Greek TCAS incident & query re Automatic Collision Avoidance System

[big fraidy cat]

Apparently, there was a near miss today over the skies of Greece. I've posted the article below, which comes from government radio. No mention of aircraft type, but Olympic has a considerable number of 737-400s.

"The 116 passengers on board an Olympic Airlines plane en route from Athens to Dusseldorf via Thessaloniki were alarmed when the pilot was forced to make an abrupt manoeuvre, because according to the instruments, the plane was on a collision course with two fighter jets. A stewardess was lightly injured on the leg, while two passengers suffered shock. The plane, which landed as normal at 09:30 in Thessaloniki, is still at "Macedonia" airport until the injured stewardess is substituted by one of her colleagues. As per the Defence Ministry, two fighter jets (photo-reconnaissance phantoms), flying somewhere over the area between the island of Skyros and Thessaloniki at 18,000 feet, crossed paths over air lane Blue 1, where the passenger plane was traveling on, and came closer than international safety limits, thus activating the automatic airborne collision avoidance system. However, there was no actual danger of a collision. Still, the Ministry is running an investigation into the exact causes of the incident."


I just have a couple of curious questions. When this automatic system takes over, do the pilots get an aural warning, or are they taken by surprise? Is all of this done without the assistance of ATC, using only the plane's radar? Lastly, how does this system interface with the autopilot?

[Hotel Mode]

Its an aural or rather various different aural warnings, ATC should not intervene when the ACAS is set off, its purely aircraft based and is real time unlike the controllers radar. It doesnt do anything with the autopilot its a manual manouvre. Most autopilots arent capable of abrupt movements

[Tacan400]

A Layman's description of How ACAS Works -
ACAS is a last ditch collision warning system (when all else fails, i.e. eyeballs, procedures, ATC). It does not take over from the pilot.
It identifies and monitors transponders on other aircraft and displays them in the cockpit for pilots to monitor. It depends on conflicting aircraft having operating transponders that are turned on.
It issues an aural Traffic Advisory (TA) when an intruder is 20-48 seconds away
Issues an aural Resolution Advisory (RA) when intruder is 15-35 seconds away). This tells the pilot to climb or descend in order to resolve a potential collision.

ACAS is not connected to the autopilot-the pilot must take the action to avoid a collision

ACAS can issue an advisory warning or RA, even though a collision may not have occurred, due to the proximity of the other aircraft. However, a TCAS warning is very serious and should be obeyed, even if ATC is involved and giving conflicting instructions.

ACAS is not infallible, unpredictable actions of intruder aircraft, such as military aircraft, can defeat the system.But it's absolutely essential for passenger operations.

ACAS is believed to be about 70% effective at avoiding collisions; 30% is still down to the skill of the pilots and all the other operational rules/procedures (such as level restrictions; ATC intervention etc) that are in place to prevent conflicts. But it's better than the eyeball which is only up to 20% effective, and less so, for high speed and IFR aircraft.

[Danny]

Just to add that a TCAS manouver should not be a violent one but a smooth pitch up or down at no more than 1.5g if in fact the RA even suggests anything other than not changing your vertical speed which could even be zero. A normal TCAS RA manouvre would probably not be felt or noticed by anyone on the a/c outside of the flight deck.

If a flight attendant was injured in the TCAS incident described in the article above, the crew probably had to use a much more violent manouver than was indicated by TCAS.

Try this link (http://en.wikipedia.org/wiki/Traffic_Collision_Avoidance_System) for a better description.

[Barry Cuda]

its purely aircraft based and is real time unlike the controllers radar.


And it isn't a radar system, despite the fact that some pilots believe that it is.... (Not suggesting you H-Mode, just using your quote).

[big fraidy cat]

Because it's picking up transponder signals? So, if a transponder is turned off, then that aircraft can only be picked up by ATC radar?

Actually, what I'm thinking is that military aircraft might not send a transponder signal, if they didn't want to be detected. Or perhaps a hijacked airliner, as well.

