I always turned the egpws off once visual in the valley at TGU Honduras. It always started going off on base leg next to the hills so was a distraction. It is much easier to turn it off at 3,000 ft rather than 400 ft on base leg. TGU made the second most dangerous airport airport in the world last week on that 2 hr history channel program last week. Once you see the terrain you don't need it any more. I wonder why they flew into terrain. It couldn't happen at TGU because everybody was looking out the window, not fiddling with automation.

C-C

What you have written is inflammatory rubbish for the new generation of Airbus.

Of course Boeings never get map shift! Well they did when I flew the 757/767!

The Question is: can you disable IRU updates on your Glass Boeing or y. Glass Airbus?
 

Iceman50 raised a bullshat flag. Good. Since I have never flown a Glass Boeing I wasn't aware that 757/767 had 18 mile map shifts on the procedure turn.

But I stlll need the questioned answered: on Glass Boeings can you disable automatic nav updates after you I.D. the primary navaid and commence approach? Of course in the developed world, maybe the navaid maintenance is good enough that you don't need to do this.

I have already explained that the first airbuses were not really designed for disabling auto updates; and initially, with many carriers, this precaution was not taken.

Although map shifts are increasingly rare in the Western world, they are not unknown in the developing world. I keep citing the first Airbus because that is where Toulouse cut it's teeth on Glass. IIRC, the honeywell FMC box would take a polling of all three IRU's and then average their position for the new center IRU position. Then it would skew the map with updates based on a complex algorithm. If one IRU update position was skewed 50 miles the wrong way because it decided to update itself off a VOR in a war torn country which was overheating, the whole moving map central position might get moved 5 to 10 miles the wrong way (depending on geometry and other logic).

Right?

Yes, you're supposed to keep the runway in sight. but if your moving map says you are close, it is a prelude for overconfidence. It was well documented in July, and August 1995, I believe, AW&ST that Airbuses where designed to reduce crew workload but instead, vastly increased it.

Last minute runway changes are the devil for a green First Officer to deal with. A new F/O will be damn lucky to get all his crap done before base turn. His head is now down pushing buttons and nobody is watching if his Airworthiness is even still alive over there. With that cursed sidestick he's already slumped down on the armrest like a patient donating blood at the Red Cross. How would the pre-occupied F/O even know if he was having a stroke? Who's watching him?

Only ATC is watching him, and one poster said they tried to save him.

Who's listening to the morse code?

Nobody.

Who's looking at the raw DME to see if it jives with the Moving Map?

Nobody.

Who's looking out the front windscreen while the old boy is looking over his shoulder behind him trying to reacquire the threshold?

Nobody

Crunch

Note: All my posts are just my opinion only.

Crash site location thanks to Machaca, aterpster, BrooksPA-28. Brooks, your lat and long are probably the more accurate - this is a small amount further West, now that I look at it. The lat/long in the image below is in decimal degrees. Machaca's posted photo below, to compare.

The chart from aterpster showing the protected circling area indicated is helpful if only to see how far outside the circling area the crash site is. Wally Roberts' discussion of FAR and ICAO PAN-OPS Circling Procedures is well worth reading and may be found here, (1996) and here, (1997).

Cap'n Crunch, well said.

PJ2


http://i277.photobucket.com/albums/k...clingArea2.jpg



http://i277.photobucket.com/albums/k...clingArea1.jpg


http://i277.photobucket.com/albums/k...777/202-04.jpg



http://i277.photobucket.com/albums/k...-28_214751.jpg

PJ2
Thanks for bringing some good factual information to this thread. There has been far too much (yet again) idle speculation from too many of the contributers.

Captain Crunch
Please leave out the Airbus/Boeing red herrings! Map shifts can occur on any aircraft and are irrelevant to this debate. The whole point of a circling approach is to remain visual with the airfield and NOT to follow any magenta lines. Once the approach path has broken off from the ILS (or other electronic guidance) The pilots MUST remain visual.

Circling Approaches
 

AT GF we just don't do circling approaches.... and with good reason !!!

PEI 3721:
When the Honeywell EGPWS was being certified, the senior Honeywell engineer on that project tried in vein to make a GPS sensor for the unit mandatory for otherwise non-GPS aircraft. The FAA was more strongly lobbied by a major, cheap airline to not require the optional $1,500 GPS add-on for non-GPS aircraft.

Absolutely correct.

