[galaxy flyer]

OK465

Oops, you're correct on right being standard? "Off right, jinking left", then :E

[oceancrosser]

Annex14 said:
But in my - past world - there was a preference and that was called left hand standard pattern.

Strange, in my current and past 34 years of instrument flying, standard holdings were and are right hand turns. (Looks like OK465 beat me too it).

[Annex14]

Okay gentlemen, before we go too far off track I admit my phrasing was misleading.:ugh:

For the sake of a correct discription I cite what is written in ICAO DOC 8168 PANS - OPS Vol. II - Part VI, chapter 1:
The criteria contained in this part are related to right turns holding patterns. If no operational considerations prevail, right turns holding patterns should be established. For left turns holding patterns, the corresponding entry and holding procedures are symmetrical with respect to the inbound holding track.

Hope that helps.

[ATC Watcher]

Thanks Annex 14 ( nice appropriate name btw ) I know myself of a few published left hands Holdings around some large airports .

BOAC : OK, " crazy" was perhaps not the best adjective to use, grant you that, what I meant was that if you look at the terrain , and you do not depend on a fixed ground Nav aid , I pesonally would think that in this case, drawing a virtual IAF 10 NM away and with LH pattern would make a lot of common sense.. But I am not a procedural map producer.

[aterpster]

ok465:

These procedures are all flight checked by people with 'real flying experience' before release, as well as 'flight validated' in a simulator even before the in-flight check using stringent guidelines that include checks at temperature and wind extremes.

They cannot, however, be checked for all possible specific RNAV system operational issues.

I think you mean to say that flight inspection cannot possibly check for possible errant blunders.

There is nothing wrong with the design of this procedure under FAA policy and design criteria.

[aterpster]

Jepp chart with overlay of Anchorage Center's MIAs for the area:

http://i201.photobucket.com/albums/aa214/aterpster/PADLMIAsonJeppRNAV19_zps59c3d9f5.jpg

[Annex14]

Thanks for the update. This overlay confirms my assumption which led to my post # 68.
Will be very interesting to read the findings of NTSB and FAA.

[galaxy flyer]

Aterpster

With those MIAs, how could the controller issued the "at or above 2,000 feet" clearance? Is there a lower MVA?

[aterpster]

GF:

Aterpster

With those MIAs, how could the controller issued the "at or above 2,000 feet" clearance? Is there a lower MVA?



No MVAs. This is center airspace. MIAs are used as MVAs when they have radar contact.

Consider those to be center MVAs on the chart.

[galaxy flyer]

That I understand, what doesn't make sense is the clearance "at or above 2,000'" when the plane was in an area where the MIA was 4,000 or 5,000.

From the NTSB preliminary:

The ARTCC specialist on duty subsequently granted the request by issuing the clearance, with instructions to proceed direct to the Initial Approach Fix (IAF) to begin the approach, and to maintain an altitude of 2,000 feet or above. A short time later the flight crew requested to enter a holding pattern at the IAF so that they could contact the Flight Service Station (FSS) for a runway conditions report,

[pattern_is_full]

what doesn't make sense is the clearance "at or above 2,000'" when the plane was in an area where the MIA was 4,000 or 5,000.

My reading is that ATC gave that clearance while the aircraft was still in the southeasterly swath near the coast where the MIA is 2,000. It was heading towards higher terrain, but not yet in that part of the chart.

My reading is also (for the moment, pending further details) that both the controller and the crew lost situational awareness - they didn't notice that the terrain situation was going to change rapidly in the near future.

I don't know for certain what information the controller had (or should have had) in front of him - I presume that will come out eventually. I do know what information the pilot had (or should have had) in front of him - and it is a chart that in TWO places says, "If you are going northeast of ZEDAG, you must be at 4300 feet to remain safe!"

Given two advisories that do not actually conflict:

ATC - "at least 2000 feet or higher"
Chart - "at least 4300 feet"

- which altitude would/should a pilot choose?

This may be of interest - although it does not go very far in clarifying the situation:

A flightcrew should be generally aware of terrain and obstruction elevations and should never accept instructions that do not ensure adequate terrain clearance. A controller is required to issue a safety alert to an aircraft if the controller is aware the aircraft is in a position/attitude that, in his/her judgment, places the aircraft in unsafe proximity to terrain or obstructions. Air traffic controllers issuing specific altitude instructions and clearances accept responsibility for terrain avoidance.

