2006 Mid-Air Collision Brazil - Part 4

Cockpit Voice Recorder

I am going to post on this page the actual cockpit voice recorder recordings.  You can access the YouTube web site to listen to them as well. Click here to listen to N600XL and click here to listen to GOL Transportes Aereos Flight 1907.


I would like to suggest for you to listen to N600XL recording first to gain a sense of how the operation was being conducted.  It contains approximately two hours and 5 minutes  of the flight.  The sound of the actual collision can be heard at 1 hour and 23 minutes and 45 seconds into the recording.  


The crew of N600XL had not talked to Brasilia Air Traffic Control for almost an hour.  Many attempts to regain contact were made however. There was also no attempt to contact other aircraft for assistance on the air to air frequency of 123.45.  


Please note the pilots of the Legacy only use the word "emergency" in all 10 of their transmissions after the collision had happened.  They did not use the International Civil Aviation Organization (ICAO) phraseology standard word "MAYDAY" in any of them.


The fact the none of the authorities involved mentions or acknowledges the deliberate application of randomness in aircraft tracks is shocking to me. I personally have been applying randomness to my aircraft track for more than 50 years. (Once again, see my post on January 26, 2010 "Randomness is good!")


 The application of Strategic Lateral Offset Procedure by either or both crew would have prevented a collision.  The application of randomness in  some form is clearly an option of the Pilots-In-Command that is rarely talked about or applied but may be the last resort to assuring separation in the event of multiple errors or procedural failures.


Do you apply randomness to your aircraft's track? Your  comments please!

2006 Mid-Air Collision Brazil - Part 3

Radio Management Unit (RMU)

Transponder Status Indication

Primary Flight Display (PFD) TCAS FAIL Indication

The single item that could most likely be determined as a primary cause of this accident was the fact the transponder was not operating at the time of the collision.  As a result, the Traffic Alert and Collision Avoidance Systems  (TCAS II) with which both aircraft were equipped was not able to alert either crew of a possible collision.

There are several conspicuous indications of the TCAS/Transponder status in the aircraft instrument panel – eight visible indications in all, with two in the RMUs, two in the PFDs and another two in the MFDs (when the MFD was set to display TCAS), and the blinking amber transponder reply light in the “ATC window” boxes on both RMUs.

The Brazilian Crash investigation was unable to discover why it was not turned on even though considerable resources were expended.  A special ergonomics committee was established to see if there was the possibility of a leg or foot movement that could accidentally turn off the transponder.

One positive result of that study discovered that not enough consideration had been spent to make the on/off status of the transponder clearly displayed to the crew under all lighting conditions.


The authors of the Brazilian Air Force Accident Report, a 268 page .pdf document failed to even mention two very important points.  The first one they missed was the failure of either crew to apply the principle of randomness in their aircraft's track.

The concept of randomness is recognized in the form of the Strategic Lateral Offset Procedure (SLOP).  SLOP is in use today and much of the world's airspace, especially the non-radar coverage oceanic airspace as well as in Africa and in China while in radar contact with ATC's approval.

See this Blog's post entitled "Randomness is good!" posted in January of 2010.

They also failed to mention the value of a third cockpit crew member.  Without a doubt, the ideal size of an aircraft's crew is three.  (I recognize I may be beating a "dead horse" when bringing up the subject of minimum cockpit crew size of three.  Today's trend of automating and remotely controlling aircraft is a move to decrease the size of the crew to one and even to completely eliminating the crew.)

If a third crew member had been on board each aircraft, there is increased opportunity for enhanced safety in operations. The richness of the intellectual results of three pilots working together far exceed the results of just two pilots working together.

Since the advent of two man crews in the early seventies, this consideration has been dropped and not ever mentioned in National Transportation Safety Board (NTSB) accident investigation reports. The subject of possible three man crews is a taboo subject.

2006 Mid-Air Collision Brazil - Part 2

Boeing 737-8EH

Photo Credit: Boeing Image

1,514 words

I ended Part 1 of this post with the question: Why did this collision happen?

The Brazilian Air Force Accident report quoted the contributing factors as:

5.2.1.1.2. N600XL

Relatively to the crew of the N600XL, the following active failures were identified: lack of an adequate planning of the flight, and insufficient knowledge of the flight plan prepared by the Embraer operator; non-execution of a briefing prior to departure; unintentional change of the transponder setting, failure in prioritizing attention; failure in perceiving that the transponder was not transmitting; delay in recognizing the problem of communication with the air traffic control unit; and non-compliance with the procedures prescribed for communications failure.

a) Training – a contributor

(Participation of the received training process, due to a qualitative or quantitative deficiency, for not providing the trainee with full knowledge and other technical skills required for the performance of the activity).

b) Air traffic control – a contributor

(Participation of the air traffic service provider, on account of inadequate service provision).