Problem is, IMHO, that many airlines (mine included) train for such procedures in simulators where you lose sight of the runway going downwind and only pick it up again when you're halfway round the turn onto finals. The display of ground/terrain is rather half-hearted as well. The end result is that you fly it partly on the map/DME, etc. because there's little else to refer to, so it becomes a hybrid instrument/visual manoeuvre rather than a purely visual one. Not good procedure.

If you do plenty of this kind of approach in real life, no problem, but if you were not current and forced into circling at short notice in bad weather, you might try and make it easier by using the FMC/map "as a backup for SA", maybe even drawing a little circuit on the screen. Easy to press '7' when you meant '4' as they're next to each other...

@CPT-Crunch

Very well said!
That's why in our airline during NP App one pilot has to use the raw data (rose) display.

Anyhow, I'm not sure that MAP shift was the problem with this accident. Didn't they say "rwy in sight", not to mention what to do during a circling app if you lose contact.....

Dessert Aviator & Circling Approaches
 

Sooo why do you call yourself "Dessert Aviator" ?

More suitable name would be "Dessert Autopilot" !:D

PS.

If for whatever reason, one day,
you stay without instruments while "aviating",
what shall you do.....?

Land with any tailwind (can be 60 kts or more) or stay in the air forever ?

Thanks PJ2,

Good to see you again. Nice Graphics w/Machaca.

Thanks guys for entertaining my unpopular anti-automation position. Bergerie1 makes a valid point that I shouldn't contrast too strenuously Airbus verses Boeing; but I maintain that they are in fact different. My indictment is really against over-reliance on glass and automation when manual flying skills are called for. But distractions can always happen when ATC issues a wind/runway change close in. At some outfits, the tyranny of company standardization will require the non-handling pilot to break out the laptop and recompute the landing distance, redo the ref, rebuild a centerline on the FMS, set up the missed approach while simultaneously talking to ATC, and participating in config changes, etc, etc, etc....

Of course, on a tight circle-to-land a new guy won't get half that stuff done right. If the crew didn't pre-brief the circle 100 miles earlier, it may be impossible to do. Even if he's good, with a patten that tight, he might not "rejoin" you until short final. I guarantee "Bergerie1" he won't know which way the aircraft is pointed when his head comes up. He also won't know if the handling pilot has lost the runway unless the handling pilot orally states that. Only the guy on the left side of the airplane is going to be able to stay visual with the Threshold if it's a left hand circuit.

This accident so far reminds me of the Wamea, Indonesian BAe-146 accident not long ago. That crash was not an FBW Airbus, so that should make the stickjockeys forgive me a little bit for my scorn of the joystick. :E

CC

- as far as I can tell, unless 'Airblue' have their own TERPS charts, the circling radius is PansOps.

While c-c has made very valid points regarding 'uncontrollable' and 'unmonitored' automation, as I have said before there is something quite odd about this crash in its inexplicable distance from the field, just like the Yemeni at Moroni a while back. There we appear to have had another strange 'excursion' outside the normal area.

Have we seen that report yet?

The terrain visualisation for this crash (allowing for its limitations) does give me some puzzles in respect of the hills and valleys en-route to the site and I find myself wondering how they impacted where they did - it is almost like some 'hill and valley flying' that went wrong - aka 'letterboxing' - but why? Do we know if the ILS DME was working?

Yes...... so you tell ATC you need 5 mins, hold somewhere safe and re-brief your imminent challenging wx affected circle to land. You don't just plough on with the PNF working like a one armed paper hanger!!

It's called non-technical skills. Workload management, cooperation, situational awareness. These are essential elements. A lot of discussion so far has talked of technical aspects (did this aircraft have GPS?) and possible mapshift. The only reason I can see that they ended up 7 miles north was a breakdown in the non- technical side of things which would of course be exacerbated by having a brand new FO in the other seat.

A4

Hi,

you can get some pretty interesting tailwinds when on the ILS to runway 30.

Here is some data from ISB airfield.

Elevation approx 1700ft.

Approx bearing of 320 degrees and 12nm from the airfield is a peak of approx 3900ft.

Landing dist on Rwy 12 is 3017m. Landing dist on Rwy 30 is 2743m. PAPI on both runways. Very basic approach lights avail on Rwy 12 (H-S).

Second half of runway 30 is downhill.

For a Cat c aircraft the circle to land is 850ft agl and with a vis requirement of 2.4km (not sure if aircraft is cat C or D).

Dont fly over the city to the south or through prohibited area 254 which is south also.

We have a note in our manual regarding possible GPS outage due jamming in ISB.

Heavy rain can cause mud on the runways and the obvious dangers of rubber deposits.

Its seems the wind favoured the circle to land in marginal conditions.

Rgds.

