Peocedures for lost altimeters

[woocash]

Hello all,

I am a trainee, learning to get S-ATCL license here in Poland.

I have one question that i cannot get a clear answer to: are there any procedures for ATC to follow, when the pilot reports no altimeter(s) reading?

I suppose ATC cannot separate vertically and emergency should be declared by the PIC?

[Spitoon]

Can't think of any particular procedures for complete loss of altitude info. I've come across problems with altimeters giving different reading and the pilot not being sure which one is giving the correct info. In this case cross checking with mode C data is the best solution.

If it's a case of how to get an aircraft down safely without altitude data I'd try to get the aircraft into an area without obstructions - ideally over the sea, but I'm from the UK and this is more practical than in other parts of the world - so that the pilot can descent into visual conditions and then vector or whatever toward the nearest suitable runway. This will work in all but the worst wx conditions. Obviously I'd keep the problem aircraft away from other aircraft in the area.

[BOAC]

With Spitoon's help, since most a/c have a radio altimeter which would normally start reading at 2500', recovery should be possible with care. Try to direct the a/c (if you can) to an airfield with no significant terrain issues on the approach and near to the same altitude as the surface below the aircraft to make life easier.

[Daysleeper]

In this case cross checking with mode C data is the best solution


There is no guarantee that Mode C would be accurate, it is after all fed from the same air data sources as the altimeters.

You might want to scramble your air defence QRA and have them fly wingman, thus reporting the actual altitude.

[Spitoon]

Daysleeper, I was making the assumption that it was the pilot that had lost access to the information. Obviously if the controller has doubts about the accuracy of the mode C data there are well established procedures to follow. But if the pilot has conflicting info and there's nothing to suggest that the mode C is inaccurate then a cross is the best way to pin down which altimeters can be relied upon. And even in the simplest aircraft fit the encoding sensor is separate from the altimeter display in front of the pilot.

BOAC's point about RADALT is good. When I offered my first thoughts I had in mind a less sophisticated avionics fit but in such a situation every bit of help is useful.

I avoided woocash's question about declaring an emergency originally because I didn't have a lot of time but I while it's clearly up to the pilot to establish the severity and seriousness of the problem I'm not sure that it will always count as an emergency. In good wx without additional problems navigating I would class it as an unusual event requiring special handling (I'm speaking as a controller). In solid Cat III conditions at all practical landing sites it's a very different situation. Certainly the pilot should tell the controller - good Team Resource stuff to call on anyone who can help - and then I would expect a solution suitable to the conditions to be worked out.

[BOAC]

Actually, Spitoon, unless I have misread your post, 'solid CATIII' does not make the problem much worse, since the radalt will be used by the autopilot for the approach.

It is definitely a major emergency for the a/c crew unless VMC/day and is extremely demanding in terms of checking routing and flying at the correct altitude in IMC.

There are, in addition, other little tricks for pilots like the ground return on the radar which can give a coarse indication, not forgetting GPS altitude where available and for both sides, mil height finders..

[Spitoon]

Interesting points BOAC. You're right that getting down the ILS in solid Cat III is not a problem but from the controller's perspective getting the aircraft onto the ILS safely would be my concern at most airports. Given that RADALT only works at low levels - perhaps below the normal safe intercept level - and airports tend to be in 'terrain rich' environments, from a controlling perspective it could be somewhat challenging.

As you say, in some conditions it will be a major emergency for the crew, but in others less so. As a controller, if I were faced with this situation I would do everything I can to ensure that the flight can be continued without the crew needing to be concerned about maintaining a level accurately. A combination of VSI and mode C (I'm assuming no major failure of an ADC in a sophisticated aircraft here) should be good enough if the levels above and below the problem aircraft are kept clear. If it's a major emergency I presume that the crew would be looking for a nearby suitable airport at which to land so continuing the flight and navigating become less of an issue.

But as I say, some interesting discussion points. And everything that I've offered is based on a well supported ATC system - because the question was about ATC procedures. For a pilot flying over some parts of the world, I imagine there would be a whole range of additional considerations!