The authorization to maintain flight level FL370 was given to the crew of the N600XL, as the result of a clearance transmitted in an incorrect manner. The vertical navigation conducted by the crew ended up being different from the one prescribed in the flight plan that was filed and activated, on account of the instruction incorrectly transmitted that led the N600XL crew to maintain flight level FL370.

c) Cockpit coordination – a contributor

(Error resulting from an inadequate utilization of the human resources for the operation of the aircraft, on account of an ineffective distribution and management of the tasks affecting each crew member, failure or confusion in the interpersonal communication or relationship, inobservance of operational rules).

The attention of both pilots of the N600XL focused on solving the question relative to the performance of the aircraft for the operation in Manaus, as they had learned of a NOTAM limiting the length of the runway of that airport. This hindered the routine of monitoring the evolution of the flight, because both pilots got busy with the same subject, creating the environment in which the interruption of the Transponder transmission was not perceived.

d) Judgment – a contributor

(Error committed by the pilot, resulting of an inadequate assessment of certain aspects of the operation, despite his being qualified for that operation)

The pilots judged that they would be able to conduct the flight even with their little adjustment as a crew and with their little knowledge of the aircraft systems, mainly the fuel system and the calculations of the weight and balance. They believed they could hasten the departure, resulting that they had just a short time to verify the flight plan and other documents, such as the NOTAM informing about the reduction of the runway length available at Manaus airport.

The PIC left the cockpit and stayed away 16 minutes, not considering the consequences of overburdening the SIC.

e) Planning – a contributor

(Pilot error, resulting of inadequate preparation for the flight, or part of the flight)

The planning of the flight was inadequate. Before the departure, there was not a monitoring of the elaboration of the flight plan that was being prepared by the Embraer employee, not allowing the pilots to have a previous knowledge of the proposed route and flight levels, although, in accordance with the Excelaire Manual of Operations, the PIC had to open and close the flight plan at the nearest FAA FSS or ATC office.

f) Oversight – a contributor

(participation of third parties, not belonging to the crew, on account of lack of adequate supervision of the planning or execution of the operation, at administrative, technical or operational levels)

The oversight conducted by the operator for the flight proposed was inadequate. The composition of the crew, with two pilots that had never flown together before, to receive, in a foreign country, an airplane in which they had little experience, with air traffic rules different from those with which they were used to operate, favored the lack of a good adjustment between the pilots, along with the already mentioned difficulties of cockpit coordination.

g) Little flight experience in the airplane – Undetermined

(Pilot error, resulting from little experience in the aviation activity, in the aircraft, or, specifically, in the circumstances of the operation)

The CVR indicated that, shortly before the moment of silence and the moment at which the Transponder discontinued the transmission, the PIC was looking at the fuel page of the MFD, and solved a doubt about fuel management with the SIC. It is possible that the PIC may have continued to look at other pages of the MFD and, possibly, to pages of the RMU.

The little experience of the PIC in this aircraft possibly made him look for information about the fuel consumed on the RMU fuel page, and, when leaving from this page and pushing the pertinent buttons, he unintentionally changed the setting of the Transponder from TA/RA to STANDBY, thus interrupting the altitude information of the mode C;

The insufficient adaptation of the crew with this type of aircraft and with the DISPLAYS of the respective avionics may have contributed to the unintentional selection of the STANDBY mode and to the subsequent lack of perception of the Transponder/TCAS status.

See Part 3 in my next post.

2006 Mid-Air Collision Brazil - Part 1

Embraer Legacy 600 

Photographed by Adrian Pingstone in March 2009 and released to the public domain.

(364 words)

     Since I brought up the subject of mid-air collisions in my last post, I thought it might be appropriate to look at this particular example of a recent one.  I have been following the developments of the investigation of this mid-air collision that took place in the airspace of Brazil on September 29, 2006, almost four years ago.

The fact the Internet has made a lot more information easily available is a taken for granted today.  If I attempted to gather information concerning this specific accident which happened in Brazil without the Internet, I would have most likely given up.

In particular, I used three internet sources for the majority of my information.  Wikipedia was my most used site however the richest source was the 280 page Accident Investigation Report by the Brazilian Air Force.