I just had a look at the plates, not having operated to ISB for a few years, and the place cries out for some sort of instrument approach for RW12. You probably could engineer a VOR/DME but it looks ideal for an RNAV: just come in from the W or SW and intercept the centreline about 6nm out, normal 3deg should suffice...

C-C and A4 - you are risking 'frightening the horses' here with your tales of what dreadful tasks PNF would have to do to circle onto 12.

1) If you can land on a wet 30 with a tailwind you will be ok on a wet 12 with a headwind - therefore no re-computations

2) You do NOT need anything re-programming - a simple fix on the field and 2 mile circle if you must will more than do for a VISUAL manoeuvre - plus the 737 at least has a 'built-in' dotted white runway c/line in BOTH directions
3) G/A + missed does NOT change!!!! Do not confuse the great unwashed please:ugh:.

If the P2 was ex mil and knew Islamabad, is it possible the mil had some sort of 'in house' proc for 12?

the FO was a ex mirage pilot which is now confirmed. and also that he had been in Airblue for 2 year and therefore not a new FO

i was looking onto the circle to land procedure for the A320 series,which stated that when you come abeam the threshold u start timing 20 sec for 500 feet and then make a turn for the base and then finals.what does 20seconds for 500 feet mean relative to this incident that has occured.Please shed some light on this matter since the previous post tell me that he flew 7nm out.

Nothing significant or new in the Pakistani papers.
ATC apparently warned pilot of getting too close to high ground to which the captain said "we see that". If he had the hills in sight why would he fly into them?

I don't think these are the actual words spoken. Probably a paraphrase.

An ILS on 12 would be good of course but they won't do that even now since they are building a new airport out west near a city called Fatehjang. No high ground around in that area.

pilotsaab

The procedure is 20 secs per 500' AGL, corrected for wind. If 10kts tailwind component on the downwind, remove 10 secs from your timing. Flying downwind at the circling minima (860' AGL), when you get abeam the threshold start your stopwatch, fly for 25secs and turn base. Your config should be Gear down & Flap 3.

7 miles is way too far downwind. If this is a accident whilst circling it bears so much similarity with the Air China B767 crash at Pusan, Korea

Ah, BOAC,

You'd make a great Chief Pilot because your logic is good. :ok: Unfortunately, they'd probably fire you for speaking such legal department heresy as not computing the landing distance like a good little line soldier (on some properties.)

I like your idea, and that's what I'm angling at: reduce meaningless cockpit chores, paperwork and standardization babble. At some places, the number of required briefings is absolutely ridiculous now. A4 has the politically correct training answer: Go into holding and start over. But I prefer briefing the circle to land and dispensing with as much distracting garbage as possible before let down. If you don't work for an airline where the Captain has authority to dispense with some of this happy horsechit down low, then you risk being turned in by your F/O for being non-standard.

BTW, when I flew the old bus, even though the MA is the same on Circle to 12, you still had to manually reset heading (by re-centering the blue heading reference AFTER rolling on to final), and THEN turning the heading bug to the right to 240 degrees (preset), or believe it or not, that old airbus would remember a left turn to MA heading where you originally set it and turn flight directors the wrong way! Boeing knew to take the shortest path to the heading preset. Airbus did not. One more friggin distraction!

You also had to check and see that the MA was still in the FMS, IIRC. Sometimes it just didn't stay in there with the runway change. So all this automation actually loaded up the crew where everything was "azzholes and elbows" on a tight circle to land. We finally outlawed them.

I know what I'm saying sounds silly, but early glass Airbuses were very, very different from steam-gage boeings. Having flown both, I can assure you that the FCP autoflight had more bugs than a bait store. Every time a new operator went into a Bob Hoover tail slide, I was not surprised at all.

Don't get me wrong. I loved hand-flying the airplane. I just couldn't trust it's automation, it's nav display or it's flight guidance, that's all. :}

Now another poster says the F/O was a two-year veteran, so if that's true, please disregard my rookie F/O theory.

Crunch - out

Map shift has nothing to do with this accident if a circling approach is involved.

For circling into ISB, is done as a procedure (as all circling approaches are done), commences from a point on an ILS (VOR), thereafter the landing runway must be in sight and visual, thereafter, it is a timed visual procedure, flown with a constant speed and configuration. The landing runway remains visual and further descent from circling altitude may only commence on base provided sufficient cues remain available.

If the MCDU's fail, or the FM's fail, (no map available/map shift, you name it) this circling procedure can still be flown in the exact same way as it has always been flown, or will be flown.

