Hacker News Reader

Comment by yladiz | Hacker News Reader

Short summary: the two planes were very close, “200 feet vertical and 0.2nm horizontal,” and according to the article the main reason the incident happened is due to miscommunication and oversaturated communication channels.

Initially the Wizz Air plane was supposed to descend to FL300, but missed the first instruction, and the second one was incorrectly given and confirmed (read back by the crew means confirmed?) as FL200. The Air Europa flight was at FL290, and the Wizz Air plane got to FL292 (200 feet from FL290) before being instructed to ascend. The cause of the miscommunication was that the frequency used to communicate from the air traffic control to the planes was saturated and many communications collided, meaning multiple communications happened at the same time and were impossible to differentiate, necessitating repeating.

Edit: wasn’t misheard, instruction was incorrectly given.

View on HN

> read back by the crew means confirmed

Read back does not mean confirmed, it's basically a double check. What pilots do when they get an instruction they'll respond with the same instruction (but abbreviated).

Example:

ATC: Descend and maintain FL300

Pilot: Cleared for FL 300, XZ 123

Whereby XZ 123 indicates the flight. By reading back the instructions ATC should notice if the pilot heard everything correctly.

That's entirely different than "confirmed". With confirmed there's no double check. Further, unless there's an emergency AFAIK the pilot must follow ATC so no need to "confirm".

'nm' is 'nautical mile', for anyone wondering

Yeah I did a double take. Close but surely not that close.

missed by a whisker -> 90,000 nanometers apart

Clearly a mistake. Abbreviations for nautical mile are M NM or nmi. I guess the author has not heard about nanometer

I guess you could take the context of the article. There would be no way to measure in nanometers in that situation. So if someone does not know that nautical miles was implied, it should be easy to infer that nm was not nanometer.

I did infer that, but was extremely confused because I didn't know what else it could be.

The source isn't really intended for general consumption. It's a briefing intended for pilots and others in the industry. The goal of these kind of documents is to help pilots and air traffic controllers to learn from their mistakes.

Sure, but that didn't stop me from thinking "nanometer" whenever I saw it. Initially, I was wondering what font screwup could turn "km" into "nm".

Same! I was wondering A) how in the world they managed to get that closr and B) how they managed to measure it.

In an aviation context, `nm` is almost universally used for (and understood to mean) nautical mile.

NM or nmi, but not nm.

Confusion in units is never a good thing.

The FAA uses "NM" (capitalized) in federal regulations.

I don't think I've ever seen someone use M, NM, or nmi for nautical mile. I've only ever seen nm. I used nautical miles a lot in a previous career, in the context of aircraft mission planning.

Telling people in the industry to use nmi or NM seems a lot like telling computer people to use MiB or GiB instead of MB or GB, or to pronounce gif "jiff". They're just not going to change.

It is going to change. I use MiB and GiB. I also use the ISO standard for dates, etc. Precisely because they are now standard.

Saying it is not going to change is what makes things harder.

The aviation industry: not exactly sticklers for metric usage.

The Apollo flights used NM on their screens for nautical mile.

[pic (right-most screen)] https://cdn.arstechnica.net/wp-content/uploads/2012/10/liebe... [article] https://arstechnica.com/science/2019/12/apollo-flight-contro...

Also "KFPS", presumably for "thousands of feet per second"

Thanks. 2 nanometers of distance would indeed be a bit low.

It's only a problem if the planes are both fermions.

Nautical miles are a fun unit. They are one minute of arc of the earth's surface (idealized), so it makes sense why they took hold in the two forms of travel that regularly travel great circle routes.

There is a similar relationship between gradians (metric angles) and kilometres https://en.wikipedia.org/wiki/Gradian

One minute of latitude for those wondering how this works.

Thank you! I was wondering how these work around the poles!

1nm is about 1/60th of a degree of any great circle on the earths surface

(Earth isn’t a sphere so it’s not exact, but it’s pretty close)

Not nano meter :)

1nm is just over 6000 feet, so that is about 1200 feet horizontally. 1nm is also about 15% more than an sm (statute mile), the thing we usually call a mile outside of aviation or nautical circles.

> 200 feet vertical and 0.2nm horizontal

So that's 375m (1230ft) total for anyone wondering.

What do you mean by 375 meters "total"? They were 0.2 nautical miles apart vertically, which is 370.4 meters, and additionally they were separated horizontally by 200 feet, which is ~61 meters.

https://www.wolframalpha.com/input/?i=%28%28200+feet%29%5E2+...

Presumably they plugged it into the Pythagorean theorem.

You can't really compare horizontal and vertical distances.

This is correct, and it is implicitly recognized in the vertical separation intervals being less than the horizontal ones. What matters even more than distance, however, is time, and in this case, fortunately, the aircraft were travelling in the same direction and so had a low closing speed (it was actually negative, I believe.)

Planes aren't spherical and don't change directions in all ways as easily... so taking the straight line distance isn't as helpful as you'd think.

Wow, usually when there are reports of "near misses" the planes are relatively far apart (still potentially dangerously close, of course) - I don't think I've seen a report with the planes this close to each other before.

Planes don't actually have to strike each other to cause issues.

The turbulence from the engines is enough to cause issues for the next plane. Enough so, that turbulence alone can cause a crash, even if the airplanes never actually touch.

https://en.wikipedia.org/wiki/Wake_turbulence

This is only really an issue for planes of highly disparate sizes and I've never heard of it causing issues (a crash or near miss) with a commercial airliner (edit: in good mechanical condition where the pilot does not make things worse from the get-go) sized craft being on the receiving end.

Wake turbulence was the cause of the second deadliest air crash in US history:

https://en.wikipedia.org/wiki/American_Airlines_Flight_587

And involved in this one, though the situation is harder to assess:

https://en.wikipedia.org/wiki/USAir_Flight_427

>Wake turbulence was the cause of the second deadliest air crash in US history:

>https://en.wikipedia.org/wiki/American_Airlines_Flight_587

Funny you mention 587. That was actually the first thing I thought of. That was the one where the pilot slammed the pedals lock to lock in response to turbulence and ripped the vertical stabilizer off. Aircraft of similar size take off similar distances behind 747s all the time. The pilot really reacted badly in the case of 587 though his training was partly at fault. From wikipedia:

"The first officer repeatedly moved the rudder from fully left to fully right... The resulting hazardous sideslip angle led to extremely high aerodynamic loads that separated the vertical stabilizer. If the first officer had stopped moving the rudder at any time before the vertical stabilizer failed, the airplane would have leveled out on its own, and the accident would have been avoided"

>And involved in this one, though the situation is harder to assess:

>https://en.wikipedia.org/wiki/USAir_Flight_427

427 was a hydraulic failure and any source of turbulence (e.g. weather) requiring rudder movements could have raised that issue. IT was part of a larger issue.

https://en.wikipedia.org/wiki/Boeing_737_rudder_issues#Secon...

I originally said "Thank you from changing the wording from "implicated" to "cited" because they mean very different things." but you since edited your comment to say "cause" and "involved" which IMO is as accurate on average as "implicated" but with greater standard deviation.

I don't think it's reasonable to say that wake turbulence has never caused an issue for an airliner, when many hundreds of people died because pilots misapplied rudder in direct response to wake turbulence.

The correct systemic response is to note that wake turbulence is very dangerous for airliners because it may cause pilots to catastrophically misapply rudder.

(Sorry for my many edits to the parent, I didn't realize you'd already replied.)

I don't think it's reasonable to say that N=1 warrants a systemic change when the N in question was a human (as opposed to a mechanical failure which is as repeatable as the components and the failure conditions are).

I don't know but I suspect United did change their training after that. The fact that the pilot reacted the way he did is as much their fault as the pilots. IMO it is a training/human factors issue, not some fundamental issue with turbulence and current following distances.

But the point of systemic thinking is to see when you've hit an area that's vulnerable to human and training factors and try to do the best you can to reasonably remove the risk. There are surely hazards that have caused less loss of life that we still consider dangerous at N=0.

Do you know whether ATC large plane same-size spacing guidelines around airports have changed much over the years? I don't know whether airliner spacing would have generally increased or decreased, but I'd be surprised if it hasn't changed at all.

Edit: I read more about these flights and I think I agree with you that I'm overstating the risk to airliners from wake turbulence. I was expecting to find increases in separation distances in response to the American flight and instead it looks like the only change made was a training change for rudder control.

American Airlines' training division had taught its pilots to react vigorously to wake turbulence, and among pilots generally, there was a widespread mistaken belief that, below the maximum maneuvering speed, it was impossible to do structural damage through the primary flight controls. Now that these misunderstandings have been corrected, we can hope this particular problem will not occur again.

On the other hand, I would guess that there is still a risk of people-thrown-around-the-cabin type incidents from abnormally close encounters with wake turbulence, and I do not know if the possibility of structural damage can be ruled out, especially if this were to happen at cruising speeds. I think it is probably safe to say that any structural-risk incident would have been within a whisker of being a collision anyway, in which case the measures to avoid the latter cover the former as well.

FWIW, the dangerous part of wake turbulence is from the wing vortices, not the engines.

The Air Canada near miss in San Francisco is another one.. but the other planes were on the ground. Apparently they missed by ~20 feet and it would have likely been the most deadly accident in aviation history as 4 planes were in the path with something like 1,000 passengers in total.

Video: https://www.youtube.com/watch?v=aGQlQFn0euI

Wiki article with the story: https://en.wikipedia.org/wiki/Air_Canada_Flight_759

It was not misheard, the controller got confused and said FL200 instead of FL300

Thanks, I updated my comment.

Dont they have ADSB in on the commercial flights? They could have known another plane was so close?

Yes.

I believe that’s how TCAS (Traffic Collision Avoidance System) works; ADS-B broadcasts aircraft data which is received by other aircraft, and an “advisory” is issued to the crew if a potential for collision is detected. This system operated in this case, for the lower aircraft of the two.

I think the parent commenter is pointing out that the descending aircraft could (and arguably should) have refused the descent clearance on noticing that it would lower them into FL290 traffic that they could see via ADS-B display in the cockpit, long before TCAS fired.

So that means the planes were intended to be at levels 300 and 290. Is it normal to have only 1000 feet of vertical separation? It seems like the higher plane could easily lose 1000 feet in a few seconds before ATC could tell the lower plane to move away.

Yup; https://en.m.wikipedia.org/wiki/Reduced_vertical_separation_...

A plane can ascend/descend fast, but generally stick to either 500 or 1000ft/min.

Fractions of nanometers close? That's some insane precision.

In aviation “nm” stands for nautical miles.

It’s usually written NM.

Regardless of what’s technically correct, I see “nm” way more often than “NM” in aviation contexts.

Reply on news.ycombinator.com