The third source was the most dramatic source since it contained the actual cockpit voice recorder recordings of both aircraft. This source was an article titled "The Devil at 37,000 Feet" that appeared on the Vanity Fair Magazine's web site written by William Langewiesche dated January 2009.I have been around the flying world for along time and this is the first time I have ever listened to the actual recording of an aircraft accident and it turned out to be personally an emotional experience.

These recordings have also been placed on the YouTube web site.

This mid-air collision had seven survivors.  All on board the Embraer Legacy 600, N600XL survived. The number of fatalities was 154.  All 148 passengers and the six crew members on board Gol Transportes Aereos Flight 1907 were killed. No one on the ground was injured or killed.

Both of these aircraft had the latest and greatest collision warning and collision avoidance equipment installed.  Both of these aircraft were being controlled by a modern air traffic control system using radar for separation purposes.  And both were crewed by professional pilots with extensive flight experience.  Neither aircraft had strayed off the course cleared to fly by Air Traffic Control. Yet, a mid-air collision actually happened and 154 people are dead.

Why did this collision happen?
(See Part 2 in my next post which lists contributory issues.)

Mid-Air Collisions Are Too Rare to Worry About or It is a Big Sky!

Alfred E. Neumann

625 words

How often do mid-air collisions occur?  I went to the  Nall Report on the AOPA's web site to see the most recent General Aviation statistics available for 2007. The Nall Report on General Aviation accident statistics only covers fixed-wing general aviation aircraft weighing 12,500 pounds or less. It says there were only 10 in all of 2007!

Surprisingly these 10 mid-air collisions involving 20 different aircraft with 21 pilots and passengers resulted in only 4 deaths. Two of the mid-air collisions were by four aircraft involved in formation flight.

The fact mid-air collisions may be survivable is rarely written or talked about. In 2007, there were a total of 17 survivors from six different mid-airs involving 12 aircraft!

My flight career took of in January of 1961 when I first soloed a Piper PA-12 Super Cruiser as a member of the Kent State University Flying Club.  I began to fly for Trans World Airlines in 1964. First as a co-pilot on the Lockheed Constellation and finally as a Boeing 747 captain.  Along the way  I accumulated a lot of flight time over the next 50 years.

In that 50 year period of time, I never experienced a close call from a mid-air collision. Why?

The continental United States is comprised of 3 million square miles.  If the airspace to a level of 10 miles is available to be used by all aircraft the available amount of airspace is 30 million cubic miles.  This is a heck of a lot of airspace or another way of saying it is to say: "It is a big sky!"

Now how many aircraft are using the big sky at the same time?  Today, it is possible to obtain the answer from a web site that counts  aircraft in the air called FlightAware.com.  I just went to the web site and right now (Sunday, May 22, 2010 at 7:30 PM there are 3,672 airborne aircraft including 205 which are operating under Visual Flight Rules (VFR).

The highest number of aircraft that are in the air at the same time counted by FlightAware is 5,650.

Assuming the aircraft were distributed in only half the available airspace or 15 million cubic miles it would mean each aircraft has approximately 3,000 cubic miles or a space 20 miles long, 20 miles wide and 7 miles high.

This is the primary reason for the lack of mid-air collisions. The Big Sky!

What role does Air Traffic Control play in mid-air collision prevention?

Another little talked about and written about is the possibility that it might play a negative role meaning ATC procedures may actually contribute to the mid-air collision issue.  


ATC will normally assign aircraft to fly along an established airway such as Victor 210 separating us by altitude,  lateral and longitudinal separations and watched over by radar. If either an air traffic controller or a pilot makes a mistake in the altitude flown or assigned, a mid-air collision is much more likely to happen due to the traffic being compacted by the present system. 

In the real world, it means they are actually packing us closer together than if random flight paths were being flown by each aircraft. Normal vertical separation between Instrument Flight Rules (IFR) aircraft is just 1,000 feet. The normal vertical separation between IFR traffic and Visual Flight Rules (VFR) traffic is just 500 feet.  

As a result of being aware of these facts, I will admit to not ever being overly concerned about mid-air collisions while I am flying my airplane.
What level of mid-air collision concern to you experience while flying your plane?


By the way, there is a matter of certainty about a comet hitting the earth at some point in the future and wiping out all life forms.
Do you worry much about it happening?


"What me worry?"

Polish Air Force Tu- 154M crash in Russia latest information!

I discovered the following article by By Simon Hradecky, created Wednesday, May 19th 2010. The article appears to have the most factual information of any I have read so far. 

The Russian Interstate Aviation Committee (MAK) in cooperation with the Polish Accident Investigators published first preliminary results of their investigation stating, that there is no evidence in support of any inflight breakup, inflight fire or any mechanical malfunction prior to first impact with an obstacle 1100 meters before the runway threshold (see the sketch of the impact marks). 

