PDA

View Full Version : Final report of near-miss (Bern) June 2011


Squawk7777
7th Nov 2013, 20:18
link (http://www.avherald.com/h?article=43e37d03/0000&opt=0)

Incident: Germanwings A319 near Bern on Jun 10th 2011, near collision in flight
By Simon Hradecky, created Monday, Nov 4th 2013 16:13Z, last updated Monday, Nov 4th 2013 16:13Z

The Swiss SUST (former BFU) have released their final report reporting the minimum separation between the two aircraft reduced to 50 feet vertically and 0.6nm horizontally with a high risk of collision. The SUST concluded the probable causes of the serious incident were:

The serious incident is attributable to the fact that air traffic control gave clearance to an aircraft which led to a hazardous convergence with another aircraft. The fact that one of the flight crews then followed the resolution advisories of the traffic alert and collision avoidance systems (TCAS) only initially and instead followed the instructions of air traffic control meant that the convergence involved had a high risk of collision.

The following factors have been identified as causative for the serious incident:

- The air traffic controller entered the flight level prescribed for this flight as a clearance into the air traffic control system, but gave a different clearance to descend by radio.

- None of the five air traffic controllers who were involved in managing the aircraft concerned noticed the discrepancy between the descent clearance in the system and the radio message.

- The crew of the aircraft followed the instructions of air traffic control instead of continuing to follow the diverging resolution advisory of the TCAS.

The following factors contributed to the genesis of the serious incident:

- An Air Force exercise led to an increased workload and a more difficult overview for civil air traffic control.

- The use of a non-standard phraseology by air traffic control.

The following factors contributed systemically to the occurrence of the serious incident:

- Air traffic control had no safety net at their disposal which would have been able to detect the working error of an air traffic controller at an early stage.

- Working processes in air traffic control which made it difficult to detect routine working errors.

The SUST reported that air traffic control cleared 4U-2529 down to FL250 with a descent rate of 2000 fpm or more on radio however entered FL280 into the radar console. The aircraft subsequently was handed off to the adjacent sector and reported "... descending FL250 ..." the controller did not detect that the crew message did not match the cleared FL280 seen at the radar console, FL280 being the generally agreed hand off level for approaches to Stuttgart.

In the meantime the Raytheon had been cleared to climb to FL230, then FL250 and further to FL270. While climbing through FL240, in contact with L12 sector, the crew detected a descending aircraft heading towards them in opposite direction slightly to their left. The crew wanted to query with ATC but was unable due to high radio traffic. A short time later their TCAS issued a traffic advisory, the pilot disengaged the autopilot.

The Germanwings Airbus, while talking to sector M2 controller, was just being given a direct to waypoint ARSUT when the TCAS issued a traffic advisory, the crew acknowledged the direct instruction with "stand by" as result. At that moment the controller recognized the Airbus had already descended below FL280 shown as assigned flight level on her radar and in a reflex action queried "Germanwings two five two niner, please, ah, confirm maintain level two seven zero?", FL270 chosen as the nearest level the descent could be stopped at. The crew confirmed levelling off and maintaining FL270. The first officer, pilot flying, disengaged the autopilot and levelled the aircraft manually. The crew attempted to acquire visual contact with the conflicting aircraft but was unable to do so despite clear weather.

The L12 radar controller recognized that the Germanwings Airbus was closing towards FL280 still with 2000 feet per minute and concluded the aircraft would descend below FL280. He immediately called L12 controller to stop the descent of the Germanwings stating their Raytheon was cleared to FL270. In response M2 controller advised they would turn the Airbus to the left, the Raytheon should also turn to the left.

The short conflict alert (STCA) activated at the control center.

5 seconds after the STCA the TCAS of the Airbus issued a resolution advisory "maintain vertical speed, crossing, maintain", which required to maintain a rate of descent of 1500 to 2000 fpm. The crew complied with that resolution advisory.

The M2 controller issued an instruction to turn left 30 degrees immediately with no reply, then repeated "Germanwings two five two niner turn left immediately". The crew confirmed they had initiated a left turn and were following a TCAS resolution advisory.

28 seconds after the TCAS RA the A320 TCAS issued "Increase Descent! Increase Descent!", the crew complied with the instruction.

In the meantime the Raytheon TCAS, evident by the Mode-S transmissions to the radar station but invisible to air traffic controllers, issued a "Climb! Crossing, Climb!" resolution advisory to the Raytheon crew (both pilots holding CPLs, captain with 8,120 hours total and 120 hours on type, first officer also rated a captain with 11,279 hours total and 415 hours on type) requiring a rate of climb of 1500 to 2000 fpm, a slight increase to the present rate of climb.