Pilot charts depict minimum en route altitudes, minimum obstruction clearance altitudes, off route obstruction clearance altitudes, and other altitudes. Minimum vectoring altitudes, available only to ATC, can sometimes be the best and lowest altitudes available for instrument flight. Pilots receiving altitude assignments below charted altitudes typically assume they are being assigned the minimum vectoring altitude.

Source: U (http://www.faa.gov/air_traffic/publications/atpubs/atbarc/04-3.htm)

[aterpster]

pattern is full:

I don't know for certain what information the controller had (or should have had) in front of him - I presume that will come out eventually. I do know what information the pilot had (or should have had) in front of him - and it is a chart that in TWO places says, "If you are going northeast of ZEDAG, you must be at 4300 feet to remain safe!"

Given two advisories that do not actually conflict:

ATC - "at least 2000 feet or higher"
Chart - "at least 4300 feet"

- which altitude would/should a pilot choose?

The controller should have had the FAA approach chart in front of him as well as training and knowledge that the FAA 8260-2 holding pattern form specified the same holding pattern and minimum holding altitude as shown for ZEDAG.

If the pilot can actually choose the altitude, based on the chart he should have been not less than 5,400 prior to entering the south TAA area. If the controller decided to clear the flight direct to ZEDAG the assigned altitude should have been not less than 5,000.

The controller had the primary responsibility to not let the flight be below 5,000 if cleared direct to ZEDAG, or not less than 5,400 prior to the flight crossing the boundary of the TAA if it was to be "own nav."

"At or above 2,000" conflicts with the altitudes required to be assigned in the circumstances.

No doubt the crew was also asleep at the switch.

The lawyers will have a field day with this one.

[604guy]

"If you are going northeast of ZEDAG, you must be at 4300 feet to remain safe!"


Interesting to note on the chart that the profile view indicates a 5nm holding pattern at an altitude of 4300'. The plan view however indicates that 4300' in that area northeast of ZEDAG is good out to 4nm only. Beyond that in that direction the plan view says 6300'. Yes that is for straight in and no procedure turn but that is the type of information that would be a strong cue to me that there is likely rapidly rising terrain in the vicinity and that any altitudes provided by ATC less than depicted would warrant a query to verify that ATC will be providing terrain separation at all times.

[aterpster]

604 guy:

Interesting to note on the chart that the profile view indicates a 5nm holding pattern at an altitude of 4300'. The plan view however indicates that 4300' in that area northeast of ZEDAG is good out to 4nm only. Beyond that in that direction the plan view says 6300'. Yes that is for straight in and no procedure turn but that is the type of information that would be a strong cue to me that there is likely rapidly rising terrain in the vicinity and that any altitudes provided by ATC less than depicted would warrant a query to verify that ATC will be providing terrain separation at all times.

Where do you come up with 4 N.M?

[BOAC]

Isn't that what the Jepp chart says?

[aterpster]

No, the inner TAA area is 4 n.m. in the straight-in TAA area (the north side). The holding pattern is 5 n.m, shown in the profile view. The holding pattern would generally be used only when arriving from within the south TAA are (the one with a single altitude of 5,400).

Check the FAA's version of the chart. Perhaps it is clearer in this respect:

http://aeronav.faa.gov/d-tpp/1303/05166R19.PDF

[BOAC]

No, the inner TAA area is 4 n.m. in the straight-in TAA area (the north side). - so if I have this right, on these 'new' RNAV charts, the min hold alt will always be the lowest of the TAA values? NB I have never flown one of these procedures, so, if I am approaching ZEDAG on the 103 for an RNAV19, I assume I maintain 6300 until 4nm when I can start descent to 1900 and begin the turn onto final? If I want to join the hold from that inbound, what then - the same but this time to 4300'? Is that the way it works?
I take it there are no other instrument approaches onto 19 there?

[Zeffy]

No, the inner TAA area is 4 n.m. in the straight-in TAA area (the north side). - so if I have this right, on these 'new' RNAV charts, the min hold alt will always be the lowest of the TAA values? NB I have never flown one of these procedures, so, if I am approaching ZEDAG on the 103 for an RNAV19, I assume I maintain 6300 until 4nm when I can start descent to 1900 and begin the turn onto final? If I want to join the hold from that inbound, what then - the same but this time to 4300'? Is that the way it works?


Not quite.

