PDA

View Full Version : Nasty Incident


250 kts
4th Jul 2002, 08:40
I understand that there was an Airprox last night on A25.

Traffic was descended to a safe level head-on to other traffic. The descending traffic either bust the level or received a TCAS,but whatever,he continued the descent and appeared out of the garble 100 feet :eek: :eek: above the other traffic.

With the very sad accident earlier in the week and this very close one over here does this bring into question the reliability and accuracy of the TCAS system?

It just goes to show that it could be any one of us at any time.

But for the grace of God.

eyeinthesky
4th Jul 2002, 09:09
From what I heard, the descending traffic was descending at a high rate to 1000 ft above the opposite direction traffic. The TCAS of the descending aircraft considered that the conflict required descent (due to the high RoD??) so the pilot complied (in contradiction with the instructions of the controller) and passed through the level of the opposite direction less than one mile in front of it!

Given the tragic events in Germany and the concentration of speculation on the reliability or otherwise of TCAS this should perhaps serve as a wake-up call to all those who think technology is infallible.

Capt Pit Bull
4th Jul 2002, 11:08
One of the things I stress when giving TCAS training is the importance of not using high descent rates near a cleared level. Not only is it dubious airmanship, but it invites a scenario where TCAS, being predictive, extrapolates the flight path will pass underneath the aircraft beneath it.

Lets call the descender a/c A and the level a/c B.

One of TCAS' priorities, when selecting an RA, is to avoid crossing the other aircrafts altitude. Clearly, if the two aircraft are never coaltitude, they can not collide.

However, TCAS is also constrained in terms of how big a manoeuvre it can ask for, and how rapidly - for obvious reasons - it can't just ask for anything between +/- 3 G!

Scenario 1:

If a/c A is predicted to pass slightly below a/c B, then TCAS, knowing its manoeuvre constraints, will figure out that it is possible to get a/c A above a/c B, and since this is non altitude crossing, that is what it will do.

So, a/c A gets "Adjust Vertical Speed" and an RA reducing its descent rate to pass over a/c B.

Meanwhile a/c B gets either "Monitor vertical Speed", with red arcs saying not to climb, or perhaps "Descend - Descend".

Scenario 2:

If a/c A is predicted to pass well below a/c B, then TCAS may well see that i.a.w. its manouevre constraints, a/c A can not be kept above a/c B.

i.e. reducing As descent, and making B descend, will bring the a/c closer together by the time of Closest Point Of Approach.

So, (reluctantly), TCAS has to choose an Altitude Crossing solution.

A is told to maintain descent, B is told to climb.

The problem is, Altitude crossing RAs are strongly counter intuitive. This is because A sees an aircraft below it, yet is being told to descend. Yet B sees and aircraft above it, yet is told to climb. Historically, non compliance with crossing RAs has been worse than other types.

To help let the crew know what is going on, the word 'Crossing' is added to the verbal bit of the RA.

So, A gets "Maintain Vertical Speed, Crossing, Maintain"
B gets "Climb, Crossing, Climb"

As long as everyone follows their RAs, and accurately, the outcome is safe, but if either does not, or worse manouevres opposite, then things will get dicey. TCAS should be able to sense teh non compliances and issue reversals is necessary, but clearly its a dodgy path.

So, my message is this. If you are descending towards your cleared level, and you want an RA that takes you 1,300 or more feet past your cleared level, in a scary altitude crossing manoeuvre, then use big descent rates.

If you use airmanship to reduce your descent rate earlier, then you will get a non crossing RA that keeps you above the aother aircraft. However, remember TCAS doesn't know about your cleared level, and aims to change your flight path as little as possible. So, it will engineer a vertical miss of a few hundred feet. So it'll still be pretty scary.

Better still, reduce your descent rate some more. Then you'll just get a preventative RA telling you not to increase your descent rate. Now you can level as cleared, and not have an airprox. You still have to file an MOR though, because you';ve had an RA.

Don't like paperwork? Use a lower descent rate. Keep below 1500' per minute within 1,000' of your cleared level and your chance of having even just a TA, let alone an RA, diminish rapidly.

If you get a TA as you approach your cleared level, that should be an immediate prompt to check descent rate, and reduce if necessary, avoiding all sort of fright and paperwork!


What happened in this incident? Sounds like not quite an altitude crosser.

To my mind, the 'thousand foot level off' scenarios are a good example of when a pilot might make use of the given authority to disregard an RA!

Following the green arc in this situation generates an airprox of a few hundred feet. Obviously I'm not suggesting flying in the red band, but if flying in the black band allows you to meet an ATC clearance, then that is (a) sensible and (b) legal.

