OK thanks, I knew it was something I did.  I'll monkey around with that knob - I just couldn't figure it out.

And you have to set those minimums on the Capt side for the callouts to work (unless you have turned Capt side displays off!)

haha - clearly you've never been to Phoenix :wink:

 

Hey! Easy there...  :biggrin:

Hey! Easy there...  :biggrin:

 

haha - I wasn't referring to the GYR group of foreigners...more DVT... :wink:

haha - I wasn't referring to the GYR group of foreigners...more DVT... :wink:

No worries, just messing with you  :wink:

 

Back to the topic: If you listen to european ATC (Zurich [LSZH] for instance] what you'll hear 99,9% of the time will be something like: 

 

"Zurich Tower, guten Tag, Swiss 39 established ILS 14 8 miles"

"Swiss 39, hallo, continue approach"

....

 

At least in europe ATC will always assign you an ILS approach and most of the pilots will prefer an ILS or GPS approach over any other - even in best VFR conditions. However that doesn't mean that you have to autoland - feel free to disconnect the AP as soon as you're cleared and stabilised  and have visual of the runway. 

 

 

No worries, just messing with you

 

Figured :wink:

 

 

 

At least in europe ATC will always assign you an ILS approach and most of the pilots will prefer an ILS or GPS approach over any other - even in best VFR conditions.

 

My main criticism of the practice is more that it would require higher separation requirements, therefore incurring delays.  That, of course, is all based on our rules, though, where we can drop the standard (3nm) separation for visual (subject to pilot judgment, really) separation and pack planes in.  Additionally, that also only really matters when you have the traffic levels that some of our airports do.  Finally, as mentioned earlier, I'm sure CFMU planning assumes a constant path from runway to runway, and visual approaches could throw off the assumed plan versus the actuality, so using the ILS makes sense to ensure those match up.

Figured :wink:

 

 

 

 

My main criticism of the practice is more that it would require higher separation requirements, therefore incurring delays.  That, of course, is all based on our rules, though, where we can drop the standard (3nm) separation for visual (subject to pilot judgment, really) separation and pack planes in.  Additionally, that also only really matters when you have the traffic levels that some of our airports do.  Finally, as mentioned earlier, I'm sure CFMU planning assumes a constant path from runway to runway, and visual approaches could throw off the assumed plan versus the actuality, so using the ILS makes sense to ensure those match up.

 

I agree, the airports over here are by far not as busy as some in the US. (See LAX SFO JFK etc.) 

 

 

My main criticism of the practice is more that it would require higher separation requirements, therefore incurring delays.

 

It does not incure delays - the system is rigged with it in mind in the first place. Consequently when weather drops somewhere between CAT I and typical VOR minima, (ie. not LVP yet), it is more or less "business as usual", while in the US, system would just be overloaded (inasmuch I can tell).

 

 

It does not incure delays - the system is rigged with it in mind in the first place.

 

In the FAA world, as soon as you start handing out block times as the regulatory agency, that's a delay.  No delay would mean the operator got to leave at its chosen time, unrestricted by ATCSCC/ARTCC/TRACON TMU (CFMU equivalent).

 

 

 

Consequently when weather drops somewhere between CAT I and typical VOR minima, (ie. not LVP yet), it is more or less "business as usual", while in the US, system would just be overloaded (inasmuch I can tell).

 

This occasionally happens, but the ATCSCC/ARTCCs/TRACONs are generally in tight communication, particularly when lower ceilings or poor weather are expected and overloads are prevented.  There are hourly calls during normal periods to address changes to the daily weather call (early morning) evaluating the CCFP/AWWD and operational plan.  The prevailing thought over here is to not handicap traffic flow the majority of the time for situations that happen as part of the minority.  Sure, it's an easier transition if you meter at all times, but that's why we still have controllers and give them the latitude we do: they're much better at putting planes in the right spot at the right time than computers are without unnecessary restrictions.

 

I get the CFMU concept.  It makes a ton of sense, but it comes with consequences.  I'm not trying to knock it, because we use something very similar when necessary, but it is only used when constraints are expected.

In the FAA world, as soon as you start handing out block times as the regulatory agency, that's a delay.  No delay would mean the operator got to leave at its chosen time, unrestricted by ATCSCC/ARTCC/TRACON TMU (CFMU equivalent).