The engines were working until final impact. The Terrain Awareness Warning System (TAWS), the Global Navigation Satellite System (GNSS) and the Flight Management System (FMS, UNS-1D) were working until final impact.

Airfield and navigation facilities were found suitable to receive the aircraft as well as the Polish Yak-40 that arrived 90 minutes prior to the Tupolev.

The cockpit voice recorder revealed, that persons not belonging to the flight crew were on the flight deck about 20 minutes prior to impact. Polish Authorities are working to identify the voices.

The captain of the flight had a total flying experience of 3480 hours, thereof 530 hours on the Tupolev TU-154. The first officer had 1900 hours of flying experience with 160 hours on the Tupolev, the navigator 1070 hours total with 30 on the Tupolev and the flight engineer had 290 hours total flying experience with 235 hours on the Tupolev.

The crew did receive weather forecasts for Smolensk North during the preflight briefing, but did not have the actual weather information. The crew did not have current aeronautical data for the airport nor did they have the current NOTAMs.

The crew interaction was unusual due to the introduction of the navigator. No procedures were available for that type of crew combination and interaction, as the airplane is usually flown with a 3-man cockpit (captain, first officer, engineer) and all training, documentation and manuals were set up only for a crew of three.

The flight departed Warsaw with a delay of one hour (actual departure at 07:27L CET [06:27Z] while departure was planned for 06:30L [05:30Z] and arrival for 09:45L Moscow time [06:45Z]). The crew interacted with air traffic control in Minsk and Moscow in English and communicated with the controller at Smolensk North Airport in Russian.


Minsk Control told the crew during the descent, when the airplane was at about 7500 meters (FL250), that Smolensk reported a visibility of 400 meters due to fog. The crew of the Polish Yak-40, which carried journalists and had already landed at Smolensk Airport 90 minutes prior to the accident, told the crew of the Tupolev 16 minutes prior to the crash, that a Russian Ilyushin 76 had gone around due to a ceiling of 50 meters and visibility of 400 meters. 


11 minutes prior to impact the Yak-40 crew radioed the Tupolev reporting the Russian Ilyushin had gone around a second time and diverted. 4 minutes prior to impact the Yak-40 crew radioed the Tupolev crew again reporting they estimated the visibility to 200 meters only. Smolensk Tower told the Tupolev crew, that visibility was 400 meters due to fog. The crew decided to carry out a "trial" approach to Minimum Descent Altitude (MDA) and then to decide whether to conduct a second approach. The final approach was flown on autopilot and autothrust. 

18 seconds prior to impact with the first obstacle the TAWS alerted "PULL UP! PULL UP!" after it had already warned "Terrain ahead!" prior to this. 5 seconds prior to impact with first obstacles the decision was made to go around and the autopilot was disconnected. 


The airplane first impacted a barrier 40 meters left of the extended centerline of the runway and 1100 meters before the threshold of the runway at an altitude, that was 15 meters BELOW the runway elevation. The third impact with an obstacle, a birch tree of 30-40cm trunk diameter contacted by the left hand wing, 840 meters before the runway threshold and about 80 meters left of the extended runway centerline, led to the first break up of the aircraft in flight, the aircraft rolled inverted and impacted ground 5 seconds later, that final impact occured at 10:41:06L (07:41:06Z). 


Impact forces were estimated in excess of 100G and were not surviveable.

Emergency services arrived at the crash site 13 minutes after the impact and cordoned the area off in a distance of 500 meters around the crash site, 180 personnel and 16 vehicles were on scene. Small fires at the crash site were extinguished 18 minutes after impact.

The airport of Smolensk North had been inspected on March 16th by Russian Experts to determine whether the airport was suitable for receiving Tupolev 134s and Tupolev 154s. The experts found runway 259, lighting, navaids and procedures suitable.

On March 25th test flights were performed which confirmed the suitability of the aerodrome.
On April 5th the aerodrome was again inspected in preparation for the special flights, the MDA was set to the 100 meters AGL (328 feet) with a visibility requirement of 1000 meters.

Specialists were on duty at the airport of Smolenks on April 10th starting at 07:00L to assist the incoming special flights. Those specialists inspected the runway lighting at 08:00L and found it operational with no defects. The lighting was subsequently operated at high intensity, specialists therefore were not able to inspect the lights immediately following the accident, but conducted the examination on April 11th. 


Pilot reports of other aircraft arriving on April 10th and 11th were acquired by the accident commission.
The investigation continues.