PDA

View Full Version : Emerg Descent... change heading?


fatboy slim
7th Jan 2007, 15:04
May I ask for thoughts regarding turning the HDG knob by default when initiating an emerg descent.

My previous airline taught not too, my new airline (and the manufacturer's QRH) has taught me to turn as part of the memory items. I fly under radar apart from the odd trip home oceanic back from TFS. Flying oceanic routes it is clear a 90 deg turn is the safest maneuver.

My feeling is that, under radar, a controller will see the descent happening very quickly (way before he gets our mayday) and sort things out below. As we spend lots of time on hdgs anyway is changing HDG the best thing to do? Will TCAS clear the other chaps from underneath me as I hurtle earthward?

This is simply a hypothetical question as my SOPs are very clear.

Many thanks.

Rainboe
7th Jan 2007, 15:37
I go for the turn option. Reasons:
*Most likely to have traffic tracking same track below you. Inertial navigation is frightening for having everybody on the same line over the ground. A rapid descent could give TCAS a real test- turning onto a random heading will ensure less conflict that remaining on original track, so get off the centreline.
*Nosing over straight ahead will cause significant loss of 'g'- whilst the cabin crew and people standing are trying to move. A big turn will reduce the loss of 'g' as you nose over into the descent.

wileydog3
7th Jan 2007, 15:55
BEST course of action is to coordinate with ATC *if* possible. Here in the states, no big deal about maintaining course or turning if you are out in the west but in the NE Corridor, it is filled with flying metal and either choice carries potentially significant risk. Of course, with TCAS available, you have an additional tool to assess risk and course of action.

Mr Moustache
7th Jan 2007, 16:29
On the Oceanic Route (like T16 to TFS and the like) maybe using the offset procedure might be beneficial in reducing collision potential.
In busy skies over Frankfurt I think you have a problem either way as there are a lot of people out there but turning increases the risk as initially ATC have no idea where you are going.
Direct routings help keep people from flying identical tracks, so go direct whenever possible and don't turn.

I think I just decided that I wouldn't turn.

Rainboe
7th Jan 2007, 19:03
You can have same direction or opposite direction traffic 1000' or 2000' below you. Going into an ED, you have a changing VS which is putting a lot on the TCAS system in a very short time as you start descending to start computing threat levels and suggested avoidance action. I would suggest a large plane would have trouble initiating the avoidance in the very short time available to it before you come blundering down through his level. You may well also be descending at some 7000fpm through the direct tracks of many other lower flying aircraft, and with INS systems, all will be on exactly the same line over the ground! Why not turn off that track and take most of the immediate problem away?

ICT_SLB
8th Jan 2007, 03:54
The latest version of Autopilot of the Global Express has an "Emergency Descent" mode triggered off the Cabin Altitude - it turns 90 degrees off the FMS track & does a pre-programmed descent to 9,000 ft (if memory serves). Beleive this is per ICAO norms.

cribble
8th Jan 2007, 07:33
Thread creep, but I wonder....
We do a lot of oceanic work on comany preferred routes (off the airways, taking best advantage of winds). On an SFO/AKL, for example, we may be on an airway for only an hour or so (where it suits the winds). At other times, we are crossing/paralleling(sp?) or nowhere near airways.
I recon that a preconceived decision to turn off 90 could increase risk. I brief accordingly.

wileydog3
8th Jan 2007, 12:03
The latest version of Autopilot of the Global Express has an "Emergency Descent" mode triggered off the Cabin Altitude - it turns 90 degrees off the FMS track & does a pre-programmed descent to 9,000 ft (if memory serves). Beleive this is per ICAO norms.

That will work just fine for US crews flying west over the Rockies... once.

haughtney1
8th Jan 2007, 13:01
My own experience (sim exercises only) makes me conclude that there is no right answer here.
I personally view this on a case by case basis.

If you are in congested airspace..ie 30West or US/EU RVSM area's then the 90 degree turn off track is a good option. If you are trans-pacific or somewhere over the middle of Africa etc then in all likelyhood there is no one else within a 100nm anyway...so its less of an issue:ok:
This comes down to your overall SA, which on a longer sector, can be reduced due to fatgue, boredom etc...

con-pilot
8th Jan 2007, 18:43
That will work just fine for US crews flying west over the Rockies... once.