Having circled into ISB with A332 myself, it is no problem, and I have flown with former Air Blue pilots in my current airline. That being said, if you forget to turn downwind from the initial offset of 45 degrees from runway track, you might land up into those hills, which means the circling procedure and visual sighting of the runway never happened and the procedure was never adhered too. This was a very sad day for all of us mere mortals.

While recognizing that Circling to Land is a visual maneuver, could this crash share aspects that are reminiscent of Addis Abeba -- i.e., EGPWS sans GPS?

Thunderbug:
There are similarities, yet great differences. In the current accident the protected airspace was a generous PANS-OPS radius of 5.2 nm. At Pusan, the captain thought it was PANS-OPS, but it was tiny/little, politically motivated U.S. TERPs. The Chinese captain at Pusan got had in my view.

JetJock330,

You might be right; you think he forgot go parallel to the runway? :confused: We need a flight track leaked to wikileaks... Anybody?

And one point: A map shift is not likely a failure of the airplane. It's caused primarily by a failure of one of the ground stations sometimes even hours before letdown. The update system is doing what it was designed to do. But the shift might be insidious when you've got your hands full.

Q: How do you disable all auto updates in the A330?

I bet if you try it sometime you will be surprised how hard it is to do (since A/B doesn't want you to ever do that, if my hunch is correct.)

CC

Greetings Crunch

I can see why you think my answer is "politically correct". We live in a litigious society and those of us in the LHS must always remember that. We are responsible and answerable for our decisions and actions. If we choose to rapidly rebrief "on the hoof" for an unexpected circling approach in cr@p wx and something goes wrong, if you survive you better have a good explanation. I agree that there is no need to recompute landing distance on a 2700m runway but on the 'bus you do need to ensure the Secondary flight plan is set up correctly (for groundspeedmini computation). This doesn't take long but could be a distracter if PF is monitoring an inexperienced FO to ensure he's doing it right ....... and who knows, suddenly your 7 miles on your ILS break-off heading......

I mentioned non-technical skills previously. We need to recognise when it is and is not appropriate to press on or "take 5" and ensure you're ready. I've seen it in the sim numerous times - getting the balance between prioritisation and maintaining standards a bit skewed.

It doesn't sound encouraging regarding the release of factual information on this one which is a pity because it looks like there could be some fundamental lessons we could learn.

A4

Aterpster, re #127. Yes, Honeywell have promoted the use of an internal GPS card, but Airbus chose a different route.

The preferred Airbus option was for an external GPS/Nav sensor (MMR), which improved navigational accuracy and EGPWS performance (Operators Information Telex 5 Feb 04). Furthermore, at that time, they would not certificate “a solution based on a GPS card”.
Whilst this policy appears sensible in that it provides an accurate navigation (and terrain) map and thus should aid pilots to avoid situations where EGPWS may activate, the more complex (and expensive?) installation might result in fewer aircraft updates.
These aspects together with the apparent difficulty in isolating unsuitable navigation sensors might show as a difference between aircraft types.

Safety estimates circa 2000 comparing high vs low nav accuracy input to EGPWS, showed factor of two in favour of the high nav solution.
I have seen similar data for unmodified Airbus aircraft (2006) where the difference was also estimated to be two. I doubt that many aircraft would have been upgraded subsequently in the modern commercial climate. In addition, safety estimates considered the distribution of suitable ground based nav aids, the lack of which (at that time), further increased the risk of CFIT (associated with ‘map shift’) in less well equipped geographic regions.

Thus the lack of a high accuracy navigation input to EGPWS, the difficulty in isolating systems, and the very compelling features of map displays (our addiction to modern technology) significantly increases the risk of CFIT.

Several operators appear to have recognised these hazards and introduced procedures and training to mitigate the threats; noting that procedures and training are soft barriers.
Unfortunately humans forget (complacency), aircraft are sold on, but not necessarily with specific operational procedures and training, and humans remain vulnerable to the attraction of ‘salient’ displays, especially in stressful situations, and may overlook the most basic of safety behaviours when circling – staying visual.

EGPWS is one of the best (if not ‘the’ best) safety system, but it, like any other computer (and humans), depends on the accuracy of the inputs.

Whether these aspects have any relevance to this accident or not, it would be worth checking how current aircraft are equipped (first line of defence) and thus how they should be operated (last line of defence).

PEI 3721:

That is a better solution. Honeywell was simply trying to prevent "Brand U" airline from using DME/DME on their pre-GPS airplanes instead of GPS. The minimal GPS sensor is far, far better than DME/DME.