First one has to remain mindful that only aircraft on the South side of ZEDAG will fly the Hold In-Lieu of Procedure Turn (HILPT).

http://i202.photobucket.com/albums/aa92/zeffy_bucket/PADLTAA_zpsb6f228df.png

Aircraft arriving on the NoPt side do not utilize the pattern -- the TAA has already provided consideration to descent gradient as well as course alignment.
6300' minimum within 30 nm, then 4300 from 4nm to the IF (ZEDAG).

The minimum holding altitude at ZEDAG is 4300, indicated by the underscore.

http://i202.photobucket.com/albums/aa92/zeffy_bucket/ZEDAGHold_zps3d3f6951.png

The 4300' value is only a chance coincidence with the inner NoPT TAA sector on the North side.

The obstacle clearance assessment template for the HILPT is vastly different from the the TAA obstacle evaluation criteria.

[BOAC]

Thanks for the link you sent, Zeffy.

"Aircraft arriving on the NoPt side do not utilize the pattern" does not really answer my question "If I want to join the hold from that inbound, what then?" - I am thinking 'outside' the RNAV procedure as for this aircraft. Should I then abandon the RNAV chart and fly the LOC/DME hold at INDRA? IE Is the ZEDAG hold PURELY for the RNAV procedure and not for 'general' holding?

I see, out of interest from your link, that the ACTUAL MSA inbound ALMOST to ZEDAG from the south-east is only 1800' becoming 4100' at ZEDAG.

Why simple, basic MSAs are not published on the RNAV chart escapes me!

Aterpster - where did Jepp get 2725' for the dominant obstacle from?

[JammedStab]

Why simple, basic MSAs are not published on the RNAV chart escapes me!



You don't want an MSA. If you had one, which goes out to 25 miles, it would be at the higher altitude which is unnecessary. With the magic of GPS showing distance from a waypoint you can descend once past the higher terrain that may be for example, 15 miles from the airport. Plus, in the U.S. the MSA is an emergency only altitude, so the approach would be much more restrictive in terms of letting down within 30 miles.

[BOAC]

"If you had one, which goes out to 25 miles, it would be at the higher altitude which is unnecessary." - not so - see the post you have replied to - 1800/4100?

[aterpster]

MSAs are not operational altitudes in the U.S. Terminal Arrival Areas (TAAs) appear on a fairly small percentage of FAA RNAV instrument approach procedures. TAAs are operational altitudes and are considered the equivalent of feeder routes. RNAV IAPs with TAAs do not have MSAs published.

ZEDAG is the course reversal hold for this IAP (RNAV RWY) 19 and it is also the missed approach holding pattern for the RNAV Rwy 01 IAP.

There is no way that ATC can get anyone onto this procedure at less than 5,000 because of the Anchorage Center's MIA. A pilot entering the procedure must be at not less than 5,400 if entering the south TAA or not less than 6,300 if entering the north TAA.

As to the spot obstacle Jeppesen shows on the RNAV 19 chart, they do that in-house using their own topographical data. The source I have is the U.S. 1:50,000 topo map for that area. I'll post the significant portion of that topo chart.

[aterpster]

This is the Jepp "2750" terrain point. As you can see it is not 2,750...not quite. The FAA chart shows 2,550 at approximately the same location. These cartographic features are not from procedure source; rather they are added by the various chart makers for reference only.
Nonetheless, this is the general area where the BE1900 crash. 2,000 won't get you through that area, and their true altitude was surely lower than 2,000 because of cold station altimeter error.



http://i201.photobucket.com/albums/aa214/aterpster/JeppPADL2750point_zps57eb6733.jpg

[Annex14]

I got to the impression that because of differences for the used similar sounding abbreviations - e.g. MSA - on the two sides of the pond, a bit of a confusions has started and there is talking about two different items. Here in good old Europe it´s all ICAO !!
To solve that, read this:
SKYbrary - Minimum Sector Altitude (http://www.skybrary.aero/index.php/Minimum_Sector_Altitude)

aterpster, the difference in shown heights of the mountains could also come from a different charting base. Normally it all should be WGS 84, but could it be possible that some people still use other geo ref systems or simply stick to old procedures ?

[aterpster]

Annex14:

aterpster, the difference in shown heights of the mountains could also come from a different charting base. Normally it all should be WGS 84, but could it be possible that some people still use other geo ref systems or simply stick to old procedures ?

You make a good point. Most U.S. Geological Survey charts are referenced to NAD 27. So, if anyone is using a paper topo they would not be using WGS 84. The GIS program I use has corrected the bitmap USGS topos to WGS 84. No doubt that the FAA procedures design department uses obstacle data predicated on WGS 84.