The L12 controller, not knowing which frequency the conflicting Airbus was on, in the view of the STCA decided to instruct the Raytheon "Rooster two zero one, descend immediately to flight level two six zero, descend immediately to flight level two six zero!"

Instead of continuing to follow the TCAS resolution advisory the crew read back the instruction and initiated a descent to FL260 reaching the highest point at 26650 feet standard pressure and now descended instead of maintaining the 1500-2000 fpm climb as demanded by TCAS, the resolution advisory had been active for 14 seconds already and remained active for another 18 seconds.

10 seconds after the instruction the L12 controller attempted to defuse the conflict by turning the Raytheon to the right. Only after this instruction the controller recognized that both aircraft had probably TCAS resolution advisories active and instructed "Two
zero one, follow TCAS, opposite traffic one mile, follow TCAS", the crew reported back "yeah we follow TCAS, traffic in sight".

At that point the aircraft were about to reach the point of closest proximity, they were 100 feet vertically apart. Both TCAS systems reverted their resolution advisories, the Raytheon being at 25625 feet standard pressure now received "monitor vertical speed" with any rate of climb marked red and any rate of descent marked green, while the Airbus at 25725 feet standard pressure was instructed to "Climb! Climb Now! Climb! Climb Now!" The Airbus crew (captain ATPL, 9,535 hours total, 8,035 hours on type, first officer MPL, 278 hours total, 173 hours on type) complied with the revised advisory, and the distance between the aircraft began to increase again.

42 seconds after the first resolution advisory was issued both TCAS system announced "clear of conflict".

According to the TCAS readout of the A320 the closest separation had been at 130 feet vertically and 0.6nm horizontally, the read out of the short term conflict alert system was 234 feet vertically and 0.52nm horizontally, while the radar data indicated a closest separation of 50 feet and 0.6nm.

SUST analysed that four control sectors were affected by the occurrence, the occurrence itsself however occurred in sector M2. Sectors M2 and L12 painted both aircraft while the other two sectors showed one aircraft only. At the time of the occurrence Geneva control center was not yet technically able to compare the clearance as entered into the radar control desk to the actual altitude entered at master control panel in the flight deck although that information was being transmitted via the Mode-S data link. Had this feature already been available at the time of the occurrence, the conflict would have been indicated three minutes earlier than the STCA actually activated. The SUST stated: "This alarm would very probably have enabled the air traffic controllers to intervene at an early stage and resolve the impending conflict at an early stage. Thus, an important safety net in air traffic control was missing; this contributed to the occurrence of the hazardous convergence."

The SUST analysed that in sector L12 many descents are to be coordinated, most of which have been agreed to a standard exit at FL250, while the FL280 exit to Stuttgart represents an "exception".

The SUST stated: "However, this does correspond to the flight level generally agreed between Geneva and Zurich for flights with Stuttgart as their destination airport. One possible explanation stems from the question of whether temporarily effective conditioning had occurred to the air traffic controller due to a repetitive process. This would mean specifically that the repeated utterance or hearing of the expression “FL250” led to an automatic response. In many cases such a response results in an efficiency and productivity advantage as cognitive resources are available for other tasks and the load of the working memory is held at a low level. Generally it can be said in this regard that the simultaneous presence of automated routine processes and conscious reflection are part of the day-to-day activity of air traffic controllers. In the interests of economy and flexibility, in the course of professional experience there arises a style of working tailored to these requirements; it exhibits individual nuances within the given framework. With regard to the above-mentioned error, this possible explanation leads to the actual result that the air traffic controller committed an error caused by the fact that in his conscious awareness of the moment the specific event was out of focus. According to our hypothesis the subsequent consequence was the unintentional uttering of FL 250 used several times beforehand in a kind of repetitive pattern."

