Midair collision between Helicopter and CRJ (AA5342) at Washington (DCA)

[Quickstatus]

Was the heli actually on route 1 at the time of collision?.

 

[Quickstatus]

Clearly in this case the traffic was given far too early, I would have given the Blackhawk a restriction to hold north, when the CRJ got closer it would have got to the point where they could have safely confirmed identification and passed behind. Assuming no NVGs in use - and the helicopter had a controller with capacity to monitor it.

That seems an uninformed casual observer to involve a lot of potential failure points. A system should be resilient enough a few failures does not bring the whole thing down

 

[justinbrett]

That seems an uninformed casual observer to involve a lot of potential failure points. A system should be resilient enough a few failures does not bring the whole thing down

This is tower control bread and butter.

 

[jb747]

This is a simulation, but I suspect it’s extremely close to reality.

 

[justinbrett]

This is a simulation, but I suspect it’s extremely close to reality.

That audio is way out of sync with the video.

 

[RooFlyer]

Was the heli actually on route 1 at the time of collision?.

If you look at this Juan Brown video from 7 mins (he's a large plane commercial pilot), he shows the helo path overlain on the route map (ATC as well). It cuts the inside margin.

 

[jb747]

That audio is way out of sync with the video.

Who listens to the radio? I rarely want to hear what is said on youtube.... Anyway, it's about what could be seen, not heard.

 

[justinbrett]

Who listens to the radio? I rarely want to hear what is said on youtube.... Anyway, it's about what could be seen, not heard.

The point is the picture looked a lot different when the second traffic call was made. You could clearly see the individual nav lights of the CRJ (about 6 seconds from the audio). And I assume the pilots were issued with necks.

They are VFR after all. If they can’t spot aircraft at night they should be flying IFR.

 

[Quickstatus]

This is tower control bread and butter.

I have major problems with "bread and butter " situations.
Bread and butter situations should never fail because they are so well practiced, second nature and familiar.
But fail they do and often especially when the built in safety margins are eroded through normalisation of deviation.

ATC was dealing with a certain level of congestion
- it directed to CRJ from RWY1 to 33 - primarily to utilise additional runways to accomodate all arrivals
- it directed the aircraft on RWY 33 to expedite takeoff
- it notified the heli of the CRJ traffic

In other words, ATC had situational awareness - it had radar, it had communication with all aircraft in its area of control.

But putting the CRJ into RWY33 approach while knowing that a heli is on Route 1 reduced or removed the 2 layers of safety - lateral and vertical separation. Those layers of safety are there to account for variation in human and aircraft performance - pilot, aircraft, and ATC, such that it requires a large deviation for a failure to occur. By reducing /removing the separation distance, it meant that only a small deviation vertically and/or laterally = failure. In this the Heli is a little too high or the CRJ a little too low or both, and the heli appears to cut inside the Route 1/4. Bread and butter should never reduce safety margins. (I know you dont say that either - this is a comment of what happened at DCA not your practices).

ATC then approves a visual separation at night with its inherent safety issues as discussed by others- but it does not tell the Heli that the CRJ is very close or that the lateral and vertical separation is fast disappearing. ATC does not advise of a Conflict alert, rather it asks the Heli to confirm visual separation. This can't be bread and butter for ATC with a greater level of situational awareness to assume that the Heli has the same.

ATC should not have put the 2 aircraft on a collision course.
Is it the individual ATC's fault?. I say no because it seems to me that this is not the first time ATC have been directing traffic like this - in other words, it's practices has been normalised AND allowed.

I think the problem could be:
1) Procedures that allow ATC to do what it did?
2) Vertical and horizontal separation margins, and whether they are actually adequate?
3) Congestion at the airfield putting pressure on ATC to reduce safety margins?
4) The design of Heli Routes, takeoffs and approaches which have to potential to confict with one another?
5) Practice of ATC communicating with aircraft using different frequencies when they are in the same airspace?
6) Task saturation - ATC handoff of visual separation duties to a heli reduces the potential of task saturation?
7) FAA for lack of oversight?

FAA and by delegation the ATC is supposed to prevent collisions. It failed in the "bread and butter"

It is very naive to say "we do it all the time and we never had a failure". The problem is that the failure is latent.- it just has not happened yet. When it does, invariably the practice right up until the point of failure is then seen and appreciated to be unsafe.
So the proper phrase is "we do it all the time and we never had a failure, until it failed....."

 

[justinbrett]

I have major problems with "bread and butter " situations.
Bread and butter situations should never fail because they are so well practiced, second nature and familiar.
But fail they do and often especially when the built in safety margins are eroded through normalisation of deviation.