Basically, everyone is told what to do when ATC and TCAS disagree - namely to follow TCAS, stay out of the red (even if it means breaking a clearance), stick to the green, and don't 'outperform' TCAS by flying in the black to avoid causing another conflict by deviating excessively from original clearance.

But we should also consider what to do when TCAS and ATC agree!

If 'outperforming' TCAS meets your ATC clearance, then why not do it.

e.g. Head on encounter, low level. Left to its own devices, TCAS will engineer a 300' miss. Serious airprox!

But lets say, just before TCAS gave its RA, ATC tell one pilot "Avoiding action, climb now FLXXX - expedite - expedite" or some such. Just as the pilot pulls up for a 3000 fpm climb, TCAS goes "Climb" and asks for 1500 fpm.

Now TCAS and ATC are in agreement as to what sense (pitch up), but by following ATC a 600 foot or more miss will be achieved.

To my mind, this is what is meant when the crew are given authority to 'disregard' TCAS. Not to Ignore, or manoeuvre opposite RAs, but instead to amplify them if instructed to do so by ATC.

The weak link in the chain is crew training, IMHO.

CPB

ferris
4th Jul 2002, 11:20
This has nearly happened to me before.
A little while ago, high rate-of-descent 737 asked who the traffic was, as he had a TA. Passed traffic to both, explained stopping level above/ other one level below. 2nd a/c then reports RA to CLIMB, but not complying!!!! First aircraft agreed with my sep as well. Would have been ugly if one OR BOTH had complied. Never trusted TCAS since.

Loki
4th Jul 2002, 11:20
Capt Pit Bull:

Thanks for that, it addressed a lot of the questions we were asking ourselves....just what this board is for IMHO.

250 kts
4th Jul 2002, 12:03
Thanks for that pitbull-very informative. Now please start to put pressure on your airline and any others you may know to reintroduce the Fam Flight scheme for ATCOs. They really are an invaluable tool for these sort of things. Just a pity our employers don't see it that way.:mad: :mad:

unwise
4th Jul 2002, 12:27
It would appear that the descending aircraft was given a RA to continue descent and going on what the Cpt said above, the other aircraft may have been given a RA climb. In any case as suggested if both comply we survive, but in this instance the second aircraft did not action his RA.
TCAS in my opinion compromised the two aircraft and did not ensure either by pilot compliance or itself that the descending aircraft past lower than the AC in level flight at the point of crossing.
It was unfortunate in its timing, just after Germany but as has already been said whether it's your fault or not you may one day find that you too have two less returns on your display. It's not going to get any better.

sky9
4th Jul 2002, 15:51
I had a very similar incident a couple of years ago in an aircraft with the VSI /TCAS display that only gives a 5 mile range. We descended "FL200 level 10 before Monty". On the way down we picked up a very strong jetsteam on the tail and increased speed (and therefore VS so as to be level) What we didn’t know was that there was a turboprop coming the other way at FL190. Passing through FL230 (ish) the first indication was the TCAS going to "traffic" then "climb" By the time we had come out of the descent into a climb it reverted to "monitor VS" and then stopped.

It was only when we were well into the manoeuvre did we see the aircraft paint on the VSI. A good argument for having horizontally separated inbound and outbound tracks on A25 for EGCC and EGGP.

Capt Pit Bull
4th Jul 2002, 16:09
As an addenda to my earlier post, one of the things TCAS v7 does is try and spot '1,000 foot level off' scenarios.

If it spots one then it delays issuing the RA by 5 seconds. So you get 5 seconds longer TA phase, so thats a minimum of 15 seconds.

Thats enough, in any aircraft, to take a big bite out of the descent rate.

Be pre armed!

High Vertical Speed + "Traffic" + approaching cleared level = reduce Vertical Speed.



Incidentally, I'm glad you guys found the post useful - When I got to the end of it I thought "Ouch - thats a bit long and verging OT".

professor yaffle
4th Jul 2002, 16:31
not at all captain, as been said , very informative and useful to us all
must admit i'm not that keen on tcas in a known tfc environment, personally
out of interest (prob'ly a thicky question!) but why in the scenario outlined does't TCAS see the conflict a/c maintaining lvl flight and instruct the descending to lvl out above??

cheers

prof

beaver eager
4th Jul 2002, 17:51
CPB,

Could it not be argued that the weakest link is the inability of TCAS to know what your cleared level is? Presumably this would invariably be set on your altitude selection on the MCP or in the FMC if a VNAV operation. I remember being astounded during my TCAS course (Bob instructing, not you!) that this information was not fed to TCAS.

Wouldn't the addition of this information and inclusion of suitable programming remove almost all instances of spurious level-off warnings? (Apart from those associated with hand flying through the level set in your altitude select - surely a relatively rare occurance in the en-route scenario?) Not a terribly costly modification in the grand scheme of things, I would have thought.