The SUST analysed with respect to human factors: "Finally, even a simple ergonomic reason for this error in the work of the air traffic controller can be given: essentially, people's ability to carry out several activities at the same time is limited. When looked at in detail, it is impossible to deal with more than one two operation simultaneously in a focused manner. Multitask capability is characterized by individual differences depending on personality and is generally known to be of limited extent. Since it is necessary to resort to procedures which have been trained for and which have become routine when several activities are being carried out simultaneously, in such phases the risk of routine errors also increases. Whenever possible, therefore, activities should be carried out at slightly staggered times. In the management of modern commercial aircraft, ergonomic principles which aim to reduce errors have long been consistently integrated into the design of work processes. Thus, for example, the crew of such an aircraft are instructed and trained, after receiving clearance from air traffic control, to enter this instruction in the flight management system and only then to actively read back the numerical value which is then displayed by the system in order to confirm the clearance from air traffic control. This implements a method of working which ensures that an instruction is not merely read back routinely from memory; on the contrary, a 'closed loop' is created which helps to eliminate the majority of incorrect inputs before they occur. However, the method of working implemented by air traffic control - entering a cleared flight level into the radar display using a mouse and transmitting it to the crew by radio at the same time (“click as you speak”) neglects such simple ergonomic principles and offers only limited possibilities of identifying at an early stage a simple working error. These processes therefore represent a systemic contributing factor to this serious incident."

The SUST analysed that the Airbus crew was taken in surprise when the controller queried "please ah confirm maintain level two seven zero?". Although worded as a question the crew immediately understood it as an instruction, read back as an instruction, the first officer disengaged the autopilot and manually flew the aircraft to comply with both present instructions to maintain FL270 and turn left, then followed the TCAS resolution advisories. The SUST stated: "Switching off the autopilot to bring the aircraft to level flight as quickly as possible indicates that the crew were coping with the situation and were able to take appropriate measures." and summarized: "In summary it can be stated that the crew of flight GWI 2529 were appropriately skilled and trained and were able to adopt and implement the patterns of behaviour required by TCAS immediately. They followed the TCAS resolution advisories without delay and to the required extent."

With respect to the Raytheon crew, both of which had been trained and rated as aircraft commanders and both having received ACAS/TCAS training and having been made aware of the events of Ueberlingen (editorial note: one crew following TCAS, the other not following TCAS resulting in a mid air collision and the subsequent murder of the air traffic controller, who was left alone at work with a good number of systems down, by the relative of a victim), the crew "remarkably" initiated a descent despite an active resolution advisory instructing to climb.

The SUST stated: "Nevertheless, the serious incident clearly indicates that they were not able to access this knowledge and convert it into appropriate actions. The reason for this inability cannot lie in a moment of surprise or inadequate mental preparation, because these simple response patterns are designed precisely to allow them to be invoked reliably and in good time following the surprise. Rather, everything indicates that the crew of HHN 201 was essentially not sufficiently familiar with dealing with TCAS."

babotika
8th Nov 2013, 22:50
Not following TCAS?? :ugh: :mad:

Bin their licences.

Big Pistons Forever
8th Nov 2013, 23:03
I am not aware of any incident in North American airspace where a North American crew followed an ATC instruction that was contrary to TCAS RA direction. Yet it has happened at least twice in EU airspace, once fatally and here only luck avoided another tragedy.

Why is that ? Is it Aircrew training, ATC training/procedures or something else ?

BARKINGMAD
8th Nov 2013, 23:18
Sounds like the holes in the Swiss cheese are aligning nicely again?

Just thinking back to the DHL tragedy.........................

Cows getting bigger
9th Nov 2013, 07:01
Yes, I though we had sorted this TCAS nonsense after Uberlingen. :(

ATC Watcher
10th Nov 2013, 09:55
I am not aware of any incident in North American airspace where a North American crew followed an ATC instruction that was contrary to TCAS RA direction. Yet it has happened at least twice in EU airspace

All recent Mode S downlink studies both in Europe (over France - UK-Germany) and USA ( over East Coast /Boston) show the same worrying results ; 35% of RA not corectly followed, of wich 15% are ignored and 7% still flown in opposite direction.

Why more incidents reported in Europe ?: probably because of better Just culture and openess. In USA there are less reports , but does not mean less problems.

For info in another " near Ueberlingen " event reported over France near Evreux in Oct 2004 , it was the US crew of a Delta 767 that did follow the " expedite clearance " from the controller in opposition to the TCAS RA received , and both aircraft ended up also a few hunded feet from one another.

This incident was reported by French ATC ( although the initial error was theirs) but not by the Delta crew.
Interestingly today you find this incident in the Eurocontrol TCAS database, but not in the FAA one. That might explains why the US has less TCAS related incidents.

Squawk7777
10th Nov 2013, 21:07
Why more incidents reported in Europe ?: probably because of better Just culture and openess. In USA there are less reports , but does not mean less problems.

For info in another " near Ueberlingen " event reported over France near Evreux in Oct 2004 , it was the US crew of a Delta 767 that did follow the " expedite clearance " from the controller in opposition to the TCAS RA received , and both aircraft ended up also a few hunded feet from one another.