 

well - yes. And you don't need to do that, because the arrival rate stays the same, so I don't get your point. As for CFMU delays, you will hardly ever get a slot assigned when you plan to leave on time - unless there is a good enroute reason, such as the French striking again. Airport metering would get used sometimes, but that would have to be someting like runway closure, snow issue or LVP ops, not your usual low overcast day.

 

This occasionally happens, but the ATCSCC/ARTCCs/TRACONs are generally in tight communication, particularly when lower ceilings or poor weather are expected and overloads are prevented.

 

I wouldn't say that overloads (as in more planes wish to get somewhere then can) are prevented - only shifted. They may divert, they might get ground delays - but the lost slots have to end up somewhere don't they?

 

The prevailing thought over here is to not handicap traffic flow the majority of the time for situations that happen as part of the minority.

 

Not sure about the US, but in Europe it is common to be in ILS weather for days, if not weeks at time, especially in winter. Not sure that qualifies as a small enough minority not to matter.

 

 

Sure, it's an easier transition if you meter at all times, but that's why we still have controllers and give them the latitude we do: they're much better at putting planes in the right spot at the right time than computers are without unnecessary restrictions.

 

 

Your point being? Sorry, I just don't follow.

 

 

well - yes. And you don't need to do that, because the arrival rate stays the same, so I don't get your point. As for CFMU delays, you will hardly ever get a slot assigned when you plan to leave on time - unless there is a good enroute reason, such as the French striking again. Airport metering would get used sometimes, but that would have to be someting like runway closure, snow issue or LVP ops, not your usual low overcast day.

 

I was clarifying my point where I stated CFMU increases delay over an unmetered system.  Our system is not metered unless we need it to be.  Our high AARs are partially a result of being able to take advantage of visual separation (something that the UK at the very least cannot do, as I've come to find, so that point is somewhat moot for that part of the world).

 

I'm not quite sure we're on the same page here as far as CFMU goes.  My understanding is the CFMU is always used (I know you have to ensure your FP is bounced off of CFMU at all times, thus the assumption).  I'm guessing CFMU then validates the FP against all other known plans and ensures it will not be in conflict with other flights, at least per known FPs.  If there are too many planned flights through a certain constrained area (most likely an airport), you get assigned a delay to fit into the larger picture.

 

Am I incorrect?

 

 

 

I wouldn't say that overloads (as in more planes wish to get somewhere then can) are prevented - only shifted. They may divert, they might get ground delays - but the lost slots have to end up somewhere don't they?

 

I don't have empirical data, but most delay is taken through speed reductions and delay vectors.  Diversions are pretty rare.  Ground delays aren't within the scope of the discussion because ground delays would be metering, and we're discussing situations where unmetered flows could potentially cause "overloads" downline when the weather drops.

 

Lost slots do have to be made up somewhere, but again, these are normally accounted for via delay vectors and speed assignments.

 

 

 

Not sure about the US, but in Europe it is common to be in ILS weather for days, if not weeks at time, especially in winter. Not sure that qualifies as a small enough minority not to matter.

 

I never said it didn't matter.  I specifically said that we operate by default in an unmetered state and only meter when it becomes necessary (and only in the area in which it is necessary, moreover).  Since visual days far outnumber ILS days for the majority of the country, metering based on ILS use would be unecessarily restrictive.

 

Even if you're ILS for weeks at a time, that doesn't mean metering should be the status quo on severe clear days.  That's what I'm questioning here.

 

 

 

Your point being? Sorry, I just don't follow.

 

It was your point to begin with:

"Consequently when weather drops somewhere between CAT I and typical VOR minima, (ie. not LVP yet), it is more or less "business as usual", while in the US, system would just be overloaded (inasmuch I can tell)."

 

I was simply responding to that by saying that CFMU sets controllers up for success even if the weather gets poor by metering to the standard of the ILS.  We choose to let the controllers set themselves up for success across varying conditions for better NAS efficiency.  The controllers are better at adjusting to changes in the NAS than a computer would be.

 

Things don't usually take us by surprise in the traffic management realm, though, so when we anticipate needing metering, we'll use it.

 

Again, my only point is that we don't meter on days (or more generally "periods of time") when we don't need to. There's no sense in constraining traffic when the entire country has high pressure parked on top of it and there isn't a cloud in sight.  Planning based on assigning the ILS, assigning the ILS, or even utilizing it at all is simply sacrificing capacity that could be taken advantage of.  Again, that may just be so much of a concern here because of the amount of traffic that would be affected.