All the systems with this capability that I am familiar with the data in the FMS knows where the aircraft is and will level the aircraft off at the appropriate altitude for the area.

Mad (Flt) Scientist
8th Jan 2007, 19:57
I believe the GX system has a higher target altitude than 9000ft, but the problem will always remain that there are areas of the world where the ground level is too high to permit unassisted breathing; if the automatic system arrests the descent at, say, 20,000ft, it might have stopped you hitting a mountain but no-one's going to regain consciousness anyway. There's no point in levelling off above the altitude for the crew to recover.

Max Angle
8th Jan 2007, 21:37
Unless you are very certain of where you are in relation to high ground or you can pull the chart out and check the MSA (impossible in the circumstance I would suggest) you are taking a big risk by turning. The MSA printed on most nav. logs is only good for 10nm either side of the route and an emergency descent from 350 to 100 is going to put you at least 30-40 miles off track if you turn through 90 degress. I reckon it's better to stay on the planned route unless there is a very good reason to do otherwise, NAT tracks would be a good example.

Contract Con
9th Jan 2007, 22:43
Gday,

I hadn't considered the use of an "offset" for the ED before, but now you mention it, a 5-10nm offset and LNAV your way into it will give a turn as you pitch into the descent and then stop you from blundering off into the great unknown.

Must give it a try next time I'm in the cave.

Cheers,

Con:ok:

JackOffallTrades
10th Jan 2007, 00:20
Like a lot of these questions, when it happens there is very little time to decide if to turn or not and in which direction. Helps if you've analised and made this decission before it could potentially happen. Like all good crm exercises it helps if you let the other(s) pilot know what you're planning on doing when everything gets a little foggy on the inside.

mutley320
10th Jan 2007, 00:41
Maybe someone from EURO control or NATS could give an opinion on this one.
Bar NAT airspace & Oceanic with its own special procedures, i think i'd tend towards descent straight ahead.Especially in the ever busier european airspace.(Thank God for TCAS though.)

Rainboe
10th Jan 2007, 08:26
MaxAngle
The MSA printed on most nav. logs is only good for 10nm either side of the route and an emergency descent from 350 to 100 is going to put you at least 30-40 miles off track if you turn through 90 degress.

I think this post is materially incorrect. All the Nav Logs I know give MSAs far higher than 10nm either side. An emergency descent turn to either parallel course or reverse course is not going to take you 30-40 miles off track unless you keep heading out there. I would hope if in a mountainous region a decision is taken early about blundering on into terrrain! Meanwhile, I hope those who descended rapidly through everybodyelses tracks got down OK! I'm not sure TCAS could handle it.

FlightDetent
10th Jan 2007, 13:35
The answer is rather simple.

Amendment to ICAO Supplementary Regional Procedures Doc 7030/4 - EUR changed somewhat about three years ago.

Now: (EUR, Part 1 Chapter 6)

6.1 Emergency Descent Procedures (supplement to PANS-ATM, 15.1.4)

6.1.1 Actions by the P-I-C

6.1.1.1 When an aircraft operated as a controlled flight experiences sudden decompression or a (similar) malfunction requiring an emergency descent, the aircraft shall, if able:

a) initiate a turn away from from the assigned track or route before commnecing the emergency descent
b) advise the appropriate ATC unit ASAP of the emergency descent
c) set transponder to Code 7700 and select the emregency mode on ADS/CPDLS system if applicable
d) turn on all exterior lights
e) watch for conflicting traffic both visually and by reference to ACAS (if equipped) and
f) coordinate its intention with the appropriate ATC unit.

----------------------------

That is why the new SOP with new aircraft ask you to turn-pull-turn-pull-pull, right? :)

Max Angle
10th Jan 2007, 16:51
All the Nav Logs I know give MSAs far higher than 10nm either side. Your ops. manual should have a note about how the nav. log is constructed. The two I have used both say that the MSA on the log is calculated by inspecting an area 10nm either side of the planned track and taking the highest grid safety altitude (from an AERAD or other nav. chart) that falls into that area for each route segment. Obviously one would hope that any high stuff around would be in the pilots mind anyway and they don't go too far off track but a much higher MSA than you have printed on the log could in theory be very close.