[Hotel Mode]

Yes its totally transponder based so depends on each aircraft being so fitted. In terms of hijacked aircraft ATC will still have a primary radar blip on their screen but its no good for TCAS

[Capt Pit Bull]

And it isn't a radar system

Well, I guess that depends how you define 'radar'. I would consider it to be one. At a most basic definition it uses radio waves to detect range and direction, even though the later isn't particularly accurate or vital.

OK, it doesn't have a rotating antenna, but neither do phased arrays.

OK, it relies on the cooperation of the target, in the form of a transponders, but so do most aspects of ATC radar use. That just makes it a secondary radar rather than a primary.

Going to have to disagree with you there Barry.

CPB

[quickturnaround]

Yesterday Olympic Airways (flt OA181) Followed a valid TCAS RA in order to avoid 2 Greek Airforce fighters above the island of Skyros. One Flight Attendant was lightly hurt and the pax were shocked. The B737 landed safely at Makedonia , the airport of Thessaloniki.
This has happend before and will continue to happen, until it is made clear to the Greek Airforce that not ALL airspace is theirs. It is about time that the Greek HCAA (YPA) sits around the table with the airforce in a genuine effort to eliminate this kind of airprox mishaps. The communication between Military and civil ATC should also improve.
This year Greece will benefit of heavy Charter traffic on top of the heavy scheduled traffic, the tourists and other passengers must not be put at any risks because the airforce likes to play with their planes.
This time it was again a near miss, can Greek tourism endure a hit?
Fly safely, QTA

[Kalium Chloride]

If this had been a newspaper article, it would be written off as a non-event, TCAS doing its job, media hype, business-as-usual, with assorted criticism of the reporter. Carry on, men. :hmm:

[Charles Darwin]

Followed a valid TCAS RA in order to avoid 2 Greek Airforce fighters above the island of Skyros. One Flight Attendant was lightly hurt...
Exceeded a valid TCAS RA, I would rather say. When following a RA (I´ve done it) the cabin crew, let alone the pax, will hardly notice.

[quickturnaround]

Dear KCl, It has been on the front page of allmost all Greek-newspapers, and on TV and it surely not treated as a ''non-event''.

QTA

[Piltdown Man]

Our military friends generally have their transponders on so their own people can see them! It also helps us see them as well. Only when they are going to do naughty things to other people do they turn them "off" (or park them in a sneaky way so that only their own side can see them). ACAS/TCAS is an SSR application. Big Fraidy Cat - I'd suggest that you try and ignore any press article which has the words: alarmed, violent, storm, frightened, collision, narrowly missed, emergency, stalled, tailspin, etc... In fact any any article with any emotive or sensational terms.

Unless the threat traffic very rapid vertical closure rate, the information and warnings you receive are generally as follows: Visual (shown on a sort of map display): Hollow blue diamond, other traffic, Solid diamond, nearby (proximate traffic), Solid Amber circle: Threat traffic and Solid red square for traffic which you being instructed to avoid. The aural warnings only start with "Traffic! - Traffic!" when the traffic is first shown as a treat. When the system starts to instruct you take avoiding action, the intruder is shown as a red square. There is a simple write up on http://www.aerowinx.de/html/tcas.html.

And finally, ACAS stands for Aircraft Collision Avoidance System and TCAS, which is the same thing, is Traffic Alert and Collision Avoidance System.

[big fraidy cat]

Piltdown - I actually do avoid those words and that is not what caught my attention to the incident. When they reported an injury to the stewardess, it made me curious to inquire if the avoidance maneuver was designed to be sudden or gradual. A gradual climb or descent shouldn't cause injuries.

It's also an interesting question for this neck of the woods, as there is a lot of military traffic over the eastern Aegean, especially if you include the Turkish incursions into Greek airspace. I don't want to start a political war here, but these are crowded skies, and commercial aircraft are caught in the middle.

It will be interesting, if we eventually do see a report here, as to what actually happened. Since there was an injury, perhaps we'll get to see that report after time.