I am not sure we are talking about the same point.

 

 

 

I'm not quite sure we're on the same page here as far as CFMU goes.  My understanding is the CFMU is always used (I know you have to ensure your FP is bounced off of CFMU at all times, thus the assumption).  I'm guessing CFMU then validates the FP against all other known plans and ensures it will not be in conflict with other flights, at least per known FPs.  If there are too many planned flights through a certain constrained area (most likely an airport), you get assigned a delay to fit into the larger picture.

 

Yes - but usually the constrained area will not be an airport, rather enroute. Airports not being constrained is kind of my point, in fact.

As for usage, yes, you would have to fill FP to CFMU system, but usually you would not see a delay if you are departing within planned time. Delays are more usual for late flights, that just don't fit into the enroute picture anymore. You could re-plan though I guess. Or you can always send a ready message. (Oh, btw. I had no idea when I wrote the comment about the French striking... but there is an active strike warning for the next week)

 

 

 

I don't have empirical data, but most delay is taken through speed reductions and delay vectors.  Diversions are pretty rare.  Ground delays aren't within the scope of the discussion because ground delays would be metering, and we're discussing situations where unmetered flows could potentially cause "overloads" downline when the weather drops.

 

See, I think this is where we are not on the same page - of course you are metering. Because if you didn't - you would have too much planes trying to get down somewhere there isn't enough space. That is my point - that would not happen in Europe until you go down past 550m RVR or so.

 

 

 

I never said it didn't matter.  I specifically said that we operate by default in an unmetered state and only meter when it becomes necessary (and only in the area in which it is necessary, moreover).  Since visual days far outnumber ILS days for the majority of the country, metering based on ILS use would be unecessarily restrictive.
 
Even if you're ILS for weeks at a time, that doesn't mean metering should be the status quo on severe clear days.  That's what I'm questioning here.

 

It's not metering as you seem to understand that is the thing I am talking about. Flights are scheduled so that you don't need reduced separation in the first place. Then you won't run into problems when you can't get reduced separation.

 

 

 

It was your point to begin with:
"Consequently when weather drops somewhere between CAT I and typical VOR minima, (ie. not LVP yet), it is more or less "business as usual", while in the US, system would just be overloaded (inasmuch I can tell)."

 

Well - yes. But:

 

 

 

I was simply responding to that by saying that CFMU sets controllers up for success even if the weather gets poor by metering to the standard of the ILS.  We choose to let the controllers set themselves up for success across varying conditions for better NAS efficiency.  The controllers are better at adjusting to changes in the NAS than a computer would be.

 

I still have no idea how you come to this conclusion.

 

 

 

Things don't usually take us by surprise in the traffic management realm, though, so when we anticipate needing metering, we'll use it.

 

That is the whole point - we don't need to use metering untill the weather gets really bad.

 

 

 

Again, my only point is that we don't meter on days (or more generally "periods of time") when we don't need to. There's no sense in constraining traffic when the entire country has high pressure parked on top of it and there isn't a cloud in sight.  Planning based on assigning the ILS, assigning the ILS, or even utilizing it at all is simply sacrificing capacity that could be taken advantage of.  Again, that may just be so much of a concern here because of the amount of traffic that would be affected.

 

Well - of course we don't "meter" in the sense you seem to understand. We just don't schedule more than we can take on ILS. Do you schedule more than you can take, even considering reduced separation? 

As I see it, we are not sacrificing capacity - you are just conjuring some where it might be, or might not. As I see it, all you do is set yourself up for all those gridlock situations and multiple hour delays (from the passenger view of "delay", not from the "it's not delay, it's metering" view) as soon as weather gets a bit down.

It's impossible to do a like for like view between the US and Europe when it comes to delay management.  The US is one big agency with a plethora of airspace, Europe is made of tens of ANSPs all with minute airspace in comparision.

 

I guess having the worlds busiest single and duel runway airports we must be doing something bordering on efficient .

Good info here - thanks

I am sure both systems have their positive and negative sides.

 

But I like to think that since JFK has adobted the European way, as in metering/slots for departure, things are much much better there!

 

A few years ago it was normal to be number 50 for take off.

During your 1 to 1/12 hour taxi, all those aircraft were burningw/asting unbelievable amounts of fuel. Many had to return for refueling!