But, the chart makers, for terrain point reference data? Beats me.

Having said that when Jepp adds color terrain contours they appear to be WGS 84 compliant.

And, of course, the FAA sectional visual charts are WGS 84 compliant, as are similar charts in developed nations.

[BOAC]

Interesting, when important elevations are not agreed. All other things being equal, if Mr X impacts 50' below the top of a 2750' spot height charted at 2550', is the chart provider liable?

[aterpster]

BOAC:

Interesting, when important elevations are not agreed. All other things being equal, if Mr X impacts 50' below the top of a 2750' spot height charted at 2550', is the chart provider liable?

They would probably be sued, but with a couple of good expert witnesses for the defense, the defense would (or should) prevail. The segment and TAA altitudes are all correct and "shout" on paper to not descend below 4300 in that area. And, it is higher than 4300 if not in the holding pattern, but in that sector or area.

Jeppesen lost at Durovnik because of a shadow box issue with one of the NDBs. But, that had an operational nuance.

[wilyflier]

Please pm, explain "shadow box " , and which / when,.. was that a Britannia in early 60s?
I landed there the next day and visited the site. I seem to remember an unusual approach , (ILS working, but notammed as out of service) , taking 4 approach markers instead of 3, which could have lead to incorrectly crossing them each at the height specified for the next.
They CFIT about the inner marker , querying on the RT about 'ground fog' when they lost sight of the runway , it was a gin clear night and they were below the trees

[aterpster]

wily:

Please pm, explain "shadow box " , and which / when,.. was that a Britannia in early 60s?
I landed there the next day and visited the site. I seem to remember an unusual approach , (ILS working, but notammed as out of service) , taking 4 approach markers instead of 3, which could have lead to incorrectly crossing them each at the height specified for the next.
They CFIT about the inner marker , querying on the RT about 'ground fog' when they lost sight of the runway , it was a gin clear night and they were below the trees

1996 CFIT of US Air Force 737 carrying Clinton's Secretary of Commerce Brown. Because of the war the ILS was missing in action. So, the approach was a dual NDB approach. Jeppeson designates the primary nav aid with a shadow box. That was messed up in this case, or at least so alleged in the lawsuit that followed.

It's been a long time. I am sure you can Google it.

[galaxy flyer]

Actually, I think they stuck terrain beyond the airport. The 737 flown only had 1 ADF onboard when dual ADF was required.

Back to AK.....

[pattern_is_full]

All other things being equal, if Mr X impacts 50' below the top of a 2750' spot height charted at 2550', is the chart provider liable?

That would be why the FAA adds a 2,000-ft "cushion" in the published minimum safe altitudes - that, plus the possibility/probability of altimeter calibration errors, or imprecise barometric information, etc.

As in this case, where the high point was 2,250, and the minimum safe altitude was 2,050 feet higher (4,300').

[BOAC]

As in this case, where the high point was 2,250, and the minimum safe altitude was 2,050 feet higher (4,300'). - I think you are looking at different charts to the rest of us. Try ours - FAA or Jepp, which ever you believe.

Try 2550' (or 2750') to taste, and 6300'? (or 4100', or....). Nice and easy, isn't it?

It would appear that the TAA figures bear no resemblance to 'min safe' altitudes, since according to the FAA LOC/DME R19 it is 'safe' at 4100' over that peak.

[aterpster]

gf:

Actually, I think they stuck terrain beyond the airport.

The arrow indicates the point of impact:

http://i201.photobucket.com/albums/aa214/aterpster/LDDULctrRwy12_zpsfff71728.jpg

[aterpster]

BOAC:

- I think you are looking at different charts to the rest of us. Try ours - FAA or Jepp, which ever you believe.

Try 2550' (or 2750') to taste, and 6300'? (or 4100', or....). Nice and easy, isn't it?

It would appear that the TAA figures bear no resemblance to 'min safe' altitudes, since according to the FAA LOC/DME R19 it is 'safe' at 4100' over that peak.

TAA "apples" and holding pattern "oranges." The outer TAA area of 6,300 covers vastly larger topography than does the holding pattern. In any case, the TAA inner area is 4,300, the same altitude as the holding pattern, because the same terrain controls for both the inner area and the holding pattern.

Although the 6,300 area goes out 30 miles for the pilot, the obstacle assessment area extends out an additional 4 miles. That is a lot of real estate.

[pattern_is_full]

OK, my bad - the FAA cushion is more like 1,500 feet.