[anotherthing]

Captain Pit Bull,

I think what BarryCuda was alluding to is the way ACAS/TCAS should be used, not the machanical or electrical theory behind the system.

It is purely an aid to assist pilots in the visual detection of intruder A/C via a TA (in the first instance) and in the second instance, it is an aid to help avoid collision via an RA.

Some pilots - and I would like to stress, it's not too many, believe that TCAS is an instrument that can also be used to give them a picture of what is going on around them. When being told about traffic by ATC, the phrase "we have it on TCAS" means nothing - it might be the wrong A/C!!


In the vertical sense, it can help the situational awareness, but in a horizontal sense ACAS/TCAS is very unreliable in azimuth. It's misuse has caused incidents in the past, when A/C that were on headings to miss each other by 5 or more miles have actually had an airprox because one of the sets of crew decided to turn the A/C on the strength of the information they saw on the display.

It is a good system, it is not infallable, it has caused incidents in TC airspace where several A/C are in the same vicinity.

[big fraidy cat]

QTA - An article in today's Kathimerini online edition indicates that, due to fog, the military aircraft could not be identified by the pilots of the Olympic flight. So, in defense of the Greek Air Force, we really don't know yet if the fighter jets were Greek or Turkish. The foreign ministry is keeping silent for the moment.

[Capt Pit Bull]

anotherthing,

I think what BarryCuda was alluding to is the way ACAS/TCAS should be used

fair enough.

Some pilots - and I would like to stress, it's not too many, believe that TCAS is an instrument that can also be used to give them a picture of what is going on around them.

<nods> Overuse of the TD is a major issue.

However, we should appreciate that the TD can form one of many inputs into a pilots situational awareness. It certaining gives a picture, trying to tell flight crew it doesn't is I feel counterproductive. The problem comes when people think it forms a complete picture and don't understand the limitations.

The TD should never be the sole reason for initiating a change of flight path, but that does not preclude it being the starting point of a course of action, e.g. a query to ATC.

When being told about traffic by ATC, the phrase "we have it on TCAS" means nothing - it might be the wrong A/C!!


I couldn't agree more. Pretty much the only sensible response is "Roger" - takes up less airtime!

It's misuse has caused incidents in the past

<nods> Absolutely. I sense a mutual preaching to the choir between the two of us ;)

[Tacan400]

And just a reminder (although clearly not the case in this instance, 'cos TCAS clearly worked and the military aircraft had serviceable transponders) you can't be too reliant on TCAS ....not all aircraft are transponder-equipped, not all transponders are functioning, not all transponders are accurate and not all pilots turn them on. Australian ATC sees many aircraft each day where pilots fail to turn the transponder on after take-off, then there are the VCA's by non-transponder equipped aircraft, then there's the unserviceable transponders enroute through poor or no maintenance, unreliability or electrical failures. And then there's all the bulldust Class G in Australia where you aren't required to have transponder at all. Hopefully, over 90% of all aircraft in proximity to passenger aircraft are transponder-equipped but they don't all function all of the time.

[GreekPilot747]

From various greek newspaper articles I gathered the following:
OA B737 in descent from 20000ft to 16000ft .
Formation of 2 RF4E on opposite course at 15500ft.
B737 at 400+ kts, the RF4E at 800+ kts groundspeed.
If the speed for the Phantoms is correct, that would have been a very high lateral closure rate.
B737 climbed 1000 ft within 2 seconds.
That would mean, they arrested the descent and climbed 1000ft within
2 seconds.
Now if you calculate that, you end up with a climbrate in excess of 30000 fpm.
How realistic is that, and what would be the g-forces implied?
Having been trained on TCAS RA manoeuvers in the sim for the last 5 years,
I never experienced anything as violent as that described by the press.
So let's wait for the investigation, to see what really happened, and how close it really was.
In any case, it is good to hear, that apart from the FA's slight injury nobody else was hurt.
I would like to wish her a quick and complete recovery.
In many parts of the world this would probably not have made it into the media, at least not to the extent these things get media coverage in Greece.