The ICAO document seems fairly plain but I still wonder in busy areas if you a creating as many problems as you solve by turning, there is traffic everywhere anyway and a turn is certainly not going to put you clear of traffic. As for TCAS, if you have the presence of mind (and I not sure I would) you should set it to TA only (as for an engine failure) so that it does not give you an RA and any unit in a conflicting aircraft will know not to try and co-ordinate with yours. Would be a pretty interesting RA for the other guy though, two emergency descents in formation, lovely!.

FlightDetent
11th Jan 2007, 10:08
As for TCAS, if you have the presence of mind (and I not sure I would) you should set it to TA only (as for an engine failure) so that it does not give you an RA and any unit in a conflicting aircraft will know not to try and co-ordinate with yours.
The co-ordination is in fact only "I say UP - you say DOWN" so a RA reversal would probably happen very quickly. I have the same idea but it is nowhere to be seen on paper. I wonder if keeping RA on is perhaps to prevent the double descent formation situation.
Imagine this: Aircraft starting ED @ FL300 v/s 4000 fpm increasing to 7000 fpm. (RA threshold is 35s to closest point of impact). Same track/direction traffic @ FL 280, one mile behind. (target RA miss distance is 400 ft vertical, lateral RA threshold is about 1,2 NM).

Case1) the ED A/C is TA only - no changes to its flight profile. The 2nd A/C TCAS has 3 options
a) climb RA: this in fact may do little to no service as far as separation goes, also it is movement against the intruder whose trajectory is parabolic(? - you get my point) as the A/C accelerates into the dive. May not be an option.
b) maintain v/s RA: Not an option, this is why the RA was issued in first place.
c) descent RA: creates a double descent situation, the lower aircraft may be pushed to such rate its forward speed will start to increase (closing in on the ED A/C), or, even worse, the lower aircraft may not be able to keep the rate great enough thus gliding towards the ED A/C. Maybe not an option.

Case2) The ED A/C is TA/RA.
Options for the FL280 A/C are pretty much the same. But the ED A/C may get REDUCE V/S RA, and the other DESCENT (other combinations also available). While this is far from final resolution to the conflict, the initial manoeuvre will INCREASE the separation unlike all three options in Case1 above.

In vertical plane it is quite hard to find critical situations and very complicated to solve them. What is for a long time obvious to any pilot, that you ultimately need to change course to solve the equations. In fact, any lateral input greatly reduces the time/space exposure to the 400 ft-1,2 NM target miss bubble.
In vertical plane, maneuvering capability is limited let's say by (sustained) +4000 / -7000 fpm. Forward speed however, is about 450 TAS i.e. 45500 fpm! To get 1/10th reduction (4500 fpm), change track by 25 deg (arccos(0,9)=25[deg]). If you change track by 45deg, you get 30% reduction in forward speed (cos(45[deg])=0,7), that is 13000 fpm!

From this point of view, turn away, start descent, keep TA/RA on sounds like a good advice. But, what have I missed? (apart from the obvious turn exactly into the trajecotry of below crossing traffic).

FD.
(the un-real)

Max Angle
12th Jan 2007, 22:00
A very interesting analysis. I should have said "you COULD set it to RA" rather than "should". I have looked in our expanded checklists and it's not part of the procedure to do it so I guess it would be best to leave RA set, as you say it may help to prevent the double descent scenario.
Lets hope we never have to find out for real how much traffic you would get in way of if you depressurised over the London TMA at FL350, one thing is for sure, turning off track is not going to solve your problem.

Rainboe
12th Jan 2007, 23:14
I don't understand this resistance to turning off track. You are going to have more traffic following your exact track directly aligned below you than crossing you if you turn off to avoid the traffic below you, then a short time later turn parallel to either head back or continue, or divert. You are asking for trouble blundering straight down through everybodies airways tracks, hence the importance of the ICAO requirement to turn off track before commencing descent if able.

