What does separation “assurance” mean?
 

I’m told by Aussie air traffic controllers that one of the major differences between our country and other countries overseas with modern aviation systems is that in Australia, the controllers not only have to separate aircraft but they also have to provide “separation assurance”.

Are there any air traffic controllers who can explain on this thread what Airservices actually means by that? Is it one of the reasons you hear the term “clearance not available”, which doesn’t appear to be in the vocabulary of air traffic controllers overseas?

From the horses mouth:

That’s an amazing statement! Is it ICAO based?
Seems to me a great way of entrapment!

It's what you do every day in Class E, Dick, apply your own rules to assure separation between all the other aircraft you know about but without applying a separation standard. You do separate yourself from other aircraft in E, don't you? For my and my punter's sakes, I sincerely hope so.

In it’s original manifestation it had two broad meanings:
1. Construction of airspace and routes such that the requirement for positive separation actions is minimised.
2. controller technique, whereby, in simple terms, traffic was not placed such that, if the controller was distracted or overloaded, a breach of separation would occur. For example, if you were radar vectoring two aircraft on reciprocal headings then you would apply level stops or slightly different headings, in case you dropped dead.
That was the simple concept, don’t know what happens these days.
LOSA was for cases where an occurrence was less than desirable, but no standards were breached, originally a form of statistical gathering to highlight areas of concern.

Making separation happen rather than letting it happen. As 40years says it's effectively another name for good controller technique. If I sit back & watch crossing traffic miss by 7 miles then I haven't had a breakdown but if I misjudged it by 15 seconds or one aircraft changed speed or jinked left by a mile or two suddenly I have less than 5 miles. So I assign different levels or use headings to make sure I keep 5 miles.

They will definitely miss rather than I think they'll miss, without me having to watch like a hawk because I have another 20 aircraft to look after too.

We still use it.

Though I've heard the term "technical loss of separation" for it as well, which covers that you did not assure your separation.

One example is the climb from ground all the way to FL400+ when someone is opposite in FL390.... the assurance is let him climb below, and look at it when he is closing getting closer to the desired level. The technical is giving him all the way, and then look at it to see if we have to do something later... hence falling dead or forgetting it can be very upsetting.

heh, thanks for that quote - now, can somebody pass that onto the trainers at the college? None of them are able to state what it is (and remember the phrase about if you can't explain something simply enough.....) (and they'd probably call that line from your quote LoS, not LoSA)

Dirty, nasty, vicious, deceitful, unclean, dangerous, venomous, pox laden words from the fundament of some lawyers rotting entrails.

It means the controller f****** up. However Airservices wants to discriminate as to whether the effup was the result if a controller failure or an organisational failure, With the unstated implication that their standard is perfect and universally applicable , hence any error is not their fault.

"Clear application" WTF does THAT mean??????????????????

Sunfish, did you not read the above explanations? If there is a LOSA, the controller may or may not have stuffed up. In the example given above, an unrestricted climb to a level above opposite direction traffic may have one of two outcomes; either the aircraft will pass safely with the required separation, or they pass (or hit) with less than the separation that ICAO tolerances require. The separation is assured if action is taken as described, eg a level stop (set and forget) to be assessed later for further action. It is controller technique, not organisational arse-guarding.
A secondary aim is collecting reports of incidents where LOSA repeatedly occurs, which may then lead to review of airspace, tracks, rostering, whatever.

It means if you chose a career in Aviation you are assured of separation at some time! -) -)

If you're both VFR they'll just give you a traffic alert and "job done" - separation assurance handed over to the pilot.

There's the VMC climb/descend for IFR flights.

Quite standard in Europe and like as mentioned, is just good controller technique.
We have occurences which are noted as 'technical' loss of seperation. No actual loss occurred, but seperation was no longer absolutely assured.

For example. Aircraft A Climbing through FL100, requesting FL390. In it's 12 o clock at 15 miles reciprocal is Aircraft B at FL380. The minimum lateral distance between the two is 3.6NM at closest point of crossing.
To clear Aircraft A to FL390 through Aircraft B, even though there is no hope in hell of aircraft A climbing 28,000ft in 7.5 miles, is a 'technical' loss of seperation or a seperation not assured event. We just don't do that.

I personally use the thought process that "if my screen goes dark right now, is everything safe?" and apply seperation techniques based on that ideology. It leads to safe controlling, without affecting efficiency. Sure we still do tight 6 mile crosses, rate of descent restriction descents etc, but we have a very good idea of the scenario at hand and have always assessed it as such before hand. Every 6 mile cross is not the same...sometimes we know an aircraft is about to hit a hell of a headwind, so to ensure seperation, we will turn the aircraft to get 8 or 10 miles before any lateral seperation is degraded by the headwind. It's simple but seperation assured just means make sure what you're doing is positive and effective, 100% of the time. No winging it.

It can be a little bit grey.. as in what’s ‘assured’ to you may be different to what’s ‘assured’ to me.
In a practical sense it usually comes down to an error/misjudgement and if something/someone saves your bacon and prevents it resulting in a loss of separation then it’ll usually go down as a loss of sep assurance.

Example:
2 aircraft converging at the same level, conflict alert activates and the controller turns one quickly enough that they don’t lose a radar standard then that would be deemed a loss of sep assurance; or, 2 aircraft crossing, one climbed through the level of the other because you don’t recognise the conflict - supervisor sees the whole thing including the climbing aircraft reaching 1000ft above with 5 seconds to spare but controller didn’t even realise.. that’s a LOSA.

Probably best summed up as ‘achieving separation not by fluke or by having your ass saved by the system tools and/or another controllers situational awareness’

It's interesting that all of the examples given are to do with radar displays. If I may I will give a couple of other issues from back in my day:
1. The ten minute longitudinal standard - the controller could not know it had been infringed until the following aircraft reported it's position only 9 minutes behind, the controller then took action to provide another standard - however to provide SA then 10 minutes between aircraft should never have been applied - work that one out!
2. Aircraft departing SYD 16R and turning right on the departures frequency are supposed to outclimb traffic inbound from the south on the approach frequency. The departures controller has to tactically monitor the departure climb rates - there is no design separation assurance, or controller separation assurance
3. Virtually all Tower separation. It depends on constant visual assesment of taxi and flight paths with built-in escape manouvres in the controller's head. E.G Final approach with an aircraft cleared for take-off ahead, the standard is that the departure must be airborne and at least 1800 metres down-runway before the lander crosses the runway threshold. Each one is a punt for the Tower controller based on experience, knowledge of aircraft preformance, prevailing weather conditions, flight paths after go-around. There is no "separation assurance"; if there was then Towers would not need windows!

I always viewed it as a training technique that had suddenly become a separation standard

Airservices introduced separation assurance to cater for the lowest common denominator. They were passing trainees that should never have passed and as a result the number of separation incidents had gone up.
When it came to Approach control, it was the biggest crock of ****, due to the relatively small amount of airspace you are dealing with. The end result was delays, unnecessary vectoring and unhappy pilots. We had approach controllers that had no idea of aircraft performance (climb/descent) and the result was they were vectoring aircraft for 20+ extra track miles.
Safe and expeditious used to be our mantra - expeditious went out the window with the introduction of sep assurance.

I know of a few controllers and trainees that lost their ATC careers due to 'sep assurance.'

Yup. It was a nebulous concept. Yet another unique Australianism that has contributed nothing to aviation, apart from added cost of course.

It's not unique - maybe the name is but the concept of good controller technique isn't