[Aristomenis]

...the tourists and other passengers must not be put at any risks because the airforce likes to play with their planes...

They play well all right, 24 hours a day, 365 days a year, keeping your (singular) country's airspace free so that you can fly your NG with safety...

[alexban]

Having a RA with an airliner it's a bit different than that with a military fighter jet. I've know of a 747 cpt having a RA with a Mig21,just before the jet launched for supersonic flight,RA that caused injuries to a nr of pax and FA.
You can get very fast a 'climb,climb ' command followed by a fast 'descent,descent' ,due to high speed and manouvrability of the fighter jet .
I guess the fighter pilots should be trained,and told about the risks involved in playing 'targets' with the airliners,cause there are a lot of pax and crew walkind,standing in those planes.
Luckily for the Olympic they were descending,with the fasten belts sign on,i guess...

[Desert Diner]

B737 climbed 1000 ft within 2 seconds.
That would mean, they arrested the descent and climbed 1000ft within
2 seconds.
Now if you calculate that, you end up with a climbrate in excess of 30000 fpm.
How realistic is that, and what would be the g-forces implied?

Not very! :bored:

Probably more of a quick change of direction with all the momentum related mishaps on board.

[anotherthing]

Captain Pit Bull,

I think I mis wrote what I meant to say!!

TCAS is a useful tool in helping with the situational awareness, if used correctly, as you said.

Alexban

UK forces are no longer allowed to 'embellish' civil A/C. I don't know about overseas

[Kit d'Rection KG]

Captain Pit Bull,

In the vertical sense, it can help the situational awareness, but in a horizontal sense ACAS/TCAS is very unreliable in azimuth. It's misuse has caused incidents in the past, when A/C that were on headings to miss each other by 5 or more miles have actually had an airprox because one of the sets of crew decided to turn the A/C on the strength of the information they saw on the display.

No, in my experience it is usually extremely reliable in azimuth. Other than in manoeuvreing flight, I have never seen a TCAS target which couldn't be confirmed in both range and azimuth using airborne radar, almost always with the two targets overlying each other. I'll admit that this is perhaps a variable feast depending upon the aircraft and equipment, but I can't agree with your statement Captain PB.

It is perhaps weakest used as as an SA tool in the vertical sense, because some manufacturers and operators set the display parameters very low, typically own altitude +/- 2600ft.

[Capt Pit Bull]

Kit,

I can't agree with your statement Captain PB

Huh? That quote you've made isn't from me.

Having said that, whilst I agree that it can be quite reasonable in azimuth, it isn't for all circumstances (this has been discussed several times here and on tech log over the years).

The main issue IMHO is not so much the accuracy of the traffic the display is showing you, but rather being aware that there may well be traffic the display simply isn't showing you at all: - non transponding aircraft, interference limiting mode, suppression of other traffic during RAs, ultimately even limits of graphical symbology.

Regardless, there have been some real howlers and near howlers caused by misuse of the Traffic Display. In particular collision courses established by flight crew making up their own avoidance manoeuvres based on their interpretation of the TD.

PB

[Kit d'Rection KG]

Sorry PB,

A 'quote' function plus human factors problem there!

I should have quoted anotherthing, who did make the post. No offence intended.

[Sensible Garage]

how sure are we it was an RA at all? my understanding of the system is that a relative old fighter jet like the RF-4E will not have an encoding altitude transponder (mode C), just a military mode 3 that implies just a symbol and a "trafic, trafic" in the 737....

[OXOGEKAS340]

They play well all right, 24 hours a day, 365 days a year, keeping your (singular) country's airspace free so that you can fly your NG with safety...

keeping it free from.....what?:uhoh:

[lhchris]

Hi everone!
Eurocontrol released a pretty good article about the misuse of TCAS and the consequences.
If you have a look at the section 'moving reference display', you will also see that optical 'illusions' can lead to incorrect interpretation of the displayed traffic.
The article finally tries to say that TCAS is not a radar replacement, but more an assisting system and guide!!
If you like, see the article here:
http://www.eurocontrol.int/msa/gallery/content/public/documents/Safety/ACAS_Bulletins-BUL6-D-1.1_21Apr2005.pdf

[Aristomenis]

keeping it free from.....what?:uhoh:

Well, I suppose you have found the BIG solution for a great aviation future.... Since you find no meaning in their existence, we should seize every country's air force...

[cloudcruiser]

Studying TCAS at this very moment for the ATPLs :eek: so this is pretty intereresting.

Just curious, does the RA specify a speed/rate of descent, or is there a standard?

[DraggieDriver]

Just curious, does the RA specify a speed/rate of descent, or is there a standard?

Cloudcruiser, depends on the system. Some have a red arc appear on your VSI to indicate the vertical speed you should aim to reach, some bring up red blocked out regions on your AI and you need to place your attitude outside the red area.

[OXOGEKAS340]

Well, I suppose you have found the BIG solution for a great aviation future.... Since you find no meaning in their existence, we should seize every country's air force...

Again, to protect from WHO?
It is a reason, to spend thousands of millions of
euros, every year! And as far as I know, at the same position, I filled 3 times AIRMISS, a few years ago.......trying to escape from the rushing towards us, F5's!:}

[rodthesod]

Cloudcruiser,

TCAS II gives CLIMB or DESCENT RAs which basically call for a change of pitch attitude within 5 seconds which will achieve 1500 fpm ROC/ROD when stabilised. The pitch change depends on your TAS and can be calculated by (pitch change = 1000/TAS). For most civil jet transports this gives a range of approx 2.5 - 7 degrees pitch change. An INCREASE CLIMB or INCREASE DESCENT RA calls for a further similar pitch change within 2.5 seconds to achieve 2500 fpm. INCREASE RAs usually occur only when one a/c does not respond within the above parameters or is non-TCAS.
Most TCAS problems are pilot induced by under-reacting or, more frequently, over-reacting in pitch and 'zooming' into someone elses airspace. Pilots should always be aware of their TAS so that they can get their initial response correct. TCAS will give corrections if you get it wrong, but not all systems have 'damped' IVSIs.
Incidentally, don't be confused by some previous posts. TCAS II IS NOT RELIABLE IN AZIMUTH. Having used it and trained pilots in its use for several years it usually appears to be so, but manoeouvre based on TCAS display alone is outside the design parameters of TCAS II and expressly discouraged by the manufacturers.
rts

[DraggieDriver]

TCAS II IS NOT RELIABLE IN AZIMUTH. Having used it and trained pilots in its use for several years it usually appears to be so, but manoeouvre based on TCAS display alone is outside the design parameters of TCAS II and expressly discouraged by the manufacturers.

Absolutely. I've seen TCAS paints with 40 degree bearing error, and once even saw a 4200 foot altitude error on a paint. The traffic was showing correct height to the air trafficers radar, but my TCAS was interpreting the altitude incorrectly. It was not conflicting traffic, and I noticed it mainly because the traffic was showing up as -2700 AMSL when it wasn't really underground. System ground checked serv.

[Clandestino]

I have never seen a TCAS target which couldn't be confirmed in both range and azimuth using airborne radar (...)
It is perhaps weakest used as as an SA tool in the vertical sense, because some manufacturers and operators set the display parameters very low, typically own altitude +/- 2600ft.

WOW! I have never seen TCAS target which azimuth&distance I could verify by using airborne radar but perhaps that's because WX radars are very lousy at detecting & tracking other airplanes and there are no other radars on ATR. What are you flying, Kit? My guess is either sentry or foxhound. As for situational awareness, with speeds and RoCs normaly used by transport category aeroplanes, 12 nm and +/- 2700 ft are more than adequate.

B737 climbed 1000 ft within 2 seconds.
That would mean, they arrested the descent and climbed 1000ft within
2 seconds.
Now if you calculate that, you end up with a climbrate in excess of 30000 fpm.
How realistic is that, and what would be the g-forces implied?


Absolutely unrealistic.

To climb 300m in 2 secs, one needs average vertical velocity of 150m/s. Let's assume acceleration and decceleration are constant. It's oversimplification but it will give even lesser normal acceleration than peak acc required in "realistic" scenario. So in the first second of climb one needs to accelerate fom zero to 300m/s and in the second the same amount of decceleration is required. And that's pull-up of 30G, followed by -30G nose-over. That kind of normal acceleration would pulverize 737 (and pax too).

Most fighters in use today carry transponders compatible with civllian mode C. A colleague of mine was once practice intercepted by pair of Armee de l'air Mirage 2000C's. First ATC asked them if they agreed to be intercepted. After they agreed, soon they got TCAS targets on ND and ATCo warned them that fighters will soon turn off their xpndrs to avoid triggering RAs. Mirages did that at about 6 Nm distance and turned them back on after disengaging.

[wrongthong]

A question about the future here, guys, sorry to interrupt. Does anyone know whether the promised ADS-B/TCAS hybrid systems will improve this azimuth inaccuracy issue? And accuracy of TCAS generally? Grateful for some info-I'm doing some research on the issue.

[rodthesod]

Wrongthong,

The azimuth 'issue' is only an issue when pilots try to do what they're expressly advised not to; i.e. manoeuvre in azimuth by interpreting the TCAS II display. TCAS II does not require to know where the other a/c is, only its altitude, distance from your aircraft and rate of change of that distance.
There will be no TCAS III.
My understanding is that TCAS IV will incorporate an air-to-air data link of the aircraft GPS position, so enabling much greater accuracy of azimuth presentation.
I'm a bit out of touch, having retired from flying 2 years ago, but I'm sure Collins or Honeywell would be happy to assist with your research.
rts

[k3k3]

The TCAS system is never going to be particularly accurate in azimuth due to the design of the antenna's. The TCAS antenna is in fact four in one, one antenna per quadrant, the TCAS computer sees which quadrant the signal is received in most strongly, then compares this with the strength of the signal received in an adjacent quadrant and "guesstimates" the bearing of the other aircraft.

[Evvy]

Can someone please clarify the following:

ACAS/TCAS systems fitted to commercial aircraft work on either modes A, C or S and only an "s to s" contact will trigger the full cockpit TCAS aural and visual alerts.

Military aircraft typically operate a IFF (friend or foe) system, so when switched on, this system has therefore to be mode S to issue alerts to commercial aircraft.

Are all military aircraft Mode S equipped then?

Evvy

[rodthesod]

Evvy

ACAS is a generic term and, so far, TCAS is the only system within the category.
TCAS needs the other aircraft to be transmitting at least mode C in order to issue an RA, although TAs should be available on receipt of mode A transmissions. In order to facilitate fully programmed RAs (where both aircraft respond; typically one CLIMB and other DESCEND in a head-on situation) both aircraft must have serviceable mode S transponders. If one has u/s mode S or merely mode C only the serviceable mode S equipped a/c will receive an RA. This may well rapidly escalate to an INCREASE CLIMB/DESCENT because the other a/c has 'failed' to respond.
Someone else will have to clarify with regard to military a/c - I believe they're mode C.

rts

[Evvy]

Rodthesod

My earlier post clearly states my understanding of what triggers the full cockpit TCAS aural and visual alerts.

To clarify therefore my question, the Greek aircraft took avoiding action, was this from an "s to s" contact (do all military operate Mode S?) or from a "c to s" (pilot initiated descent?)

Evvy

[Capt Pit Bull]

Sorry to split hairs, but TCAS does not see mode A transponders at all. It sees Mode S or Mode C, either with altitude data (in which case an RA can be generated) or without altitude data (in which case it's TA only).

You have to have mode C (or S) to send altitude data, but you don't have to have altitude data to reply to a mode C interrogation, you just send an empty pulse frame which is enough for tcas to do a speed of light / range calculation. TCAS doesn't interrogate mode A - the 4096 code is useless to it and just wastes bandwidth.

CPB