lateral separation in DOC4444

[ferris]

Look up the definition of 'diverge'. Especially in the context of lateral separation.

Imagine what would happen if the thread starter really is an instructor somewhere, and is learning how to do/teach procedural control via pprune? The mind boggles.

[10W]

Whilst I fully agree that the information on PPRuNe cannot be assumed to be correct, I'm not sure why you pick on DFC's post. Or am I missing something here?

What Ferris said

[Rule3]

Missing Something

Spitoon, yes you are missing something. Read zkjaws Post 3 and DFCs extrapolation to the ridiculous. I believe we are talking about lateral separation, not opposite direction passing. Divingduck has the answer in a nutshell. I knew all that time at Henty House wasn't wasted.

[RustyNail]

Just to Confuse things more

DFC

The radials in your example are for Reciprocal tracks - 15 degree sep doesnt even apply in your case until A/C are on the same side of the VOR and the tracks are diverging.

In Canada the 15nm sep applies for inbound and outbound, but, for 2 A/C inbound to the VOR the 2nd A/C has to be 15nm or more from the VOR and also 10 minutes flying time away.

Yes, the book says NM (not DME)

[DFC]

ferris et al,

The point is obviously lost on you - if you work to the exact letter of an isolated paragraph of DOC4444 without taking the consequences into account, you end up with what I said.

The point is that one can not extract a small paragraph from DOC 4444 and use exactly what it says (which I did) as the sole basis for establishing a certain separation standard otherwise one can have someone using the exact words to do something like I described.

So that there is no doubt, an aircraft flying outbound along the 270 radial is diverging from an aircraft travelling outbound on any other radial.

Similarly, an aircraft inbound on the 270 radial is converging with an aircraft inbound on any other radial.

DOC 4444 sets a minimum angular difference there is no specified maximum.

Same direction, opposite direction and crossing tracks are defined with respect to Longitudinal Separation. That is not what we are talking about here.

The debate for years regarding that paragraph was if the aircraft had also to be diverging. Many providers required the aircraft to be diverging. With such a stance one could never end up with the situation I described.

Separation is established according to an array of issues which can not be established based on a single paragraph from one book.

That is what the planning manual etc etc etc provides the guidance for.

Regards,

DFC

[ferris]

The point is obviously lost on you

Errr.. No, my point was that you can't use pprune to learn about ATC on a professional basis.

So that there is no doubt, an aircraft flying outbound along the 270 radial is diverging

Err, that's not what you used in your example??

Many providers required the aircraft to be diverging.

The aircraft, or the radials...?? Or does the word "diverging" in the standard somehow sleep? Keep digging that hole.

Separation is established according to an array of issues which can not be established based on a single paragraph from one book.

Now you've really got me intrigued. Doc 4444 is full of paragraphs that can be used for separation. That's, pretty much, what it is.

I'm just going to give DFCs confusing posts a wide berth, and merely re-state that it is highly irresponsible to be seeking guidance on pprune at a professional standard. Punters/pilots/whatever seeking answers out of interest/ad hoc- fine. Alleged instructors- not on.

[DFC]

The aircraft, or the radials...??

The aircraft obvously.

VOR Radials both diverge and converge at the same time. i.e. The 270 radial and 260 radials diverge away from the VOR. Those very same radials coverge towards the VOR.

In order for the aircraft to be diverging in all cases they have to both be outbound.

As I said, DOC 4444 is just one of many documents that ICAO provide so that Authorities can set separation standards.

Annex 6, 10, 11, DOC 8168 and the various Planning Manuals are very important parts of the process.

I do agree with you on a very important point - everything one reads in discussions on Prune is simply people's opinions.

Regards,

DFC

[topdrop]

Now my point about quality of response is nicely made.

So ferris, and others, tell me are these two "opposite direction" aircraft always separated or not?
One acft inbound on 270 radial, the other inbound on the 090 radial. Second aircraft is assigned 7000 on descent. First is cleared visual approach, requirement to reach 6000 or below by 15NM (In Oz you could use 8DME based on 5.4.1.1.1 from doc 4444)

If you agree they are separated, tell me, are they separated using a longitudinal standard or a lateral sep standard?