[woocash]

Thank You all for replying.
In concusion it seems You have to rely on Your own judgment and creativity.
In some polish forum, one atco also send me a link to Aeroperu 603. On wikipedia there is an information that the controller sent a B707 to guide Aeroperu down safely.
Would You (as an ATCO) approve such flight? Would You agree to willfully loose separation for them to see each other? Would You do it if the pilot asked You to?

[Denti]

Spitoon, mode C information on commercial airplanes is exactly the same as displayed to the pilot as long as the altimeter subscale is set to standard. We can switch the altimetry source for the transponder from left to right and in my outfit it is usually set to report the information from the flying pilot. It is not independent from the two main altimeters on the left or right PFD, the only independent third altimeter on the flightdeck is the standby altimeter.

To use Mode C as a crosscheck is somewhere between useless to dangerous if it instills confidence in the malfunctioning air data computer simply because mode c is still the same as the displayed information.

It is a different thing in smaller planes which have independent blind encoders for their mode c altitude read out where mode c can indeed act as a third independent altitude information.

[Spitoon]

Thanks for that Denti - last time I had to deal with anything like this I was told that encoding sensors were independent of displayed data because it mitigated some failure modes.

[BOAC]

Given that RADALT only works at low levels - perhaps below the normal safe intercept level - don't forget rad alt is reading height AGL, independent of QFE, or QNH etc so if it reads 2000' you are 2000' from the ground (at that moment). Most pilots would be more than happy with 2000' vertical separation from terrain for intercept.:)

woocash - formating on an unknown aircraft is fraught with danger but is a good last resort. If you have access to a mil pilot you would stand a better chance.

[sycamore]

You could also get the aircraft to be vectored to cross a known ILS at a given distance,ie,20nm,should be 6000ft...

[BOAC]

Not much point if there is no altitude information available to the crews, and there is always the possibility they might fly into the ground before they get there!

[NigelOnDraft]

Spitoon... See 757 Accident Link (http://en.wikipedia.org/wiki/Aeroper%C3%BA_Flight_603)

In this case, lack of systems knowledge by both Flt Crew (who maybe should have known?) and ATC (less likely) meant that Mode C readout (relayed by ATC) was given credibility, yet of course it was suffering the same error, and played it's part in them flying into the sea.

last time I had to deal with anything like this I was told that encoding sensors were independent of displayed data because it mitigated some failure modesDitto - another cause to be careful! Even if these "encoding sensors" are independant, unlikely they will have a separate static source.

In short, as per ATC practice to verify Mode C, I would suggest it is the least reliable indicaiton of altitude.

2 examples:
1. Homebuilt I fly has 2 altimeters and a Mode C/S Xpdr. All 3 have their own static inputs, but are all untilately connected to the same static ports.
2. On an Airbus, Xpdr is driven from 1 of 2 sources - which in normal use, is the source of the altimeter the AP is using*.

The logic for this is strange - it is done to ensure that where the altimeters differ somewhat, the Mode C reports the same FL as the AP maintains, and ATC do not whinge. Should that altimeter (ADC) be in error, then both the aircraft and Mode C will drift off, and nobody is the wiser i.e. the practice is to stop inconvenient ATC reports pointing out the ADCs seem to be different. Nobody seems to mind the fact the aircraft is actually at a different FL ;)

NoD

[ShyTorque]

Having taken off from a private HLS in the local area, I was once flying level at 2400 feet under the "stub" of controlled airspace (CTA which began at 2500 feet) and had asked for a flight information service. Transponder set to 7000 / + Alt.

After giving a position report, ATC immediately asked us to confirm our altitude. I replied "2400 feet", as we were, on their QNH. They asked us to squawk, which we did. They asked us again to confirm our altitude, sounding stressed.

"2400 feet!" we replied.

ATC replied "Your mode C is reading 7500 feet and climbing!"

Us "(Click). How now?" (changing transponder select to read from altimeter 2).

ATC "That's better, you are now reading 2400 feet!"

Posting this because our altimeters were both reading 2400 feet at all times.

Mode 'C' can fail, too.

[woocash]

I'll ask for clarification.

If all altimeters fail, the crew is unable to reliably detemine their level. How should i separate other traffic? Assume, that a/c with malfunction can be at any level between gnd-460 (top of our FIR), or maybe ask the pilot to guess their altitude if in vmc? He/she wont be able to tell FL360 from FL380, but FL380 from "somewhere around FL150" should be doable (with huge error, but still it would allow me to treat them as being under any other traffic")?