This incident was reported by French ATC ( although the initial error was theirs) but not by the Delta crew.
Interestingly today you find this incident in the Eurocontrol TCAS database, but not in the FAA one. That might explains why the US has less TCAS related incidents.

Hello Watcher! Nice to see a familiar poster after a longer absence from pprune.

In the US it is common to file a NASA report and an ASAP (if you are employed with an airline aka 121 carrier) after an incident. ASAP stands for Aviation Safety Action Program, and is based on self-reporting and is usually supervised/managed by a panel of three: One member from the FAA, one from the company, and one from the pilot's union. It is somewhat similar to the UK's ASR program(me) if my memory serves me right.

The NASA report is actually called Aviation Safety Reporting System (http://asrs.arc.nasa.gov/), also a self-reporting non-punitive program that also encompasses ATC, maintenance, general aviation etc. incidents.

While most ASAPs are not made public, ASRS has a database search available. Right now, I can't remember if incidents in foreign airspace/countries are included.

Locked door
11th Nov 2013, 13:39
Big Pistons,

It's funny, my company has nearly been kebabbed by two aircraft from US majors in seperate incidents. Both times the US aircraft failed to follow the RA, and in one case the parent company behaved in a less than honerable fashion by 'losing' FDR data and not cooperating with the investigation.

It happens all over the world, and it needs to stop. I agree with a previous poster, pull the licence of any crew that fails to respond to an RA. But that would require the cooperation of the parent authority, which might be hard to get.

Squawk7777
11th Nov 2013, 15:11
Link (http://www.flightglobal.com/news/articles/inquiry-details-confusion-behind-777-airprox-with-c-17s-375717/)


Inquiry details confusion behind 777 airprox with C-17s

US investigators have detailed the confused co-ordination between air traffic controllers that led to a close encounter between an American Airlines Boeing 777 and a pair of US Air Force Boeing C-17s off the New York coast.

The 777 pilots received three resolution advisories from their collision-avoidance system, two ordering a descent and the third demanding a climb, after air traffic control efforts to separate the three aircraft failed.

After a refuelling operation with a McDonnell Douglas KC-10, the westbound C-17s were cleared to descend to 21,000ft (6,400m). Shortly beforehand, the controller had also taken a radar hand-off of the 777, but could not contact its crew because the pilots had not been told to change radio frequency from 128.3MHz.

Before the hand-off the 777, heading southeast, had been cleared to climb to 23,000ft. Unable to reach the 777 himself, the controller sought to halt the climb by asking the 777's previous controller - now busy with a separate flightplan verification - to stop the jet at 20,000ft.

But this request, relayed via an assistant, went unheeded for 46s, and the controller instead ordered the C-17s to turn right in an attempt to increase horizontal separation. He also told the C-17s to stop descending at 22,000ft.

The 777's previous controller, still in radio contact with the jet, overheard the discussion about the C-17s' altitude and mistakenly identified 22,000ft as the height at which to stop the 777's climb.

As a result he ordered the 777 to maintain 22,000ft and also instructed it to turn right as a traffic-avoidance measure.

The C-17s' collision-avoidance systems were operating in "traffic-advisory only" mode, as required by US Air Force multi-ship formation rules, and did not deviate from their altitude, while the 777 crew received three resolution advisories in succession, the third reversing a previously ordered descent.

One of the pilots of the lead C-17 told the controller that it had come "within approximately 2,000ft" of the 777, as the two crossed at the same altitude. The 777 crew transmitted: "That guy passed us now, and that was not good." Radar data indicates the horizontal distance was only 2,300ft and there was no vertical separation. The second C-17 was 500ft to the right and 4,000ft behind the lead.

The controller handling the 777's communications told the crew: "I apologise, I am not working that other airplane." He told the National Transportation Safety Board that the plan to separate the aircraft "was not clearly conveyed to him" and that he was "just guessing what the [other] controller was doing with the [C-17s]", without understanding the "urgency" of the situation.

In a factual disclosure, issued in August, the NTSB attributes the 21 January 2011 encounter, which occurred 88nm (163km) southeast of New York, to a "misunderstanding" between the controllers.

ATC Watcher
11th Nov 2013, 15:41
Hi A7777, still there just getting older :E
Thanks for the explanations.

Yes the 7% are still there and will mathematically cause another "Ueberlingen" one day, although with version 7.02 , reversals RAs might help save the day.

Locked door : the parent company behaved in a less than honerable fashion by 'losing' FDR data and not cooperating with the investigation.

Interestingly in the incident I mentioned earlier , something similar occurred I was told.