Air France/ Egypt Air air prox incident

[Chronus]

On 01/01/2017 serious incident reported by BEA. Separation +/- 0.74 nm /300 `

Source : https://www.bea.aero/en/investigatio...-aaiu-belgium/

[SeenItAll]

I guess TCAS only works if you follow its instructions. The question will be, why did MSX keep climbing when its TCAS told it LVL.

[Mlambin]

Sorry if it's seems stupid but can you translate me this please:

" +/- 0.74 nm /300' "

More or less. 74nautical miles / 300feets ? I'm not sur to understand

[Herod]

A little more than one kilometre, and a little less than 100 metres. Too plurry close.

[ATC Watcher]

The rate of climb of the MSX in the last 1000 ft is most probably why RAs were triggered I would say . So far normal, but why the MSX continued to climb above its cleared level, against the initial RA to level off, and were they subsequent corrective RAs issued and to who, will the interesting bit, hence the investigation . Could be pilot over-reaction (most likely) or a TCAS bug.. will not be the first time . Curious of the result.

[ATCO1962]

Just to clarify, Mlambin. They are referring to the closest the aircraft got with reference to lateral separation (.74nm) and with reference to vertical separation (300ft). This is a common way of expressing how close aircraft have come in such a situation and the .74nm/300' is not a conversion equation.

[Mlambin]

Thanks for the answer. So first is horizontal and second vertical :-)

[Yaw String]

As professionals,when being aware of a stepped climb,due proximity of other traffic,we look for that other traffic..don't we..and make a conscious decision as to whether it is necessary to reduce our climb rate for the last few thousand feet,in order to avoid uneccesary TAs...don't we......
As one who has been on the receiving end of a crew who acted opposite to their RA,nearly taking me out,..I am,perhaps more attentive...

[Chronus]

Originally Posted by Mlambin

Thanks for the answer. So first is horizontal and second vertical :-)

1370 meters horizontally and 91 meters vertically. Much too close for comfort. Once upon a time it was said that if the fly grew wings, it was too late even to say Hail Mary.
What we don`t as yet know is the crew actions.

[ATC Watcher]

Chronus :

Much too close for comfort.

Yes in ATC separation standards values , but from the air, over one Km away on crossing tracks is not really that " too close" .
As to the vertical , TCAS is designed to give you 3 or 400 ft , not more . It is not a separation tool it is an anti collision system. So here they got what TCAS planned.
One of the Belgian newspapers printed a map of the event ( probably from FR24) showing as I expected a very high rate of climb ( 2500-3000ft/min) of the A300 which initiated the RAs.

[misd-agin]

V/s selected, with situational awareness, would prevent this. Energy management and good neighbor policy.

Had a corporate jet descend from FL410 to FL340 towards us. We were at FL330. TCAS "descend". Notified ATC. ATC apologized - "we didn't expect him to descend 7,000' in one minute. Jet 123 please call this number..." Violation of RVSM standards.

[pax2908]

Assuming AF did not act contrary to their RA, does this mean Egyptair overshot by at least 700 ft so it is not only a too high rate of climb problem?

[Capn Bloggs]

Originally Posted by pax

does this mean Egyptair overshot by at least 700 ft so it is not only a too high rate of climb problem?

Even if they got a "Level Off" RA (if they had TCAS 7.1) they would have been hand-flying the RA manoeuvre; they could well have gone through FL210 (which is of course ignored during an RA) while complying with the RA command.

[ATC Watcher]

Capn Bloggs : I think you have it most probably .
The high rate is always a problem. When receiving a level off RA , you disconnect A/P and if you do this close to the selected altitude your chances or overshooting a few 100 feet is very high .
This is one of the cases where the incident is in fact caused by having TCAS . No TCAS and the A300 would have most probably leveled off at FL210 as cleared by ATC and (hopefully) selected in the FD by the crew . The RA changed that a bit . But TCAS assume the intruder will not level off , In fact it is there to cover you for the " most probably" and the "hopefully" which is a good thing.

We just have to live with those once in a while.

[misd-agin]

The expected response is greater for TCAS then it is for the autopilot logic which is why we flew it manually. So any overshoot should be less. Considering that they were going to level off anyway an overshoot is poor piloting and it definitely doesn't take 700' to level off an airliner climbing at 2,000-3,000 FPM.

[KayPam]

Originally Posted by ATC Watcher

Capn Bloggs : I think you have it most probably .
The high rate is always a problem. When receiving a level off RA , you disconnect A/P and if you do this close to the selected altitude your chances or overshooting a few 100 feet is very high .
This is one of the cases where the incident is in fact caused by having TCAS . No TCAS and the A300 would have most probably leveled off at FL210 as cleared by ATC and (hopefully) selected in the FD by the crew . The RA changed that a bit . But TCAS assume the intruder will not level off , In fact it is there to cover you for the " most probably" and the "hopefully" which is a good thing.

We just have to live with those once in a while.

Doesn't that depend entirely on the crew's "personality" when flying ? For instance, they could "overreact" and push hard on their stick, get a +0.25G in the cabin, and level off almost instantly, or they could push very gently, more gently than the AP, and overshoot.

But it should really depend on how they manual-fly, shouldn't it ?

[ATC Watcher]

KayPam:

But it should really depend on how they manual-fly, shouldn't it ?

Yes, I am not familiar at all with the A300-600 trim system, but if you ( or the system ) trim for 3000ft/min and you suddenly depress the A/P , I wonder how the pressure on the yoke will be. Add delay due to some surprise and possibly a bit confusion over the R/A ( level off is a new 7.1 feature) then most probably receiving another corrective RA ,( under 7.1 this is what you get ) and you could get close to the scenario. Not an excuse though, , just a tentative explanation.
misd-agin :

it definitely doesn't take 700' to level off an airliner climbing at 2,000-3,000 FPM.

Agree, but where did you get the 700 ft from ?

[PJ2]

Does anyone's air carrier/employer use their FDM Program to monitor TCAS responses?

At one time, we had programmed by our FDM service provider a series of FDM events which were used to monitor TCAS responses.

It looked at the warnings, monitored the response and compared the resulting parameter data, (i.e., pilot response), with TCAS response requirements.

Normal crew response may be:

• Appropriate; following TCAS
• Timely
• Within TCAS design expectations (g-load, altitude gain/loss, VSI rate)

Inappropriate crew response may be:

• Late
• Slow
• Non-compliant
• Opposite; contrary to TCAS
• Aggressive (beyond design requirements for TCAS response

There were approximately 25 FDM Events. The project resulted from two presentations on incorrect TCAS responses, published by Lufthansa and by Air France, around mid-2003, (Air France - http://www.developpement-durable.gou...lerte_TCAS.pdf )

[Chronus]

Originally Posted by ATC Watcher

KayPam:
Yes, I am not familiar at all with the A300-600 trim system, but if you ( or the system ) trim for 3000ft/min and you suddenly depress the A/P , I wonder how the pressure on the yoke will be. Add delay due to some surprise and possibly a bit confusion over the R/A ( level off is a new 7.1 feature) then most probably receiving another corrective RA ,( under 7.1 this is what you get ) and you could get close to the scenario. Not an excuse though, , just a tentative explanation.
misd-agin :

Agree, but where did you get the 700 ft from ?

I think he got it by deducting the reported 300`separation from the RSVM of 1000`.



I am rather curious about the following from the report:
"Approaching FL210 MSX541 reports following a TCAS/RA and is observed climbing through the cleared level."

Is this the MSX crew saying they were complying with their TCAS RA.

ATC WATCHER says:

"It is not a separation tool it is an anti collision system. So here they got what TCAS planned."

But we don`t really yet know how and what avoided a collision. After all MSX says they were following the RA. How come they changed their mind and what could have happened if they had not. All we know is that MSX climbed through their cleared level for reason that they claim they were following a RA on their TCAS.



Fortunate all were spared the cheese, the kind with holes in, on this occasion.

[Capn Bloggs]

All we know is that MSX climbed through their cleared level for reason that they claim they were following a RA on their TCAS.

Chronus, it would appear to me that you do not understand responses to TCAS RAs. If you get a TCAS RA, the "cleared level" goes out the window and you do what is required to follow the RA. If it takes you through the cleared level, then so be it.

While I sympathise with Misd-Agin's comment, his honing in on 700ft is being a bit harsh. There are plenty of variables that could cause the crew to "do well" or "do slow" when responding to an RA. To arbitrarily criticise the crew for taking so long is being a bit harsh at this point.