Doc 4444 says the radials diverge by at least 15°. It doesn't say the aircraft have to be diverging - they can be converging as long as they are on radials that diverge by more than 15°. Now let me try my maths - 270 minus 090 means the radial diverge by 180°, just slightly more than 15°.

[Spitoon]

I ank a simple question and then go away for one night and look what I started!

But topdrop has identified the key point. - Doc 4444 says the radials diverge by at least 15°. It doesn't say the aircraft have to be diverging.

Now I feel sure that the intention for the VOR separation was that the aircraft were diverging - but it doesn't say that. I might be considered to be splitting hairs with this but it is ambiguous. As Hub47 says, in the UK, where much of the content of Doc 4444 is published in local rules, the ambiguity has been removed - I believe this was done following discussion about the poor ICAO text.

[ferris]

I dunno- it seems glaringly obvious to me? If an actual student was posing these questions, here's what I would say.
Step 1 in solving any problem is to identify the circumstances, so you will know where in the tool bag to look for the solution. If you identify the opposite direction scenario described as a "lateral problem", you wouldn't be getting very far in the ATC world, in my view. i.e. can you draw a lat sep diagram? If not, it's pretty clear it is not a lateral problem, and cannot be solved using lateral standards, no matter how loosely you think they are worded.
If you are trying to be a smartarse and attempt to apply the solution described because you think everyone will think you are really clever (as often happens in ATC), then I'd suggest you will also 'come a croppa':
Firstly, that the radials must "diverge", would, to me, indicate they cannot be parallel lines. They are lines in space, and when looked at two-dimensionally (lateral problem), parallel lines cannot be said to "diverge". Even two different lines, with different start points. That's just the "reasonable man" test. I think you would fail miserably even if you were trying to say that the lines are different, and that they have an end point at the VOR. They still have an angular difference of zero. Not 15 or greater. Trying to read a document without regard to it's intent is a very poor basis for argument- legal or otherwise.
Yet even if you got past that, the killer is in 5.4.1.1.2 "Lateral separation of aircraft is obtained by requiring operation on different routes....." etc. Kidding yourself that the wording is loose enough to get away with something just doesn't cut it. Because YOU wont be the one applying the 'reasonable man' test if it goes wrong. There is even a diagram which rules out trying to say "different geographic locations".

And, once again, rams home the point about instructors learning their trade via pprune.

[topdrop]

And, once again, rams home the point about instructors learning their trade via pprune

Certainly does, because what you posted above it is basically crap.

can you draw a lat sep diagram?

Yes I can.

I think you would fail miserably even if you were trying to say that the lines are different, and that they have an end point at the VOR.

The radials do end at the VOR. Using 5.4.1.1.1 (Figures are those used inOz). Draw your VOR tolerance, 5.5° The min tolerance around a navaid is 4NM from memory and voila, there is your area of conflict. Now you need the aircraft to be 1NM outside the area of conflict, so that makes 5NM,add slant range and DME errors and you come up with 8DME as lat sep. However, if you don't give a requirement to the first aircraft to be vertically separated, you have to add the tolerances of the holding pattern.

You obviously haven't worked a procedural tower because they use this situation as lat sep every day of the week. Or do you insist it's longitudinal and they have to use 10 mins prior to time of passing??

Even if you work enroute procedural you should know that this is lat sep, or has automation (TAAATS) taken all the skills away? Do you just use the Time Of Passing function and require the aircraft to be vertical 10 mins prior. - works out to about 160NM apart for two jets as against 16NM using lat sep - how bloody inefficient is that - in fact, as an ex-checkie I'd probably fail your check.

[ferris]

OMG- WTF are you talking about?

I had heard rumours about procedural towers doing some interesting things and inventing standards, but here you are broadcasting it to the world!! I am not a tower controller, and am aware there are a few more tools in the bag/books, but if people are doing what you describe and using 5.4.1.1 (lateral separation) as the basis, I hope no-one ever finds out. Are you even aware that the opening paragraph states that aircraft are required to be on different routes to be able to apply lat sep, or that the angular difference between parallel lines is ZERO?

You are not still employed, are you?

[topdrop]

This lat sep in procedural towers is not an invention - it's obviously just something you can't come to grips with.
They are on different routes - 180° apart. In fact, they are further apart than two aircraft on tracks that vary by 80° - or can't you work that out either.

I suggest you go and talk to a controller - in fact better talk to a few of them as you won't believe just one - who has worked a procedural tower and ask him/her whether my example is laterally separated or not. I await your answer.

Still working after 30 years and by the looks of it forgotten more than you'll ever know.

[DFC]

topdrop,

how does 8nm account for slat range errors? What is the maximum level at which you can use that separation?

ferris,

You seem to say that an aircraft flying westbound along the 090 radial is parallel to an aircraft flying on the 270 radial?

How can two aircraft which are either head to head or in trail be flying parallel?

If you and topdrop stand back to back and he walks off on a track of 270 and you walk off on a track of 090 are you diverging?

If after 10 paces you turn round and walk back towards each other are you not converging towards the point you started from?

Simply put - aircraft following parallel tracks would have constant track separation. Aircraft following diverging tracks would have increasing separation and aircraft following converging tracks would have decreasing separation.

That isn't ATC it is simple geometry!

However, if you want the offical answer -

Routes are parallel when they;

Have about the same orientation i.e. the angular difference does not exceed 10 degrees;

they are not intersecting; and

traffic on each route in independent of traffic on the other route

The 270 radial from a VOR is a different route from the 090 radial. In simple terms both routes have a common point (intersection) - the VOR but that is the only point that is common to both routes. Perhaps you are mixing this up with the fact that separately, an ATS route may pass over the VOR following the 090 radial inbound and the 270 radial outbound. However if you are an approach controller with your VOR in the middle of no where you have 360 possible inbound and outbound routes!

If we look at the simple geometry of the question. the track separation between two aircraft at 15 miles fromthe VOR on radial different by 15 degrees is 3.88nm.

I have no problem with two outbound aircraft who are diverging i.e. the separation is increasing being cleared to the same level.

However, with two aircraft inbound, I would not be comfortable with vertical separation not being established until the aircraft which are converging are only 3.88nm appart.

Work that with the fact that a pilot can report established on the radial provided that they keep it within half scale deflection i.e. +/- 5 degrees and then add in the system tolerences and you can have two B747s converging and very close together!!

If on the other hand we use the more common 8nm as a target separation we end up with a figure of 30.9nm from the VOR. i.e. The latest to establish vertical separation for a theoretical 8nm separation between aircraft inbound is 30.9nm.

Sound familiar to anyone?

Simple trig - separation(nm) = distance from VOR(nm) * Sin Angle

Regards,

DFC

[ferris]

You seem to say that an aircraft flying westbound along the 090 radial is parallel to an aircraft flying on the 270 radial?

Yes, it is. The lines (routes) have no angular difference, and are by definition, parallel.

How can two aircraft which are either head to head or in trail be flying parallel?

This statement demonstrates exactly why you can't/shouldn't be using lateral separation for either example described. You are admitting the aircraft are longitudinal, and trying to say that is why they aren't parallel!! Then turn around and advocate the use of a lateral standard in a longitudinal situation!!!

That isn't ATC it is simple geometry!

You'd better believe it!!

Perhaps you are mixing this up with the fact that an ATS route may pass over the VOR following the 090 radial inbound and the 270 radial outbound

I'm not mixing it up- it is the example you gave. The wording of Doc 4444 uses "different routes" for a reason- trying to say that two aircraft flying down a single airway which passes over a VOR are "different routes" because they are on different VOR radials (the 090 until over the VOR then on the 270 radial) is really clutching at straws and trying to bend the wording to suit your argument. Once again, the "reasonable man" test would bring you undone, and I can't believe any oversight authority would allow that interpretation. Topdrop further complicated things by adding the arrival/departure element, but I still disagree with his version (for the same reason- misapplication of a lateral standard in a longitudinal situation).
You then go on to say that you believe you can't use the standard for two aircraft pointing at each because they will be too close before lateral sep is lost and the vertical must be in place!!! Either you agree with the standard, or you don't!!!! You want to infer things (from the "diverge" term), even though you are quite happy to misapply the standard by using loose interpretation of other, clearer, words!!!!
To me, it is absolutely clear what the intention of the wording of the standard is. I just hope Topdrop has misunderstood the derivation of whatever standard he is talking about.

[DFC]

Ferris,

First of all you seem to have missed the point that I do not like the idea of two aircraft inbound to a VOR with no vertical separation until 15 miles be they on radials 15 degrees, 90 degrees or 180 degrees appart!



Longitudinal Separation can be applied if you want - it is optional in the case when you can be more efficient by using a lateral separation.

No where is a certain separation required to be provided - provided that one separation standard is safely applies then the aircraft are separated.

However, lets use your idea and provide longitudinal separation eventhough we have a VOR/DME and aircraft are tracking towards and/or away from that VOR.

Firstly, if you use longitudinal separation based on time then two aircraft outbound on radials that diverge by 15 degrees are considered to be on the same track!!! So you are going to have aircraft a long way from the VOR) and a log distance appart and still not be able to change levels!!!

But of course you could use separation based on DME and then you could get as low as 10 dme between same direction aircraft even on the same radial!!

However, this is a discussion about lateral separation and as I said previously, if one restricts the 15 degree or more to outbound aircraft only then it is impossible even when pushing the written example to the maximum to end up with what I said.

Slight diversion - Using radar assigned headings aircraft A on a heading of 090 is pointing at aircraft b heading 270. There is no wind. You are going to tell me that they are on parallel headings or would you say that they are converging?

Regards,

DFC

[topdrop]

The slant range error varies with altitude - we have a table which states what to add depending on the level - the example I gave means the slant range error is up to 1NM, so we use 1NM.

Ferris, the example I gave is for 2 aircraft, one flying A to B, the other C to B. The routes are different. Tell me where they overlap and don't say at B because I'm already using a different standard (vertical) prior to B.

A lesson for Ferris
You can still use the lat sep for acft going A to C via B and one the opposite direction. You only need one form of separation at a time. Use lat separation up to the lat sep point approaching B with a requirement to have vertical prior to passing that point.
I agree with you that the aircraft have to be on different routes when you are using lateral separation. But they don't have to be on different routes when you are using vertical. So first is on route from A to lat sep pt approaching B (let's call it AB) and the other is on a route from C to a different lat sep point approaching B (call it CB). So when I'm using lateral separation one aircraft is on route A to AB, the other is on route C to CB - two different routes. From the lat sep points I'm using vertical, so they can be on the same route from there.
I've probably blown poor young Ferris's mind, he's still trying to work out how a procedural tower uses 10 min prior to time of passing for two aircraft on opposite radials inbound to B. I suppose it could be done, but you'd have to work bloody hard with all those step descents.
And Ferris you don' t need a Tower rating to use these examples. They were a common occurence on the sectors I had procedural ratings on - used by all with the rating including checkies. In fact, if you tried to used 10mins prior to time of passing all the time, the traffic would still be airborne.

[ferris]

Ok, so DFC doesnt believe Topdrop can do what he says is common practice. Moving on....