[BOAC]

How should i separate other traffic? - laterally. Whatever altitude they might be at now they may well not be in 5 minutes!

[Agent86]

Some lateral thinking. If the Pitot-static system is inop every pressurised aircraft has some form of cabin altitude (generally logarithmic in scale above 10k) and cabin rate. Depressurise the aircraft ...If the masks drop out you are above 10000! Descend until the cabin alt gives a valid reading then use the radalt on approach.

[Tarq57]

Saw a situation a few years ago where a commercial turboprop had an altimeter failure.

This was before transponders were used in NZ.

One static source had iced over, the other altimeter malfunctioned. A pure (and unusual) coincidence. The a/c was in IMC.

(This, of course, by the definition in MATS, is an emergency situation. I declared an incerfa via our ACC supervisor.)

The radar controller (I was the planning controller) used lateral separation; there is no relevant vertical separation criteria for such an event. (Makes sense.)

Next issue was terrain clearance. Although when the altimeter was last known to be working, terrain clearance was more than ample, and almost certainly was going to remain so, this could not be guaranteed until the aircraft was VMC.

The answer? Temperature.

Fairly frequent OAT readings were obtained from the aircraft concerned, referenced against the expected temp gradient for the day and area, and (most importantly) compared with readings from other aircraft in the vicinity, preferably at similar levels.

This is not foolproof. Unexpected inversions or temp gradients occur, particularly in frontal conditions. The controller was able to build quite the picture, MET-wise, of the temperatures expected along the route for the flight based on the reports he received, and frequently updated. Smart cookie.

Worked. The flight was able to be vectored (once again, with the assistance of reports of conditions from other flights in the area and along the projected route) to an area where the flight could be continued in VMC. The next ACC was informed of the situation and accepted the flight in due course. It went to where it was going, landed normally. IIRC he got a sincere round of thanks from the crew for thinking of this, and helping make it work.

A bit of met is good to have. Maintain a working (living) awareness of what the weather is doing.

[Piltdown Man]

I'll guess the answer - Initially, only lateral separation will be possible. Then you have determine if the aircraft clear of cloud and in sight of the surface? If yes, provide an ATC service as appropriate. If not, you have to determine if there is there any other altitude information available. The Mode C might work when the altimeters won't and don't forget that many aircraft have some height information from a GPS. The RadAlt, if fitted may be of some help if you can vector them over flattish terrain. However, the thing to bear in mind is that the reason for lack of altitude information may be blocked static ports which means that all systems that rely on that information will be next to useless as well.

PM

[Denti]

Quite a few other reasons for wrong or malfunctioning altitude information besides static ports. We had quite a few cases of malfunctioning AoA vanes a couple years ago and in all cases not only IAS but also altitude was missing or completely wrong (as well as G/S and W/V). And that on such a simple aircraft as the 737NG.

[BOAC]

Your standby did not work??????

[Denti]

Of course it did, no AoA vane input into the ISFD. And of course the ISFD has its own set of static ports as well. Interesting though that all incidents started either as ALT DISAGREE or IAS DISAGREE, not AOA DISAGREE. Anyway, my point was that there are some other sources for an altimeter malfunction than just simply blocked static ports.

[BOAC]

and in all cases not only IAS but also altitude was missing or completely wrong- it seemed to prompt the question:confused:

[orgASMic]

Schiphol airliner crash blamed on altimeter failure, pilot error - Wikinews, the free news source (http://en.wikinews.org/wiki/Schiphol_airliner_crash_blamed_on_altimeter_failure,_pilot_e rror)

This accident was blamed on altimeter failure leading to a CFIT because the autopilot believed the faulty altimeter and the crew noticed too late to intervene. The aircraft had a history of altimeter failure that had not been picked up.

My take on this emergency (it is a Mayday IMHO) would be to separate the aircraft laterally and get a shepherd aircraft (probably a friendly military crew, as they are used to flying in formation and will not have company issues to worry about) on him PDQ, whatever the flight conditions, then use the shepherd to escort him down the nearest ILS/PAR.

[BOAC]

org - best not to confuse radio altimeters (Schiphol) with what this discussion is about? Your plan is good, though:ok: