How to manufacture a Mid-Air.

[forget]

This business spends millions of dollars, and man-hours, keeping aircraft from bumping into one another and then, thanks to RVSM and Nav equipment that works in inches - we get this. If they’d ‘bumped’, and they don’t come closer than this, how would it be explained to the man in the street who has a rough idea of the area of the North Atlantic?

2 OCT 2000.

UK AIRPROX/AAIB SERIOUS INCIDENT. A340 and A330 in Oceanic airspace at RVSM levels at N5828 W1646 at FL370.
Severe turbulence encounter. The A330 pilot reported that, as his aircraft was overtaking the A340, which was 1000ft below and slightly left, he observed the wings of the A340 starting to bounce as it hit turbulence. A330 then hit turbulence and received a TCAS RA ‘climb’ as altimeter showed 36,800ft. The A340 was then seen to climb rapidly and pass within 200 metres of the A330 to a level 1100ft above it. The A330 pilot then turned right to allow the A340 to return to its original flight level of FL360.


[This message has been edited by forget (edited 28 November 2000).]

[ATC Watcher]

Now you understand why it is agood idea to offset (up to 2NM is allowed in NAT /RVSM airspace ) and why after all this time the whole thing is still under TRIAL...
RVSM and CAT do not mix well. We said this since the begining.

[moodymoosey]

All looks good for RVSM in Europe 2002....of course this is why Eurocontrol are trying to press ahead with 'Free Route'. This is the concept of filing a flight plan which is essentially a direct track from dep to dest. Minor route adjustments sent via datalink when conflicts arise. Sounds easy but expected delays due software lagging behind the actual idea, plus results not received from VHF datalink trials (I think they have started now)

Not sure if Free Route is planned for NAT though.

I once read an early document from a Eurocontrol think-tank which outlined their 10-15 year plan....it showed a potential system where these route adjustments were filtered through the airline ops and then passed to a/c. Just something to help everyone sleep better at night....

mm

[4HolerPoler]

Your words are a sound for sore ears ATC Watcher - I'd appreciate any reference you have in regard to the allowable offset to arm myself for the next soul who criticises my mile to the right.

[Blacksheep]

This sort of thing will become more common. There is disconcerting confidence in the ability of technology to solve all humanity's problems. Air Traffic Control Authorities place increasing reliance on new technology to overcome congestion. We have already had 8.33 KHz VHF reducing the radio band seperation to just one third of human audio perception. Basic RNAV reduces track errors to within one mile while RVSM reduces vertical seperation to 1000 feet above FL290. These were all imposed as legal requirements before the necessary avionic equipment was designed, tested and certified in older aircraft. The belief seems to be that anything can be achieved as long as it is made mandatory. The TCAS experiment continues and I use the word experiment advisedly. (For those familiar with the technical side, consider the messy 'Gilham Code' AD focussed on analogue/digital conversion of altitude data for use in TCAS systems on non-digital aircraft.) EUROCONTROL will soon announce the introduction of rules to make RNP1/3 mandatory. ATC watcher's offset will be no use then, anything more than 750 metres will put you in the opposite lane as it were!

Is the equipment capable of these degrees of accuracy? I don't know and nor does anyone else. Although new deliveries of the latest models come ready equipped with the latest technology, most aircraft need major retrofit modifications. Not everyone has GPS (not as foolproof as many people think it is!) SatCom, ACARS or the latest Laser INUs capable of meeting the tighter RNP limits. Although equipment is already available and flight testing has been conducted on new built aircraft, long term equipment reliablity at the increased accuracy levels has never been established. Yet legal requirements will be imposed regardless, and in general by authorities that have no expertise in the details of installing and testing the systems.

The US Congress forced TCAS on the aviation world in the face of protests from avionic experts that equipment development would take more time. They have a lot to answer for now that they have established a trend.

**********************************
Through difficulties to the cinema

[410]

Since I first started pushing this particular barrow, (up a seemingly very step hill), I've never ceased to be amazed at the resistance by fellow pilots, (as well as the perhaps mre understandable resistance by ATC and regulatory authorities), to this oh so simply achievable quantum leap in flight safety.

4holer, you give me heart to learn that I'm not the only one out there offsetting. The only way it's ever going to instituted in legslation is if so many people start doing it the powers that be simply can't ignore it.

[7x7]

Dream on, 410. Most pilots have a resistance to change that would put the Vatican to shame. Even if it was to be made compulsory, there'd be some who'd simply refuse to do it or offset one mile left just to be different.(!)
I've had a conversation on this subject with a mate in Australian CASA and he assures me the mathmeticians have proven to him that offsetting INCREASES the chances of collision. Something about having to re-draw the existing airways' dimensions to allow for it.

Can't see it myself, but there you go...

[Narada]

The "offset" topic comes back once in a while. It seems like a good idea. Were there any studies conducted to formally evaluate the merits/demerits of offsetting (say 1 nm to the right on en-route)?

7X7 - how was your friends convinced? What were the assumptions the mathematicians were making?

I would appreciate any references (on both the questions above).

[tired]

Yeah, I've also heard the mathematicians "prove" that an offset makes things more dangerous. All I can say is that it gives me a warm fuzzy feeling when boring through the dark African skies to see the opposing traffic slide past a few miles off to the left, rather than straight down the centreline.

I'm sure the above-mentioned maths boffs all have IQs greater than 200 - pity they can't see the practical side of things sometimes!!

[ATC Watcher]

Quick replies :
4 holer : the ref is in ICAO RVSM trial procedures for NAT . It allows pilots to offset 2 NM to the right without having to advise ATC . But only valid in that airspace for the RVSM levels currently used (not all yet )

Moody : The Free Route Airpace Concept (FRAC) will be initially limited to 8 States ( Scandinavia + Benelux + Germany ) so definitively not in NAT and above FL 285 or 295. No data link in FRAC, only todays avionics . Feasibility depends on large scale simulations to be held in Bretigny (F) next year. Basically is to officialise in daytime what we already do during night.

410 : keep the faith..

[PPRuNe Towers]

The maths is definitely correct. It is also total bunkum.

The ATC structures with the wherewithal to employ or contract these boffins are the large busy ones such as LATCC, Eurocontrol, Rhine Etc. They also end up with the power delegations at the conferences and meetings that review theses things. Their research is the only research and as such their findings are accepted as correct for all airspace.

Because of their own interests they, quite understandably, look at extremely complex airspace, for example around FFM. Through flights, multiple airports, climbing and descending traffic all enter the equations.

They totally ignore the encounter time of the vast majority of air transport traffic spending the vast majority of its flying hours sitting in the damn cruise on a standard five mile airway well away from their intellectually stimulating little hot spots.

These potherbs need to spend some time on a flight deck somewhere over the surface of planet reality. A planet with mindboggling amounts of silent, empty upper airspace while all the traffic flies frighteningly accurately through the eye of a needle (thanks to mathematicians).

------------------
Regards from the Towers

[email protected]

[This message has been edited by PPRuNe Towers (edited 26 November 2000).]

[willadvise]

Offsetting-An ATC perspective
In radar airspace. I think it very much a bad idea. Your 1 or 2nm offset may infact put you in conflict with another route. The controller is expecting you to fly a particular track and makes judgements based on your predicted track only to find you a mile or so closer to that other aircraft that you were going to slide nicely 5nm past.
Outside radar coverage:-
Not such a bad idea. Especially in areas where you don't particulary trust ATC. But here are some things for ADS/CPDLC aircraft to consider (Australian airspace)
i) When you fly an offset it is immediately known to ATC. Personally I don't give a hoot. The current rules do not allow me to ignore the fact that you are flying an offset
and because I know you are flying an offset I have to take that into consideration when working out seperation problems and this adds considerable complexity and delays to the calculation.
ii) When flying an offset the data from your ADS does not automatically update the data into the ATC computer as it would if you are flying on the route (because you are not on track it doesn't recognise that you have over flown any waypoints). You will probably be asked to give CPDLC positions.

A Singapore Airlines pilot told be the other day that it was company policy to fly an offset. Is this an official policy or just a policy of the pilots or was he telling me porkies?

Another pilot the other day was complaining to a colleague about not being able to get a particular level despite him being about 10nm away (outside radar coverage). He said that his GPS was accurrate to 50m and why should he be held up. My colleague's response was "So you would be happy to have another aircraft 50m from you?" This shut him up but I thought it was an interesting question. Just how close are you prepared to have another aircraft from you. Pilots often complain when they can't have what they want, but the line has to be drawn somewhere!

[forget]

It’s interesting to see that this ‘incident' is to be investigated by the UK Air Accidents Investigation Branch. Rarely does this happen unless an incident is considered to be a gnats away from a disaster. No doubt the AAIB will produce a highly detailed investigation. Even now, they may be pondering the ‘procedures’ by which two air-liners can be deliberately brought within spitting distance of one another – and over the ocean. Granted that off-set track isn’t part of the equation when aircraft are pointing the same way - but RVSM! – we’ll live to rue the day.

[PPRuNe Towers]

willadvise,

You don't have to be around these forums to realise the immense respect we have for our colleagues in ATC. However, your post uses all the reasoning that fills so many of us with dread.

To your way of thinking the more accurately we fly the better you can do your job and the more effiently ATC can move traffic.

We are the traffic, not data tags. Anyone travelling on the flightdeck can see that we are now using a 100 metre swathe of that 5 mile (assuming it is drawn between VOR's with the required legal distance) or wider airway.

As forget points out, throw RVSM into the mix, as we're facing here in Europe, and you must surely understand our desire to offset. The major crossing points are going to present some very interesting and salutary experiences.

Why not allow a couple of hundred metres of limited offset in that complex airspace to cover the majority of our encounter time? It must already be built into the calculations previously mentioned as a buffer. Let's use part of it.

As controllers in 'first world' complex airspace you have an entirely undue influence on the wide open airspace we spend the majority of our time traversing.

An off hand reference it being understandable in dodgy parts of the world only serves to illuminate our frustration with your senior colleagues. Get them on a flight deck to see just how our navigational accuracy combined with your tight, accurate and efficient controlling gives us a very nasty sensation in our trepidation glands.

And while they're at it let's have them in a sim to at least give them a taste of what high level windshear, CAT and turbulence from passing traffic can do to a 1'000 feet of vertical separation at upper flight levels.

------------------
Regards from the Towers

[email protected]

[This message has been edited by PPRuNe Towers (edited 27 November 2000).]

[willadvise]

Thanks for your comments PPrune Towers but I must say that I am very surprised by some of them. You are right in saying that us "first world" controllers are fortunate to have good training and equipment.

I have never had a pilot complain I have been under separating them.

Whenever I have offered a RVSM level, I have never had any of them refuse it. This is what surprises me about your statement. If you are not happy with RVSM then tell someone. But I suppose you don't really have much control over it as its the bean counters who really want RVSM.

A couple of hundred metres in an offest is not really a problem. It would be very difficult to detect on a radar screen. But when off radar coverage the same problems as listed above will still apply. I am perfectly happy for you to do it as long as I am allowed to consider you "on track" for calculations.

[410]

Nicely said, Towers. Repeating myself ad nauseum, I am still goggle-eyed at the resistance to this idea from people who should know beter. It should be built in to all FMS equipment to kick in automaticaly above (say) 10,000'.

In an ideal world, we'd always fly on unidirectional airways. (I believe Eurocontrol are working towards that, and all power to their deliberations and may it cme t pass soon.) But it ain't gunna happen tomrrow, particularly outside Europe, so please, someone in authority, introduce 1 mile right offsetting as an interim measure before another incident - or even worse, accident - occurs.

All you journos out there, this is one subject about which you'll quite likely get lots of co-operation from pilots if you chose t make a feature of it.

[willwc22]

I'm quite amazed how the A300 slipped in from nowhere..... 3 aircraft in less than 1000ft airspace. Very impressive.

[LowNSlow]

Go away willwc22 until you can contribute something meaningful to an extremely important topic.

[Groundloop]

Even if it became standard procedure to offset to one side of the track, in this particular incident as the A330 and A340 were both heading in the same direction they could have the same offset programmed and so the result would have been the same. As the overtaking aircraft was visual with the lower aircraft would it not be be a good idea, in this type of situation, to fly slightly right or left of track until past? Obviously may not be as easy under radar control.

[RATBOY]

Though I'm not familiar with the Austrailian system it seems intersting that an offset could cause the ATC computer's tracking algorithm to not detect that the aircraft was the one on a nearby planned route of flight. In the terminal environment in the U.S. (and I think enroute too) the secondary radar (beacon) gets a target report, correlates with the primary target and matches them up if they are at the same point in space, tags the data block to it and updates the track file that the display is made from. Uncorellated targets are worked on the next time around and if still unresolved taged as unknown ID speed and alt unknown.

The Automatic Dependent Surveillance bit is very attractive to the beancounters until you tell them that it is only as accurate as the aircraft's knowledge of it's own position and that though it has been made to work in tests it hasn't been used much in the real world, unless you count United transPac to Oz with FANS, which I understand was a big bust.

Combine with that the fact that the data link standards are not set (1090, VDLMode 3, VDL mode 4, etc are possibles) it will be a long time before anything like this can be really operational.

It would be very interesting and instructive to gang a bunch of simulators together with an ATC and try RVSM, ADSB, Free Route and other things in a virtual world. though it would cost millions it would be a lot cheaper and safer than the possible bent metal, let alone wasted money, it there are some unexpected fatal flaws in the concept.

[Wiley]

willadvise, can you confirm that an offset as small as 1 mile will cause the ATC computers to spit the dummy and not update an aircraft’s position on your boards? (And can your enroute radars detect a track error that small?) I can only accept that you know what you’re talking about. However, as a (very non-mathematical) flying layman, I’m surprised to learn that, given that there are many aircraft still out there that are not GPS or IRS-equipped that the ATC computers can’t cope with an aircraft that is flying extremely accurately 1 mile off the airway centreline. (Non IRS/GPS aircraft can and do wander considerably more than 1 mile off the exact airway centreline.)

I feel that there is a maybe all too human reaction on the part of many ATC people that in asking for the offset to be legalised, we pilots are somehow slagging off at ATC and impugning their abilities or professionalism. I’d like to state quite clearly here that that is not so. (The analogy would be that in practising engine failure drills, we’re impugning the abilities of our engineers – or the abilities of flocks of seagulls to find their ways into engine intakes.[!]) We just believe that it is a very easily achieved extra layer of safety – and surely, we should all be grasping any extra layers of safety available to us without quibbling? And without delay.

Groundloop, have to agree that this particular incident was not your classic opposite direction conflict. In your suggestion to pass the slower aircraft on the left, I’m assuming that you are not a professional aviator (or sailor for that matter) – but yes, yours is a good idea. However, this incident does clearly illustrate the point 410 and others have been trying to make for some time now – the extraordinary accuracy of modern navigation systems. I think we’ll have two (and possibly four) converts to the offsetting argument in the crews involved in this incident. If flying NATS tracks and overtaking same direction traffic only 1000’ above or below you, (with the wonderful aid of hindsight), commonsense might now dictate that offsetting slightly to the right until past might be a very good practice, even in absolutely smooth conditions. ‘Situational awareness’ might also dictate that in such a situation, crews keep a very close eye on exactly where traffic immediately below them is tracking to pre-plan the direction of an immediate turn (or not to turn) in the event of a sudden depressurisation or engine failure in the cruise.

Maybe we should ask for a quick survey of respondents to this thread. How many of you do fly offset some or most of the time? I’ll add my name to list – I do it nearly all the time above 10,000’, except in RNP5 airspace. I’d rather find myself standing on the mat in front of the boss or the CAA explaining the error of my ways than finding myself a (dead) statistic. I don’t want to end up in some crash comic after meeting the same fate as the Luftwaffe TU154 and USAF C141 crews who had the midair off the African coast a year or two ago.

And if someone can explain the maths to me re how offsetting increases the chances of a midair, I’d be very grateful.

[This message has been edited by Wiley (edited 29 November 2000).]

[ATC Watcher]

Here we go again..
Some more thoughts :
1 NM offs et is fine in RNP 10 Airspace or higher. In RNP 5, 4 or in TMAs this is too much, but as in nearly all FMS you only can manually off set full digits, it is not possible to off set less than one full mile.
Only solution would be enbedded off set.

The more the Navigation accuracy of your aircraft, the least ammount of offset you need. In fact with GPS/IRS 0,1 NM will do.

To avoid the problem encountered in the begining of this thread, the notion of applying RANDOM offset between 0, 1 and 0,4 NM would elimaniate that problem.

Problem is when do you revert to accurate NAV . 10.000 Ft as suggested won't work as there are may airports above 10.000 Ft...
Most collisions occur in low altitudes anyway...in TMAs or close to airports, . this question at the moment is a choking point,

Last : Mathematically, someone in the ICAO Working group demonstrated that if you offset, on multiple crossing points, the area of exposure to collision is greater, therefore the risk of collision is higher.(But do not ask me to give you back the formula )For this reason , legally ICAO cannot implement a measure that will increase the risk.At least this is the reason they give us. Common semse might prevail one day, but like many things in aviation, it will need a few more collisions.

[willadvise]

Wiley
Yes I can confirm that. Remember I am talking about FANS/ADS/CPDLC aircraft. Of course if you don't have this then there is no way of knowing (and I would prefer not to know). In radar coverage you have 3nm either side of the waypoint to be considered passing over they waypoint.
For more infomation on this have a look at the following link

http://www.pprune.org/ubb/NonCGI/Forum3/HTML/000823.html

[410]

willadvise, thanks for the information you gave in response to Wiley's query. I think most of us can understand how the maths quite correctly gives a theoretical increase in the probability of collision with the wider crossing window that offsetting causes at airways intersections. I can also see that this wider window throws the current formulae you use to calculate clearances for crossing traffic out the window. What I’m trying to get across is that maybe someone at the top should use some lateral thinking and accept that maybe they need to accept these wider tolerances and re-calculate the formulae to allow for a (preferably embedded) offset. Crossing traffic doesn’t scare me nearly as much as opposite direction traffic on the same airway. To hit crossing traffic, you’ve got to be really unlucky in having both aircraft crossing the same three dimensional spot at exactly the same time. With opposite direction traffic, many aircraft occupy exactly the same line in azimuth for up to eight or more hours.

Even if the decision I’ve been asking for was made tomorrow by ICAO, it would still be years before an embedded offset was a part of even one, let alone every FMS in the world. This is why I believe we simply have to allow the manual offset to be made a legal option – and preferably, a requirement – for all aircrew, ASAP. I don’t like the 0.1NM argument put forward by some ATC people. This will not prevent a TCAS Resolution Alert in the event of a mistake - and recent events within my own airline have proven that when surprised pilots receive an unexpected (and in this case, genuine) RA, they sometimes react not exactly as per the TCAS instructions. I want to avoid pilots having to manoeuvre to avoid whenever possible in a TCAS encounter and 0.1NM offset won’t achieve this. Anyone who doesn’t believe me, get clearance from ATC for 1.0NM right of track the next time you have traffic approaching 2000’ above or below you and take a look how close that is. Imagine 0.1NM and you’ll see that it simply isn’t enough.

[willadvise]

410
I don't see any need to increase the tolerances for the calculations to account for a 1nm offset. In oceanic airspace in Aus. you are considered to have a cross track tolerance of 50nm(yes thats 50nm left and right), in RNP10 airspace its 25, and over land it is 14nm. With current navigation systems a 1nm offsett can easily be asborbed into these tolerances. I would like to see it reduced to 25nm over the ocean and 10nm over land.

[410]

willadvise, to clear up a (very small) point of apparent contention: I don't believe anyone in his right mind would suggest pilots should stick in an offset if being radar vectored. (Your original post >>" The controller is expecting you to fly a particular track and makes judgements based on your predicted track only to find you a mile or so closer to that other aircraft that you were going to slide nicely 5nm past."<< )

Is there anywhere except a terminal area where you allow an aircraft to get within 5NM of other traffic? Even in Europe, where I've had same level cruising traffic quite close, I can't remember seeing any only 5NM away. I don't think anyone is suggesting that we as pilots should build in our own bugger factor over a controller's vectors.

Every time I throw the "automatic embedded offset above 10,000 feet" argument forward the same argument is put forth: "There are airfields above 10,000'". Not too many, particularly if capable of taking jet aircraft, but this apparent problem can be overcome with existing technology. As soon as any current FMC-equipped aircraft enters a STAR (and until it exits a SID), the offsetting facility is not enabled. The same feature would fix the problem for any airfield above 10,000'.

Could someone give me a list of such airfields? Maybe in the Andes or Nepal? I've landed above 10,000' in New Guinea, but it wasn't at an airfield - and it's not an experience I'm in any hurry to repeat. Thin, thin air and a very dramatic demonstration of how TAS is the overriding factor over IAS that far above sea level...

[TinPusher]

I suggest that if you decided to offset a mile in any airspace, excepting on approach, it's unlikely anyone would notice, or if they did the assumption would be the poor pilot is having trouble navigating (not suggesting for 1 minute that pilots navigate poorly :))
In a/space with bi-directional routes it seems sensible to offset and keep your transponder and TCAS on!

Ratboy.....NZ runs a fair portion of the Pacific Ocean and Tasman Sea with 'OCS'- (oceanic control system), using the technology you mention. It appears to work well, the major advantage the system has over Ozzy's TAATS or any of the other current systems is the 'conflict probe' software. RVSM and Flexi tracks are becoming the norm, ie: Free flight (more or less) so it wont matter how far you off set because your traffic may well be tracking on a random route too!! So.....offset anyway, keep your transponder and TCAS on and keep a sharp lookout skipper http://www.pprune.org/ubb/NonCGI/eek.gif

[This message has been edited by TinPusher (edited 03 December 2000).]

[This message has been edited by TinPusher (edited 03 December 2000).]

[This message has been edited by TinPusher (edited 03 December 2000).]

[willadvise]

410
To clarify my original point.
: I don't believe anyone in his right mind would suggest pilots should stick in an offset if being radar vectored.
You don't neccesarily to have to be radar vectored for the controller to consider your track. Some examples. You have just departed via a SID. The SID is designed to keep you away from the corresponding STAR. You offset 1nm right. The guy opposite to you on the STAR offsets 1nm right and you find yourself 2nm closer to each other (depending on how the SID/STAR routes are designed).

You depart on a SID there is only one bit other bit off trafic you conflict with. The controller leaves you on the SID for a while (intending to give you direct once there is sufficent separation with the other aircraft) and sees that you are going to slide 5nm past the other aircraft, now you are passing through 10000ft and so you put your offset in. Oh sh!t not going to pass after all.

You depart on a SID there is no other traffic so it is cancelled and you are given direct to your first enroute waypoint. Do you offset?

Tin Pusher comments that he wouldn't notice 1nm offsets and may assume pilots are having trouble navigating. I disagree. I can see what nav equipment you are carrying in your flight plan. If you don't have GPS/IRS then I will cut you some more slack, but if you do a 1nm offset will be obvious. on radar.

You ask if less than 5nm is used anywhere else other than in the terminal area. Apart from visual separation I can give one other instance. This is going to be a little difficult to explain but I will give it my best shot.
http://www.geocities.com/willadvise2000/latsep.JPG
Aicraft 1 is tracking north bound. Acft 2 south bound. They are tracking via a VOR at A. There is 15degrees between their nominal tracks. According to lateral separation prinicipals these aircraft are considered to be separated at 15nm from VOR A. At which point the two aircraft will nominally be 3.9nm from each other. Now assuming you are both offsetting 1nm to the right you will now pass 1.9nm from each other. Do you still want to offset over land when I am going to be making decisions like this.

Like I have said before. I don't really have a problem with you offsetting over the ocean but please consider what I have said above before you do it in radar coverage or over land.

Cheers
WA

fixed image link.


[This message has been edited by willadvise (edited 05 December 2000).]

[410]

willadvise, I can well understand the concerns you express in your examples. However, they illustrate very clearly why line pilots, ATC and the regulators haven’t been able to get together and solve this problem to date. Put simply, it is ignorance of each other’s problems and concerns and a lack of will so far by those in a position to do something constructive towards addressing them.

The designers of the most commonly used Flight Management Systems foresaw your objection as you detailed in your examples. Offsetting isn’t possible in a STAR or a SID. The feature is not selectable until the SID had been completed and if an offset is still in place on descent, it is automatically deleted at the beginning of the STAR.

To repeat the main thrust of my argument: as line pilots, we don’t perceive a big problem in an ATC environment where a controller is actively vectoring aircraft on radar, particularly in a ‘first world’ environment. The arousal level of all concerned in these phases of flight is usually high enough to have at least one set of eyes sufficiently on the big picture to prevent glaring mistakes being made. (I have to say there are exceptions even to this, and ‘Murphy’ is always lurking in the shadows ready to catch even the sharpest controller or pilot out, as seems to have been illustrated by the recent very near miss between the loosely formating F15 with the Britannia 757 over the UK late last month. If initial explanations of this incident prove to be true, is there anyone out there who still places his total trust in TCAS to provide separation?)

It is the other ten hours of a long haul flight, (or the forty minutes of a short haul flight), where arousal levels are not quite so high, where co-ordination between ATC agencies may be poor or non-existent, (between semi-warring agencies for example, like Larnaca and Ankara), where we want to see something done. There have been more than a few incidents already attributable to the super accuracy of modern nav systems, some of them in so-called ‘first world’ countries, even the U.S.A. In Africa, offsetting’s been recognised as an easily-achievable and much-needed fix. Do we have to wait until hundreds more die in one more single catastrophe before we slam the proverbial stale door and adopt it elsewhere in the world? It’s not as though it hasn’t happened on a large scale already – it has, but so far, ‘out of the Western media’s eye’ in places like India or Africa. So far in the West, thankfully it’s only been commuter or charter aircraft involved, like the accident over the Grand Canyon in the U.S. If it ever happens to two widebodies full of Western tourists or businessmen, you can bet the media and the lawyers! will be screaming “Why didn’t you as an industry do something to prevent this when you knew the problem existed?”

We’re not in any way accusing you as Air Traffic Controllers of being sub-standard in the performance of your jobs. We’re just acknowledging that we are all human. It’s all too possible for a pilot to misread a clearance or acknowledge a climb or descent instruction for another aircraft with a similar callsign. It’s just as easy for a harried controller to miss or misread the information passed to him on an aircraft entering his airspace. It happened off the east coast of Africa a couple of years ago and the crew and passengers of two jet aircraft paid the ultimate price.

Murphy' Law says that if such an error does occur, the errant aircraft will almost certainly be in conflict with some other aircraft. Twenty years ago, such an error might have gone unnoticed except in some pilot’s or controller’s annual fitness report. Today, thanks to the extremely accurate nav systems most of us use, that same small error could have catastrophic consequences. Offsetting in the cruise phase of flight (when not under radar vectors) would put in place a ‘bugger factor’ that might help prevent one small mistake leading to a catastrophe.

If we could just talk, (and maybe more importantly, listen) to each other, as we’re doing on this thread, we might get around the misunderstandings that could give an outside observer the impression that we’re all too busy defending our individual pieces of turf to do what’s best for all of us.

Somebody on another thread mentioned the 'Titanic' principle of management - where each department is perfctly happy so long as its deck chairs are all neatly arranged. Meanwile the overall 'ship' is sliding gracefully under the waves. The same might be said of this argument.

Edited for typos.

[This message has been edited by 410 (edited 05 December 2000).]

[TinPusher]

Willadvise
Excellent graphics. I work radar TMA airspace where, if aircraft are anywhere near each other at the same FL (in your example both aircraft are within the area of conflict, I assume they are at different levels) they will be locked on headings and know about each other.
Procedural sep's are designed with a fudge factor, perhaps someone with a penchant for ICAO Doc's and the like could advise if 1nm off track within an airway, (assuming that nav is without reference to VOR or other ground based nav aids), is within tolerance or not??

I still maintain that on bi-directional routes (even if in radar coverage) it would be prudent to offset, if only 1/2nm, especially in RVSM airspace.

410
I would be happy to fly with you on the condition you get me to destination 'unbuggered' and if that includes offsetting then go for it!! :)

[Wiley]

Shame to see this thread slip to page 3. Looks like only a very few are interested. Pity.

[LoLevel]

In the enroute case where using the least accurate means of navigation combinations, and a procedural standard is achieved, isn't the basic lateral separation still only 1nm (worst case)?

Therefore if each ACFT throws in it's own offset, this lateral sep point must be looking pretty shaky! (case is assuming no ADS/RADAR)

[Deep Float]

Concerning RVSM over the atlantic: wouldn't it be much safer, and would capacity not be greatly increased if the NAT's would be separated by 30nm laterally (instead of the current 1 degree lattitude = 60nm) than by reducing the vertical separation to 1000 feet (1/6 nm)?
I have never understood this. But please correct me if I overlook something here.

[forget]

UK investigators worry about RVSM after Atlantic airprox
David Morrow, London (07Jun01, 14:56 GMT, 536 words)


UK accident investigators are urging regulators to review operating procedures in reduced vertical separation minima (RVSM) airspace after a clear air turbulence encounter caused a serious airprox incident over the North Atlantic last year.

Shortly after hitting the turbulence a New York-bound THY Turkish Airlines Airbus A340-300, cruising at FL360, suddenly pitched up and climbed rapidly to FL384. The uncommanded manoeuvre resulted in the A340 passing through FL370, an altitude assigned to a Canada 3000 Airbus A330-200 which had been above the THY jet and overtaking it slightly to one side at the time.

UK accident investigators probing the incident are warning that there is no formal procedure in place for avoiding the risk of collision in such circumstances. The accuracy of modern navigation systems, says the Air Accidents Investigation Branch report into the 2 October incident, is such that there is a greater chance that overtaking aircraft will be separated only vertically and, in RVSM airspace, by just 1,000ft.

It stresses that the speed of the A340’s climb – initiated by the aircraft’s automatic protection systems after it hit the turbulence – was so sudden that the A330 crew might not have been able to initiate evasive action. The report describes the event as a “serious loss of separation” and says the A340 pilot estimated the aircraft to be horizontally separated by 1nm (1.8km).

This has disturbing implications for the RVSM safety case, says the report: “If the intruder aircraft continues its climb there can be no guarantee that an aircraft directly above it could respond in sufficient time to avoid a collision.”

It adds: “It is not clear whether the European or Oceanic [RVSM] safety case studies and models have taken account of the risks of clear air turbulence coupled to the response of sophisticated flight-control systems such as those fitted to the Airbus fly-by-wire aircraft series.”

There is no formal procedure listed in the North Atlantic Minimum Navigation Performance Specifications Airspace operations manual for limiting the risks while overtaking at RVSM altitudes. The report states that lateral offset flying should be considered as a way of providing an effective safety barrier.

“The collision risk could have been reduced if the commander of the overtaking aircraft had been permitted (and expected) to temporarily increase the lateral separation between the two aircraft before they reached the line-abreast position,” it says.

It points out that simulations have indicated that a lateral separation of 1.5nm (2.8km) would be enough to prevent an aircraft’s traffic collision-avoidance system (TCAS) from generating a resolution advisory to the crew, and adds that flying a 2nm lateral offset is already an approved procedure used to mitigate wake turbulence problems.

“There would appear to be a safety case for extending this contingency procedure to overtaking, particularly in regions where turbulence of any kind is evident or forecast.”

Recommendations made in late November last year urged the CAA to request a review of overtaking procedures in RVSM airspace as well as suggest the development of a standardised lateral offset procedure. A spokesman for the CAA says: “We have put forward these recommendations to the appropriate international authorities. It’s really out of our hands now and we can’t say how or when the issues will be addressed.”


Source: Air Transport Intelligence news

[Flaps90!]

...and as elected representative for the 232 of us sitting directly behind you, we had a quick vote after the meal and have decided we would also like you to fly offset :) :)

(sorry to bring a layman pax into your fascinating, superbly written and highly informative thread.) :)

[Feather #3]

I don't think anybody would be thinking of offsets below at least FL100 and my personal suggestion, FL200.

However, the smart folks at ICAO don't sit up night after night watching opposite direction traffic passing PRECISELY head-on time after time after time after time after time after time.......you get the message.

All the talk about pilot tolerances, equipment tolerances, the need to recalculate safety heights for the offset route, the need to recalculate latsep points fot ATC sounds just great. It ignores one thing, GPS equiped a/c today ARE EXACTLY ON TRACK!!! And they WILL hit head-on if someone makes a mistake.

The boffins, to use that American expression, need to get out more. So do the ICAO and IFALPA folks who sit around pontificating while the rest of the world wonders.

For mine, an official 1nm offset above FLxxx can't come soon enough!

G'day

[411A]

DeepFloat--
Must agree with your reasoning. Altho I do not do all that many NAT flights, have always thought that reduced lateral spacing is far superior to RVSM. Was it ever considered?

[cossack]

As a non-oceanic controller it does seem strange to me that separation in the vertical can be reduced by 50% to 1000 feet, but the lateral still remains at 60 miles or 10 minutes in trail which is about 80 miles.

I know its a procedural environment and commnication is difficult but surely modern aircraft are more likely to deviate vertically than laterally.

Have any studies been carried out on the advantages of reducing the lateral (track) separation by 50% to 30 miles (an awful long way even allowing for offsets)?
You could use the RVSM produced levels (300, 320, 340 etc) on the intermediate tracks so that vertical separation on the same track became 2000 feet again.

Just a thought...

[Captain Windsock]

offsetting is great as long as you are the only aircraft doing it.

[divingduck]

I been reading this thread quite closely, good instructive and informative comments from both sides of the fence so to speak.

With regard to whether we (the ATC's) can see if you are off track, here in the Middle East we certainly can. The new kit we have is pretty good, we can see to within 0.1nm if we feel like it.
Having said that, I do see many aircraft sitting comfortably 1-2nm off track on a regular basis. I personally have no problem with it unless I'm RADAR vectoring, then it would be impossible for the pilot to offset anyway wouldn't it?
Having heard of many opposite direction near misses (and same direction come to think of it) coming out of the airspace to the east of us...I am more than happy to see the offset happening.

BTW willadvise...I haven't been in Oz for a few years now...are you guys now using the ICAO 15 degrees 15 miles now? I have always felt that 3.75nm is not really enough...if we are using radar, we need 5 (usually)so I can't see that 3.75 is safer.

BTW someone asked about lateral separation...it's 1nm between the POSSIBLE positions of two aircraft (that takes into account the xcross track errors and adds one for mum).

Keep the good gen coming, I feel it is of benefit to both sides especially with the possible implementation of RVSM in most places of the world.



------------------
turn the plane! turn the plane!

[410]

Capt Windsock, I have to disagree with your (perhaps flippant?) comment, (“ offsetting is great as long as you are the only aircraft doing it. ” posted 16 June 2001 15:20 )

The fact is, the safety margin doubles if all opposite direction traffic offsets one mile right; ie, if the navigation equipment is accurate (and the indisputable fact is, it is - very, very accurate), the separation between opposite direction traffic becomes 2 NM, because each aircraft is 1NM right of the airway centreline.


[This message has been edited by 410 (edited 18 June 2001).]

[Bono Vox]

A cynical hypothesis on the "30- mile lateral" vs 1,000' RVSM- (I believe) there are ususally 5 NAT tracks, identified A-E so if you reduce separation laterally to 1/2 a degree of latitude you would add an extra 4 tracks. By introducing RVSM between FLs 290 & 430 you add 7 FLs, or 7 new "tracks" (or have I got the vertical dimensions for RVSM embarrassingly wrong?).

[RATBOY]

In the airprox incident that prompted this whole thread it was stated that clear air turbulance caused the aircraft being overtaken to climb through the RVSM levels. It also appears that very smart Eurocontrol people are talking about domestic RVSM. What will happen even at FL300+ when clear air turbulance or predictable mountain waves etc cause these kinds of conditions in greatly more congensted airspace than over the middle of the Atlantic or Pacific? Seems that offsets or some other method of allowing for this type of sudden climb or descent would be a necessary condition to provide a safety margin.

[Farside]

We started this offset topic more than two years ago on this forum and on Avweb. The articles posted were from us here in Singapore and some friends in Dubai. Our topic was posted under the name" FLY RIGHT" and I wonder if it is still on the server. I am happy to see that it is a current topic again and perhaps we can get some things moving on this subject.

[bookworm]

I don't have an axe to grind on RVSM -- I wish I flew something that got up to those levels. But I'm somewhat surprised to see this incident used to support the anti-RVSM case.

The incident revolves around the unexpected climb of the A340 after a turbulence encounter while the vertical separation between the aircraft was 1000 ft.

It strikes me that:

1) The turbulence encounter could easily have occured at the lower levels where the standard separation is and has always been 1000 ft.

2) In this case the A340 climbed 2400 ft before descending again. If no lateral separation had existed and the vertical separation had been 2000 ft, the aircraft would still have collided.

I was particularly surprised by the following comment.

<font face="Verdana, Arial, Helvetica" size="2">Such was the vigour of the A340's climb in AoA law, the aircraft could well have climbed through FL 363 (thus provoking a TCAS RA with revised software version 7.0) in a very short time, even if the crew had applied nose-down sidestick as soon as they heard the (delayed) autopilot disconnect warning. The climb to FL 363 would have been sufficient to generate a TCAS RA in any adjacent aircraft at FL 370 but, if the intruder aircraft continues its climb, there can be no guarantee that an aircraft directly above it could respond in sufficient time to avoid a collision. Therefore, the RVSM safety case should not be driven by any assumption that a different crew might have contained the situation by making an earlier nose-down sidestick command than the A340 crew involved in this incident.</font>

There didn't seem to be a great deal of evidence that an aircraft at FL380 would have been able to react in time either.

By contrast, the incident seems to demonstrate that lateral separation of tracks under all circumstances is a sensible precuation.

Or am I missing the point of the anti-RVSM case?

[411A]

Bookworm--
Would totally agree. However, the powers that be have decided that RVSM is here to stay, inspite of any problems. Afraid it will take "bent metal" before anything is done, due to a very big loss of face syndrome otherwise. Lateral spacing IMHO is much better.
Was there/is there a problem with Airbus aircraft that would indicate this is unique to the A330/340?

[bookworm]

<font face="Verdana, Arial, Helvetica" size="2">Was there/is there a problem with Airbus aircraft that would indicate this is unique to the A330/340?</font>

You can debate that one till the cows come home. The A340 went into "AoA protection law" because the AoA reached the 'alpha prot' threshold value in the turbulence. Thus it then held the 'alpha prot' value of AoA until a nose-down sidestick input was made some 28 seconds after the incident began.

It could be argued that the crew should have made a nose-down sidestick input much earlier, but that's easy to say with the benefit of hindsight. It could perhaps be argued that the "AoA protection law" is of limited value in the cruise at FL360.

[Capt PPRuNe]

Sorry, got to close this thread until 'forget' fixes their email as all the notification of replies are bouncing to me.

As and when I recieve notification that the email address has been fixed I will re-open the thread.

------------------
Capt PPRuNe
aka Danny Fyne
The Professional Pilots RUmour NEtwork

[Capt PPRuNe]

Thread reinstated.

------------------
Capt PPRuNe
aka Danny Fyne
The Professional Pilots RUmour NEtwork

[Wiley]

It's a real pity to see such an important topic slide off the top page.

I hope the day never arives when some senior ICAO official or airline manager finds himself facing some high-powered lawyer who asks him why ICAO or the airline had never though this issue important enough to acton, given that at least one fatal accident involving large passenger jets, the USAF/GAF midair off the west coast of Africa some years ago, might have been avoided if offset tracking was mandatory.

Someone has said it before in earlier threads - the sad fact is, fare-paying Western passengers, probably lots of them, will have to die before anything is done.

[mallard]

Just a few remarks after several years of oceanic and other flying.
Oddly enough, the only other case I have recent knowledge of was an own company A330, over the Atlantic, which had an altitude excursion following,"the worst turbulence the captain had ever experienced".
Oceanic traffic goes in swarms in the same direction with only occasional opposite traffic. If everyone offset it would achieve nothing.
In high density Europe under RVSM it would be reassuring to see the oppos slide down the other side. You have to be there with a joint speed of 1,000 mph to appreciate how little time you have to react to a potential conflict.
With regard to reducing lateral separation on the Atlantic by 50%, consider this: communication is still by antiquated HF via radio operators, not ATC controllers.
On a bad day you may take 20 minutes to make contact with the guy and possibly never get a response to a request to deviate off track due to severe weather.
By then you have probably had to take the initiative to take whatever avoiding action is required.
With reduced lateral separation it would just make that, officially, unauthorised deviation more risky.
You don't have to spend much time over the Ocean to reallise how little actual control there is in the continental sense.
I believe the general public would be amazed.

[Farside]

How many more “Close Calls”

The attached file is a copy of an article that appeared in last week’s USA Today. It describes another near miss over Indian Airspace and casually mentions in it’s closing paragraph that there had been some 20 near collisions in Indian skies since November 1996.
Now it doesn’t need an Einstein to realize that one of these days it won’t be a near miss but a full hit.
With “Freeflight” and the implementation of full Fans still a decade away, are we just going to sit and wait for a disaster to happen. The Airway system and communication system over India (and many more places in the world for that matter) are completely outdated and not capable of handling today’s traffic volume. A friend of mine, who is a leading design engineer in navigation equipment, told me that he sometimes had sleepless nights realizing that the equipment designed today was so accurate that, used in the present ATC environment, would one day contribute to a midair.
We cannot sit idle and wait for some beancounting bureaucrat in organizations like ICAO and JAA to change and adapt ATC procedures to create a safer environment, and avoid such disasters from happening. We all know that changes will not come soon, and only after several disasters will somebody wake up and actually start to do something.
I believe, and have argued this before, that we should start a discussion to come up with some ideas and/or procedures as to how to implement the legal and authorized changes in the present system. With the many aviation forums that we have today,let us use these forums to encourage positive discussions, resulting in policy changes, and increased safety.
For the time being, Offset Tracking is an individual choice, to increase separation, and is being used by more an more pilots (as per several articles in Flight International). But as long as these procedures remain someone’s individual technique, it might and will work for that crewmember, but is not used to it’s full effect as it is not an official procedure.
Offset Tracking, Parallel-Oneway airways, direct INS and GPS routes are all procedures that can be implemented fairly easily and without great financial inputs. Why is it not happening, and what are the political forces stopping it from happening?
Where do we start, and what is the procedure to follow?
The irony of the situation is that, while we are flying a multi-million dollar piece of equipment into the 21st century, screaming at the top of our voices over some outdated piece of HF equipment, trying with 100 other flyers to get our position known to an Indian ATC controller, you can step back into our luxurious passenger cabins, where every passenger today can swipe his creditcard through his individual armrest satellite phone and have immediate 5/5 duplex phone connection with anybody in the world.
I wonder if we have got our priorities right.

[Wiley]

People - a lot of people - have already died when opposite direction traffic has collided using the same airway. (See the Pprune Home page for details from 1996.) This was one of Pprune's earliest 'crusades', and unfortunately, little or nothing has been achieved since then except a few more line pilots have been made aware of the problem.

Let's be clear about one thing: more people are going to die in midairs unless we as an industry do something about fixing this all too easily fixed problem. It's not good enough for isolated individuals to fly offset. Everyone with the super accurate GPS/INS should do it, and the only way that will ever happen is if the equipment does it automatically.

Every time in the past this has been raised, someone chimes in 'proving' it won't work because of one thing or another. It's all too easy to be dragged off on a tangent with arguments like this. The RVSM camp see their problem, the NATS track people theirs, ATC people see the call for offsetting as an affront or criticism of their professionalism, the mathematicians can 'prove' it makes the risk of collision more likely.

Let's agree that the problem offsetting is attempting to address applies in no way, shape or form with NATS tracks, which have problems unique to the busy (and usually one way at different times of day) traffic patterns North Atlantic. I'm mainly concerned with the non-radar environment outside Western Europe and the US. And I'm not slagging off the controllers in these other parts of the world. It's just that it's a fact of life they are forced to work in a far from ideal environment with sometimes poor or outdated equipment, bad (and sometimes no) comms with neighbouring agencies because of political conflicts. Anyone with any imagination can see the potential for serious traffic conflict in such environments because of the high accuracy of modern navigation systems. The only long term answer is unidirectional airways, but I wouldn't hold my breath waiting for them.

What amazes me is the concerted resistance that people in authority (and therefore not flying the line daily) are maintaining to addressing this navigation accuracy problem. It will probably take at the very least some embarrassing disclosures by some 'muck raking' journalist to shake these agencies into doing something. (Thanks to sites like this, they won't be able to say they or the industry as a whole were unaware the problem existed.) Unfortunately, it's unlikely any such journalist will see any news value in any such story until a lot of people have died in circumstances that cannot be disguised as being attributable to anything but the accuracy of the nav systems. Let's hope that's a long time (if never) coming, but I fear that's a faint hope.

[Farside]

I completely and fully agree with all you said Wiley,but where are we going from here. I am open for all suggestions and willing to put some time in it. It is a worthwhile course and something that should be addressed.

[410]

As a start, talk to your Chief Pilot / Fleet Manager and ask for it to be tabled at the next Standards meeting for discussion, Farside.

It probably won't get up on the first try, but keep reminding those above how important you think it is.

[Ignition Override]

Blacksheep: I don't know about the CAA over there, but our head people at the DOT and FAA etc are always White House appointees, and usually consist of lawyers and/or career politicians. One guy at the FAA had been a Thunderbird pilot but he was just another personal ticket-puncher.

What the FAA might not be taking into account when they perform their 'cost/benefit' analyses is the ever-growing trend to outsource any and all major aircraft maintenance.

JetBlue sends many, if not most of their A-320s each year to El Salvador for C or D checks. Even in the US, also according to the same article in either "the New York Times" or "the Wall Street Journal", as a minimum, only the maintenance supervisor is required to have an FAA license! Even before our oldest fleet was equipped with the RVSM alimeters etc, many of our planes returned from a place (not too far north of the very white sand) and were grounded for a while. They almost lost their license to do the work, until a company Maint. Crew Chief was based down there (this he told us upon arrival from ATL).

Is any of this already the case with European airlines? About the only US major to perform most of the technical work in-house is American Airlines, and they have never yet filed for Chapter 11. Well-done American!

[barit1]

willadvise is taking his Air Traffic Control paradigm too literally.

There are things that ATC cannot control - turbulence, Comm and/or Xpdr loss, a host of other systems & environment anomolies. When this happens ATC has only minutes to provide separation info to other flights, and when this secondary level of protection is less than perfect, a tertiary level (ie offset) can save the day.