Topdrop, as I said, I can't really comment on what towers do, as I am not a tower controller. What I can comment on, is that if you are using lat sep to provide the sep you described with 5.4.1.1 as it's basis, then I can't see it being legal. If it ever went wrong, I'm pretty sure the lawyers would have a field day, and whoever ended up being responsible (obviously not the controllers if even checkies (!) believe this is an acceptable use of 5.4.1.1) would go for a row. If there is some tower voodoo that you are using and mandated by the authorities in Australia in the oz MATS, then go for it. I am aware of a couple of unique Australian 'anomalies', and maybe this falls into that category. I can see how IF THE VOR WAS THE END POINT OF THE ROUTE OF FLIGHT FOR BOTH INBOUND A/C how you might be able to construct a lateral sep diagram using "opposite sides of the aid" or some such thing, but not 5.4.1.1 (for the reasons already given). I dont really want to talk much more about it here, because it's a pointless exercise. It's a divergence (!) from the en-route example DFC gave, and that's what I feel I can speak confidently about (without blowing my young mind- love it!!).

DFC.
I have no doubt that you only need one standard, that some are less efficient than others etc etc. and do get your statement that you "don't like" the idea of nose to nose at 15 nm. But if you believe you can use lat sep VOR 15 deg. for longitudinal problems, why would you be uncomfortable with the standard? A standard exists, or it doesn't. If both a/c are approaching from the same side of the VOR on radials 15 deg apart, the aircraft are converging, but the radials they are flying on are diverging. I am more than happy to run a/c that why, achieving lat sep by the specified point. You meet the fundamental criteria of only having one a/c inside the area of conflict while vertical doesn't exist. That seems to me to be the fundamental problem with your argument- how do you draw the area of conflict with an opposite direction a/c? When I worked procedural, we had angular difference charts on the console, and even though my memory (young mind!!) may be a little rusty, I'm sure the angular difference charts didnt extend to 180 deg.

As for your radar question (which is starting to look like a straw-man argument, because the meaning of terms may change between the different disciplines)- I would describe those a/c as being on reciprocal headings. "Parallel headings" means a particular thing in radar control, whereas "parallel" in geometry is as I have described "lines having an angular difference of zero". One continuous line with a VOR in the middle of it is no different to two lines with zero angular difference which meet at a VOR. How you can say they have "a divergence of 15 deg or more" in order to meet the standard is, well, fanciful. And (critically IMHO), certainly not the intent of the standard, either.

[DFC]

OK ferris, where do you draw the line?

Is 15 degrees OK? then what about 30, 60 or even 90?

Your argument that 180 degrees between the radials then it must be longitudinal separation and not lateral separation must also mean that with a 90 degree difference the longitudinal separation relating to crossing tracks has to apply?

Where do you find a limit to the 15 degrees or more specified in DOC4444

My reading of that means anything from 15 to 345 degrees is OK.

If both a/c are approaching from the same side of the VOR on radials 15 deg apart, the aircraft are converging, but the radials they are flying on are diverging.

No. the radials are also converging towards the VOR.

The 270 radial is nothing more than a line in space. The angle measured clockwise between the direction of Magnetic North and the direction of the line defining the 270 radial (measured at the VOR) is 270 degrees. That is it.

An aircraft can track both ways along a radial. If travelling towards the station they will be converging with other radials. If travelling away from the station they will be diverging from all other radials. Or put another way when moving closer to the station the lateral distance between the radials gets smaller (converging) and when moving away from the station the lateral distance between the radials gets bigger (diverging) they are still the same radials in both cases.

If however, it is as you say i.e. radials only ever diverge then if that was the case why would it ever be necessary to use the word diverge when talking about radials and lateral separation?

Perhaps you are being confused with the NDB situation and with QDRs and QDMs? - QDRs always diverge while QDMs always converge.

Regards,

DFC

[Spitoon]

Quote:

If both a/c are approaching from the same side of the VOR on radials 15 deg apart, the aircraft are converging, but the radials they are flying on are diverging.

No. the radials are also converging towards the VOR.

When I learned about this stuff I was taught that a radial, by definition, is a magnetic bearing extending from a VOR navigation facility. Hence radials always diverge.

It's interesting that this used to be stated (maybe even defined) in ICAO Doc 8071 and in the ICAO Vocabulary - now there's no definition that I can find and it doesn't appear in the latest edition of the Vocabulary.