Capt Pit Bull
4th Jul 2002, 18:45
Hi Beaver,

You make a valid point, trouble is its the relatively rare level bust which is one of the reasons to have TCAS in the first place. Basically, people not being at their cleared level is a major cause area of airprox, so we can't just assume that preselected altitude will be achieved.

As I mentioned, v7 does look for level off scenarios (as indeed did 6.04). E.G. if one aircraft is level, and the other has relative altitude of greater than 1,000 feet then a level off is on the cards. Or if neither aircraft is level, but there are 2 available levels between them.

There are also TCAS variants (Honeywell TCAS 2000 springs to mind) that know your cleared level, but don't inhibit based on it. I believe they use it to identify these scenarios, but don't quote me.

That 5 second delay really makes a difference in allowing people to start levelling off, and since its inclusion (combined with crew awareness of moderating VS near cleared level) radically reduces the occurance of these nuisance alerts.

By way of comparison, our RA rate, with our training course and v7 equipment, is only about 10% the national average from the previous v6 years.

1,000' level offs may be a pain in the proverbial, but thats all they are as long as everyone follows the RAs. They only turn nasty if someone doesn't!

CPB

atco-matic
4th Jul 2002, 23:02
it's made the news now:

http://news.bbc.co.uk/hi/english/uk/newsid_2095000/2095724.stm

MACC 29 all the time!!!!
5th Jul 2002, 07:05
Sky9 in response to your suggestion about horizontally separted in bound and outbound tracks to EGCC and EGGP, agood idea in concept and as a rule this is done in bounds on the west outbounds on the east. However as with everything in avaiation this is the ideal, and as we all know this does not happen very often. As soon as you throw a couple of overflights in there at various levels and speeds this particular piece of airspace gets very busy and complex very quickly. Maybe you need to visit us at the manchester center to see how it works? (something I think not enough flight crew take a chance of doing, those that have say it helps tremendously)

If your interested don't hesitate to contact me, happy to help

Nugget90
5th Jul 2002, 07:45
It is generally acknowledged that, whilst changes made to TCAS II/ACAS II equipment design requirements have been effective in reducing the frequency of occurrence of RAs generated by high vertical rates, no further changes can now be made without degrading the safety benefit unacceptably.

ICAO is developing guidelines for operators, suggesting that they should authorise pilots to use a modest vertical speed throughout a climb or descent when the vertical interval is not large - such as a change of altitude in a holding pattern, or a step climb of a couple of thousand feet - and to specify how this should be accomplished. Operators will be encouraged to specify procedures appropriate to the type flown that pilots would use to reduce the aircraft's vertical speed when an autopilot is engaged.

Further information on what pilots should be taught regarding compliance with RAs and the need never to manoeuvre in a sense opposite to that posted by an RA (so as not to compromise compatible RAs) can be found in Flight Deck Forums >> Tech Log >> Tcas.

Avalon
5th Jul 2002, 08:37
I'm with SKY 9 on this one.......... surely this is a case for a lot more Controlled Airspace to allow more horizontal separation, inbound and outbound routes etc. Why not make JUST about all airspace above FL100 CAS?

HEATHROW DIRECTOR
5th Jul 2002, 09:05
I greatly appreciate the info provided on here about TCAS by CPB and a few years back I was fortunate enough to be loaned the BA TCAS video by a pilot friend. However... I wonder how many ATCOs have A) experienced a/c taking TCAS "avoiding action" for no good reason and B) have had to get in fast to prevent TCAS action causing a real nasty, or have seen potentially dangerous situations which TCAS has ignored?

I've had a good number of "A" above and had several of "B" which seriously frightened me:

1. A/c descending in the hold suddenly announced "TCAS climbing" and got to within a few hundred feet of the guy above before stopping on ATC instructions. At all times prior to the incident fully legal separation applied.

2. Two TCAS equipped a/c in the hold, one immediately above the other. Top one busted his level and got to within 200 ft of the one below. Nobody mentioned TCAS and we sorted it because our "TCAS" warned us and we were able to take immediate action. If I recall, there was no lateral separation at all.

3. A/c descending to 4000 ft suddenly announces "TCAS descending" and shot down to around 3500 ft. Only traffic was someone climbing up straight underneath to 3000 ft. Again, it got too close for comfort.

I accept fully that pilots should obey TCAS but in examples 1 and 3 above ATC stopped TCAS induced action which could have caused serious amounts of paperwork. If the pilots concerned had ignored us and continued to follow TCAS....... well it doesn't bear thinking about.

ZIP250
5th Jul 2002, 12:21
Please correct me if I'm wrong but did I just hear BBC lunchtime news say that this incident happened last night and that it was all saved by the automatic collision avoidance system.

Maybe this is all reporter speculation but the story I heard from very close to source was that the incident was caused ty TCAS. Perhaps NATS or Prospect could tell the BBC the truth before the next propaganda bulletin.


:mad: :mad: :mad:
Z

Cryolosophorous
5th Jul 2002, 14:06
Thanks, Pit-bull. Don't stop now..youre on a roll. All very attentive in here...(Or we should be..)

spekesoftly
5th Jul 2002, 14:11
Capt Pit Bull,

Like to add my thanks for the very informative post on TCAS. I've read elsewhere that pilots are trained to give a GPWS alert priority over a TCAS RA. Would the TCAS equipment not be aware that the GPWS was simultaneously giving a warning, and factor that into the RA?

eyeinthesky
5th Jul 2002, 15:00
Without wishing, or being able, to pre-empt the official investigation into this incident which has left the controllers concerned needing conselling, I understand that not a small part of the problem might have been the way the crew reacted to the TCAS TA. I cannot really say any more other than the scenario of concentrating on one issue and neglecting the other (remember that from the CRM course and various crashes??) might have raised its head again.

120.4
5th Jul 2002, 19:56
Most valuable CPB, thank you.

One of the scenarios we often find ourselves in at LL is a full stack which then has the bottom 3 dragged off leaving much needed levels vacant. LAM is the best example. It has become my common practice to ask for a good rate of descent when I drop the next aircraft down to the lowest holding level in order to assist my TMA colleagues who are eagerly watching the CCTV, waiting for the levels to become vacant. Of course, all of the stacks have outbound traffic underneath them; in the case of LAM it is the Dover SIDS of SS.

This technique, designed to make the very best use of our oversubscribed airspace, runs counter to your very good advice on rates. May I ask, what is the maximum rate of descent I should ask for which won't expose us to the risks you mention?

250k is right. The FAM flight scheme promotes safety through mutual understanding and we need it back now. There are ways around the security issues. Please do all you can through your airline, it is in their interest.

Point 4

:)

professor yaffle
6th Jul 2002, 00:21
great idea to reinstall them, however at my unit, trying to get one is extemely difficult.

admittedly haven't asked for a fam flight recently but have asked for fam visit to college to get better understanding of college courses ( and what studes are actually taught, instead of being told by some , not all , have never been shown that table - the vortex wake table for deps - yeah right! -) in order to help me be a better ojti and been told have no money for such trips, therefore i can only assume that there is no money available for fam visits either - what with car parking fees and taxes etc!

prof

Ahh-40612
6th Jul 2002, 07:54
Gave courtesy traffic info to a MYT going through LMS to EGNX a few days back - response was "thanks for that as our TCAS is on the blink at the moment".
Aren't they meant to tell us?
Probably should have put in a 1261.

Scott Voigt
7th Jul 2002, 18:25
Just for info.... In the US, if an aircraft advises that they are reacting to an RA. We are then hands off until they are finished.

regards

ferris
7th Jul 2002, 18:31
But what if they don't tell you (because they are busy doing it), or it's not clear (as happened the other day)?
It now really concerns me that we are not in the loop.

Capt Pit Bull
8th Jul 2002, 13:27
Hello All,

Sorry not to reply earlier, have been off down route for 5 days.

Spekesoftly,

Several factors on your query.

The only one which is absolute is that there is a heirarchy of audio alerts, and GPWS is superior to TCAS. So if your audio system is trying to say "Pull Up - Pull Up" (GPWS) and "Descend - Descend" (TCAS), then you will hear "Pull-Up".

However, the TCAS RA will still be shown as a command on the vertical speed scale.

Other considerations.

Several GPWS modes, which incorpate various envelopes of various parameters, including RAD ALT.

TCAS is inhibited, to varying degrees, by low RAD ALT.
i.e. you won't get any RAs under any possible circumstance below 900' Radio. You don't get DESCEND below certain RAD ALT, etc etc

There certainly are possible areas where both systems could be active but I think you can see that the more of a 'white knuckle' GPWS scenario you have, the more likely TCAS is to be inhibited.

If you'd like a fuller list of system inhibitions, I can dig one up.


120.4

Quick answer:

Our TCAS manufacturers handbook recommends not more than 1500 fpm within 2,000' of level off.

Personally, I've done some sums and quote it the other way round, (assuming there is traffic within the horizontal trigger)

"If you exceed 1500 fpm within 1,000' of your cleared level, the probability of a nuisance RA increases very rapidly".

What Flight Levels are you talking about? I'll dig out the precise triggers and give you a better answer.

Also remember the issue here is combined rate of closure.


Must dash - Mrs Pit Bull and Miss Pit Bull Junior require me at Tesco's.


CPB