Things that should've helped avoid the tragedy:

EGPWS - And associated memory items/actions. Did they switch it off? ignore it? was it not working?

ATC - Asking crew to turn immediately, ignored because crew thought or lied (hey, some pilots do) that they were visual.

Early days but evidence so far points towards this accident being caused by very experienced (and over-confident) pilots who let the rules slip because they were content with their situation. This has always been and will be the death of many! :ugh:

:eek:

I try to avoid too much 'AB bashing', but fer xxxx's sake- is it THAT bad? Can't you just fly it like an aeroplane? 'Ground speed mini'??? Don't you have an ASI? Does that mean that with an FMC failure you are doomed to crash somewhere?

I think I need stronger tablets - NURSE!

The Secondary would/should have been set and briefed prior to descent.

It seems to have air data problem.
As rain was reported ,some times Water in pitot tubes do not drain properly which can lead to un-reiable ASI or ADR failure

it seems they encountred sudden down drafts while crossing the ridge.
and we also can consider the altimetre settings was correct or not

- now, I am prepared to risk being accused of prejudging this accident, but I do think that judging by where the runway is and where they crashed - yes.

This is all becoming rather odd - now the chaps had water in their tubes and/or were unfortunate to encounter a down-draft over high ground below MSA outside the circling area? This thread is sadly going the way of all PPRune threads.:ugh: What about being shot down by the Taliban? We haven't tried that one yet.:{

Beautifully understated. :cool:
p.pilot will doubtless support his analysis anon ... :hmm:
For p.pilot: Oddly enough, most aircraft fly in rain with no hint of trouble in re the pitot tubes providing input to the systems requiring data from them. From where do you derive the assumed failure of redundant systems? :confused:
I will now ask how you relate this assumption of an anomaly to a pilot flying a visual procedure: circle to land from an instrument approach, which was already pointed out in this discussion more than once.

Did you read the entire thread before offering your analysis?
OK, maybe the Taliban shot it down using a bacon powered death ray. :E

Now that we have that out of the way, we can return to the discussion about weather, flying, circling approach, and the recently (and prematurely) departed folk on the Airblue flight. :{

BOAC,

To the Taliban story you can add what is coming out in the local papers that the captain had just been released from hospital for various ailments and was too old to fly a 'glass' fly-by-wire jet.

No cover-up
 

I have little doubt that the CVR and FDR will provide the necessary information to explain this accident.

Pressure must be brought to bear on the Pakistan CAA to release a full report. To do otherwise, as has been suggested in earlier comments, would be negligent in the extreme. (I assume the Pakistan CAA is part of ICAO.)

Yes, along with another 182 sovereign states.

This might be a case of "who pays the piper, calls the tune", and one country comes to mind!

mm43

As has just been confirmed, yes, Pakistan is a member of ICAO. However, in 62 years, there has been no published, or publically-assessible air accident report. PIA pilot's are given an internal summary memo of the salient facts and/or safety recommendations, but that is strictly confidential.

PIA's no.1 pilot in seniority Capt. A.A Khan flew his 720B into the ground on approach on the inaugural flight to Cairo. No one knows why.

My class-mate Capt. K.Wyne was departing Jeddah one night in a 707-320 and it is rumored he had an in-flight cabin fire. He turned around but did not make it. Crashed in Taif.

Capt. I. Janjua flew his A300B4 one step ahead of the step-down Sierra One approach into Khatmandu, Nepal. He hit a ridge. A classic CFIT from which valuable lessons could have been learnt.

I mentioned that very senior commander Capt. Siraj in the 747-200 who forgot to put his gear down landing at Islamabad. Why? All I know is that Capt. Siraj retired.

Finally, an F.27 Fokker Friendship departed Multan on a very hot day when climb performance is degraded and got an engine-fire just after lift-off. He crashed in a field. It was rumored that the aircraft was over-loaded. Since some big shots from the judiciary got killed, the government grounded the entire Fokker fleet. We now have the ATR42.

Not to mention the most dramatic crash of all, the VVIP C-130 of General Zia with the US Ambassador and others on board. It took off after they had witnessed a tank display in central Pakistan and after experiencing several phugoid ossilations [forgive my spelling if wrong] plunged 90 degree nose-down into the ground. Lockheed came to assist; since the US Ambassador was killed the FBI came to assist too. To this day his son asks where is the report!

I will apply pressure, speak to others in and out of government but the struggle could be in vain. I suppose the ICAO cannot force government to publish. What a pity. They should expel (or suspend) Pakistan from membership of the organization for lack of transparency and actions prejudicial to air safety.