That doesn't change the important part of my point - if the pilot stays at the charted safe altitude of 4300 feet (or even 4100 feet), the disparity between 2550 feet or 2725 feet as the highest terrain point is really irrelevant.

if you're so low that a geographical elevation error of 50, or even 250 feet, makes the difference between safe passage and a crash - you're far too low in any event. At least in IMC - we'll acknowledge that there are visual approaches that cut things that fine.

[BOAC]

if you're so low that a geographical elevation error of 50, or even 250 feet, makes the difference between safe passage and a crash - you're far too low in any event. At least in IMC - we'll acknowledge that there are visual approaches that cut things that fine. - not the point of my post! Don't think anyone would disagree. Is the Pope a Catholic?.................and I won't ask about bears.......................

Edit to add: Incidentally OK, my bad - the FAA cushion is more like 1,500 feet. - I suspect your 'cushion' will be 1000' even in FAA land, so I assume there is a 3100' peak somewhere within 25nm north of Dillingham.

[galaxy flyer]

Thanks for the chart, aterpster.

[DownIn3Green]

The DBV Approach is almost like the one in Trabzon, Turkey...When I flew a jet (727) full of passengers into there at night and/or low vis conditions...I always kept in mind one thing...If everything starts to go south...TURN RIGHT!!! Over the Black Sea...Maybe not correct, but if you lose situational awarness...at night, in IMC, Over the sea is better than towards the mountains....

As has been beaten to death here on this thread by many of us "greybeards"...if you're not sure where you are at, IMC or at night being at 2,000' is not the time or place to try to figure it out....

[BOAC]

because the same terrain controls for both the inner area and the holding pattern.- my post was about ACTUAL MSAs, not some 'TAA' figure - out of interest, why 4300'for TAA when MSA is 4100'? What are the criteria for TAAs?

[Zeffy]

TAA criteria are in Vol. 4 of Order 8260.58 (http://www.faa.gov/about/office_org/headquarters_offices/avs/offices/afs/afs400/afs420/policies_guidance/order_notice_ac/media/Order_8260.58_noCalculators.pdf)

FAA AFS-420 web site hosts all TERPS Orders on this (http://www.faa.gov/about/office_org/headquarters_offices/avs/offices/afs/afs400/afs420/policies_guidance/order_notice_ac/) page.

[BOAC]

Thanks again, Zeffy - I could not find any reference to the ROC for TAA procedures (mind you, I got dizzy with all the maths................) Overall it seems odd that the hold at INDRA (1 nm inside ZEDAG) is considered safe at 3800' and yet when we switch to TAA procedures it jumps to 4300'. I assume there is a different (increased) ROC since the terrain is no less 'challenging'.

Incidentally, I see that 8260.19E at 2.74 calls for 200' to be added as an AAO which may account for the 200' discrepancy we see between the FAA and Jepp chart 'spot heights' (basically to allow for the 199' mast), and 2750' + 1000' gives us the 3800' hold, I guess.

Just glad I never had to construct procedures..........................

Sorry for the diversion and back to the accident.

[Zeffy]

Yes, the criteria are not for the faint of heart.:)

In a very small nutshell: ROC values for TAAs are the same as for airways.:ok:

[aterpster]

BOAC:

Incidentally, I see that 8260.19E at 2.74 calls for 200' to be added as an AAO which may account for the 200' discrepancy we see between the FAA and Jepp chart 'spot heights' (basically to allow for the 199' mast), and 2750' + 1000' gives us the 3800' hold, I guess.

Chart makers don't use AAO for reference obstacles.

But, the designers made additives to the holding pattern's minimum ROC of 1,000, probably for precipitous terrain and, more important at this location, cold station altimeter errors.

[BOAC]

Chart makers don't use AAO for reference obstacles. - I was talking ROC not reference obstacles, and Jepp seem to have added 175' (to the reference obstacle)

[aterpster]

BOAC:

- I was talking ROC not reference obstacles, and Jepp seem to have added 175' (to the reference obstacle)

I was speaking of both in the total context of this aspect of the RNAV 19 procedure. I have no idea how Jeppesen came up with the value they show.

[Zeffy]

NTSB Report (http://www.ntsb.gov/aviationquery/brief.aspx?ev_id=20130308X64149&key=1)

Accident Docket (http://dms.ntsb.gov/pubdms/search/hitlist.cfm?docketID=54667&CFID=341263&CFTOKEN=49906034)

ATC Factual (http://dms.ntsb.gov/public/54500-54999/54667/541308.pdf)