ATC was dealing with a certain level of congestion
- it directed to CRJ from RWY1 to 33 - primarily to utilise additional runways to accomodate all arrivals
- it directed the aircraft on RWY 33 to expedite takeoff
- it notified the heli of the CRJ traffic

In other words, ATC had situational awareness - it had radar, it had communication with all aircraft in its area of control.

But putting the CRJ into RWY33 approach while knowing that a heli is on Route 1 reduced or removed the 2 layers of safety - lateral and vertical separation. Those layers of safety are there to account for variation in human and aircraft performance - pilot, aircraft, and ATC, such that it requires a large deviation for a failure to occur. By reducing /removing the separation distance, it meant that only a small deviation vertically and/or laterally = failure. In this the Heli is a little too high or the CRJ a little too low or both, and the heli appears to cut inside the Route 1/4. Bread and butter should never reduce safety margins. (I know you dont say that either - this is a comment of what happened at DCA not your practices).

ATC then approves a visual separation at night with its inherent safety issues as discussed by others- but it does not tell the Heli that the CRJ is very close or that the lateral and vertical separation is fast disappearing. ATC does not advise of a Conflict alert, rather it asks the Heli to confirm visual separation. This can't be bread and butter for ATC with a greater level of situational awareness to assume that the Heli has the same.

ATC should not have put the 2 aircraft on a collision course.
Is it the individual ATC's fault?. I say no because it seems to me that this is not the first time ATC have been directing traffic like this - in other words, it's practices has been normalised AND allowed.

I think the problem could be:
1) Procedures that allow ATC to do what it did?
2) Vertical and horizontal separation margins, and whether they are actually adequate?
3) Congestion at the airfield putting pressure on ATC to reduce safety margins?
4) The design of Heli Routes, takeoffs and approaches which have to potential to confict with one another?
5) Practice of ATC communicating with aircraft using different frequencies when they are in the same airspace?
6) Task saturation - ATC handoff of visual separation duties to a heli reduces the potential of task saturation?
7) FAA for lack of oversight?

FAA and by delegation the ATC is supposed to prevent collisions. It failed in the "bread and butter"

It is very naive to say "we do it all the time and we never had a failure". The problem is that the failure is latent.- it just has not happened yet. When it does, invariably the practice right up until the point of failure is then seen and appreciated to be unsafe.
So the proper phrase is "we do it all the time and we never had a failure, until it failed....."

The bread and butter skills I was referring to was my example about how this situation could have been handled safely. I’m not going to defend the actions of ATC in this incident, already been quite vocal about many of them.

 

[RSD]

I have major problems with "bread and butter " situations.
Bread and butter situations should never fail because they are so well practiced, second nature and familiar.
But fail they do and often especially when the built in safety margins are eroded through normalisation of deviation.

ATC was dealing with a certain level of congestion
- it directed to CRJ from RWY1 to 33 - primarily to utilise additional runways to accomodate all arrivals
- it directed the aircraft on RWY 33 to expedite takeoff
- it notified the heli of the CRJ traffic

In other words, ATC had situational awareness - it had radar, it had communication with all aircraft in its area of control.

But putting the CRJ into RWY33 approach while knowing that a heli is on Route 1 reduced or removed the 2 layers of safety - lateral and vertical separation. Those layers of safety are there to account for variation in human and aircraft performance - pilot, aircraft, and ATC, such that it requires a large deviation for a failure to occur. By reducing /removing the separation distance, it meant that only a small deviation vertically and/or laterally = failure. In this the Heli is a little too high or the CRJ a little too low or both, and the heli appears to cut inside the Route 1/4. Bread and butter should never reduce safety margins. (I know you dont say that either - this is a comment of what happened at DCA not your practices).

ATC then approves a visual separation at night with its inherent safety issues as discussed by others- but it does not tell the Heli that the CRJ is very close or that the lateral and vertical separation is fast disappearing. ATC does not advise of a Conflict alert, rather it asks the Heli to confirm visual separation. This can't be bread and butter for ATC with a greater level of situational awareness to assume that the Heli has the same.

ATC should not have put the 2 aircraft on a collision course.
Is it the individual ATC's fault?. I say no because it seems to me that this is not the first time ATC have been directing traffic like this - in other words, it's practices has been normalised AND allowed.

I think the problem could be:
1) Procedures that allow ATC to do what it did?
2) Vertical and horizontal separation margins, and whether they are actually adequate?
3) Congestion at the airfield putting pressure on ATC to reduce safety margins?
4) The design of Heli Routes, takeoffs and approaches which have to potential to confict with one another?
5) Practice of ATC communicating with aircraft using different frequencies when they are in the same airspace?
6) Task saturation - ATC handoff of visual separation duties to a heli reduces the potential of task saturation?
7) FAA for lack of oversight?

FAA and by delegation the ATC is supposed to prevent collisions. It failed in the "bread and butter"

It is very naive to say "we do it all the time and we never had a failure". The problem is that the failure is latent.- it just has not happened yet. When it does, invariably the practice right up until the point of failure is then seen and appreciated to be unsafe.
So the proper phrase is "we do it all the time and we never had a failure, until it failed....."

It would seem likely that there has been a normalisation of deviance - or else the procedures were just dumb and asking for trouble to begin with - or had never been reviewed and revised to account for changing circumstances over time...

And why not do these training flights in the early hours of the morning when there are probably less aircraft taking off and landing?

 

[Jacques Vert]

I can't help wondering when these procedures/heli routes were approved? Are they something that was approved in the 60's or 70's when traffic levels were, presumably, significantly less? Or were they something that was approved more recently?

 

[justinbrett]

It would seem likely that there has been a normalisation of deviance - or else the procedures were just dumb and asking for trouble to begin with - or had never been reviewed and revised to account for changing circumstances over time...

And why not do these training flights in the early hours of the morning when there are probably less aircraft taking off and landing?

Or perhaps just staff the dedicated helicopter control position that’s supposed to be there.

 

[RooFlyer]

And why not do these training flights in the early hours of the morning when there are probably less aircraft taking off and landing?

I guess 'training at night' would be part of the syllabus.

 

[RSD]

I guess 'training at night' would be part of the syllabus.

Middle of winter in DC - I reckon it would be even darker at 01:00 to 03:00.

 

[RSD]

I can't help wondering when these procedures/heli routes were approved? Are they something that was approved in the 60's or 70's when traffic levels were, presumably, significantly less? Or were they something that was approved more recently?

The obvious problem with the procedures is what @jb747 said a couple of pages back - "At night? It's almost impossible to say who you're looking at, or even to judge the range." - so any procedure that allows the Blackhawks to request permission from the tower for the Blackhawk crew to maintain (manage) their own visual separation from jets on short final is obviously a pretty stupid and unsafe procedure that is just waiting for an accident to happen... and eventually it did.

 

[Quickstatus]

The bread and butter skills I was referring to was my example about how this situation could have been handled safely.

Didn't say you were defending DCA ATC. See Last sentence in paragraph 4

 

[Quickstatus]

And why not do these training flights in the early hours of the morning when there are probably less aircraft taking off and landing?

Training should ideally occur at different times of the day/season and in all the usual flying conditions that a heli might be tasked to fly in including in congested airspaces.

so any procedure that allows the Blackhawks to request permission from the tower for the Blackhawk crew to maintain (manage) their own visual separation is pretty...

Not when the separation bubble and visual identification of conflicting aircraft invites "no room for error". A system should always have room for error.

 

[justinbrett]

The obvious problem with the procedures is what @jb747 said a couple of pages back - "At night? It's almost impossible to say who you're looking at, or even to judge the range." - so any procedure that allows the Blackhawks to request permission from the tower for the Blackhawk crew to maintain (manage) their own visual separation from jets on short final is obviously a pretty stupid and unsafe procedure that is just waiting for an accident to happen... and eventually it did.

That’s very absolutist comments. Just remember night VFR is practiced all over the world and in most countries is a specific pilot rating. While an advanced pilot skill I don’t think there’s any evidence it’s unsafe.

It’s going too far to say visual separation from aircraft on final is “obviously a pretty stupid and unsafe procedure”. Let’s say the Blackhawk was joining the circuit for runway 33. ATC tells him to enter a downwind for Runway 33, traffic is CRJ circling for Runway 33, maintain downwind report traffic in sight”. This is now tower control 101. Also note that tower don't usually need the radar to separate, up until and including this point the tower is providing visual separation. It's easy to focus on the radar as that is all over YouTube, that is not how tower control works. An almost identical procedure could have been used until the Blackhawk was behind the CRJ and then cleared to proceed south.

It may well be (and is indeed likely) that the traffic volumes at DCA simply don’t allow for these procedures to be done safely, and certainly if they did they require the dedicated helicopter control position and/or suspend Runway 33 arrivals while the helo route is active. In my mind, that is the #1 fail from ATC, if that position was staffed I'd almost guarantee this wouldn't have happened. But there's now been a few comments in this thread suggesting that core ATC procedures are not safe and that's simply not true.