So what is the decision? To 'ignore' the ICAO requirement 'because I know better'?

mutley320
13th Jan 2007, 15:46
Turning off track/airway is the best way to go, as all the procedures state but my resistance to it (in busy controlled airspace) without communication with ATC, is that nowadays you spend large amounts of time on Radar Headings or indeed direct routings that aren't "on airway."
So i'd love to hear an opinion from NATS or EURO control.

777AV8R
13th Jan 2007, 21:41
Descent procedures are spelled out in the Jepps regarding cabin pressurization failure on NOPAC routes. An offset left/right 25 NM to clear the airway, THEN descent to a lower altitude.

For those about to say 'one must descend right away', the answer is no. The procedure is takes into consideration passenger oxygen availability. In my case, 22 minutes on the 777 for the passengers. The lateral offset and descent will not use up the 22 minute availability.

Safe flying gang.

Tee Emm
14th Jan 2007, 20:21
My own experience (sim exercises only) makes me conclude that there is no right answer here.
I personally view this on a case by case basis

Wholeheartedly agree. Because of the nature of the beast (emergency descent and a fair amount of adrelenin) especially in IMC, it is all too easy to overbank and overcontrol while dropping the nose for rapid descent especially with flight spoilers deployed. A straight ahead descent does not require negative "G" at all - rather like a TCAS RA descend command. In terms of dark night IMC on instruments, straight ahead may be the lesser of two evils.

FlightDetent
15th Jan 2007, 07:42
Turning off track/airway is the best way to go, as all the procedures state but my resistance to it (in busy controlled airspace) without communication with ATC, is that nowadays you spend large amounts of time on Radar Headings or indeed direct routings that aren't "on airway."
So i'd love to hear an opinion from NATS or EURO control.
I think that is quite understandable. Why not give it a second tought.
1) The procedure I quote is ICAO "differencies to Europe". While I agree that is is a great shame that ICAO documents are not available as public domain, the sad status quo is that I cannot give you a link. Yet, the quote says as a controlled flight. In other parts of the document, the word Eurocontrol is quoted numerous times. I get the feeling, that whoever designed this procedure (if able!) and codified it, had a very good European ATC input.
2) The headings you mention, as I see it, serve to make the full use of available airspace by reducing separation to the approved standard of 5 NM. TCAS threshold is 1,1 NM /35 s for RA whichever comes sooner (now I've found it). 420 kt TAS is 7 NM/min.
Let 5 NM separation be the "paperwork" threshold.
Let 1,1 NM / 35s be the "scary-hairy RA" threshold.
Let 0 NM be the ultimate threshold.
If you are kept on a heading to achieve separation 5 NM to a conflicting (same level) traffic, in a 90° scenario it would take you 42 second to hit it. This may very well be the the start of your Emergency Descent, so the everage RoD is about 2000 fpm in the first minute. Target ACAS RA miss vertical distance (now I've found this as well) is 600 ft. By the time you reach the ultimate threshold - 42 seconds, you would have achieved more than twice (1200') the ACAS life guard separation.
Discussion regarding the other two threshold is, I'd say, superfluous (http://www.m-w.com/dictionary/superfluous). Needless to say, ATC will definitely NOT be happy about an ED. Yet, neither them nor us are payed to feel so, as opposed to finding and exploiting the safest course of action, retaining as much margin as operationally available.
There is a procedure. It may not be the optimal thing to do ad-hoc, but it will save our (a-word) and cover our (b-word). A procedure must be simple, short, unific, and uniqe. At the end of the day, it is the pilot's call.
Cheers,
FD
(the un-real)

FlightDetent
15th Jan 2007, 07:44
Wholeheartedly agree. Because of the nature of the beast (emergency descent and a fair amount of adrelenin) especially in IMC, it is all too easy to overbank and overcontrol while dropping the nose for rapid descent especially with flight spoilers deployed. A straight ahead descent does not require negative "G" at all - rather like a TCAS RA descend command. In terms of dark night IMC on instruments, straight ahead may be the lesser of two evils.
Some interesting points lurking from a deeper layer of thought, which aircraft are you referring to?

BOAC
15th Jan 2007, 08:26
Use of autopilot is encouraged and a spin of the heading pointer is a simple way to achieve a controlled banked entry.