 

Since two years or so they meter the departure flow. (you are in sequence, number 3 for pushback).

They just let you sit at the ramp (engines off) for 15-30 minutes untill there is room for you to get to the runway.

 

Only when Thunderstorms in the area close pieces of airspace (or the airport) can you still have the taxi traffic jam problem of the past.

 

Much better!

Well - of course we don't "meter" in the sense you seem to understand. We just don't schedule more than we can take on ILS. Do you schedule more than you can take, even considering reduced separation?

Seems like your definition of meter is rather...interesting...and I'd argue almost contradictory.

 

So, you don't "meter" (that is to say, restrict flights to a certain amount per hour), but you don't schedule more than you can take on the ILS?

 

In the FAA sense, any time you schedule (which is going to be based on a rate) that's metering. A rate has been set for the ILS and flights are scheduled to that rate: metering. I can't find any direct ICAO language on a quick search, though EUROCONTROL uses it in the same manner, even going so far as to say "CFMU slots" (which is, again, a metering term): http://www.eurocontrol.int/eec/public/standard_page/EEC_News_2008_2_DM.html

 

Here's the official (FAA) definition, which meets exactly what you're talking about:

METERING- A method of time-regulating arrival traffic flow into a terminal area so as not to exceed a predetermined terminal acceptance rate.

A rate is being set off of the ILS-based rate. Traffic is then held to that rate. That's metering. I don't know how else to put that.

 

Yes - but usually the constrained area will not be an airport, rather enroute. Airports not being constrained is kind of my point, in fact.

So en route is the constrained area, but the constraint is being set off of the ILS rate? That doesn't make any sense. In the ILS case (which is the topic of discussion), the constraint is still the airport (because the ILS is related to said airport), even if the effects of that constraint cause congestion in the en route environment.

 

I think I get what you’re trying to say – that en route airspace may also be constrained, and CFMU can help to avoid that – but it’s not exactly as important to the ILS-based metering discussion.

 

As for usage, yes, you would have to fill FP to CFMU system, but usually you would not see a delay if you are departing within planned time. Delays are more usual for late flights, that just don't fit into the enroute picture anymore. You could re-plan though I guess. Or you can always send a ready message.

The en route picture, of course, is set off of the schedule. The schedule is set off of a rate of acceptance. This is all based on an ILS rate.

 

Over there, you get a delay if you're late - even on a severe clear day.

Over here, you just go when you're ready - provided it's a clear day (when it's not, our practices are predominantly similar).

 

It's not metering as you seem to understand that is the thing I am talking about. Flights are scheduled so that you don't need reduced separation in the first place. Then you won't run into problems when you can't get reduced separation.

See, I think this is where we are not on the same page - of course you are metering. Because if you didn't - you would have too much planes trying to get down somewhere there isn't enough space. That is my point - that would not happen in Europe until you go down past 550m RVR or so.

 

Again, metering is time-regulated traffic flow. Another word for ‘time-regulated’ is ‘scheduled.’ What it seems that you may be misunderstanding is that in the FAA world, nothing is scheduled until we need it to be. No schedule exists until we put a program into effect. As such, we do not meter - AT ALL - until it's deemed that there may be a need to reduce traffic flow into a particular area. This is done via what we call "programs."

If you go here: http://www.fly.faa.gov/ois/

Have a look at the top table. If there’s something listed there, then metering/scheduling is in effect (for the specific control element listed, and only that element). If nothing is listed there, then no metering is in effect.

In other words, we have no schedule to follow at all until we determine that there may be a need for one. That's the fundamental difference.

 

Europe schedules at all times.

FAA only schedules when it’s needed.

 

I still have no idea how you come to this conclusion.

I came to that conclusion through the following:

 

CFMU – Severe Clear Day:

Basing everything off of how many aircraft you can throw down the ILS is setting a rate. We’ll call that 20 (our average acceptance rate in “regular IMC”). As such, you can take one aircraft every three minutes. Slots are then set at three minute intervals, and flights are matched to those intervals. If three flights had an ETA of 1200, one of those would get 1200, and the next would get 1203 and 1206, respectively. Since they were all planning on getting there at 1200, one flight got three minutes of delay, and one got six minutes of delay. Over these flights, the average delay is 3 minutes. It’s small, but it’s still an assigned delay to fit into a slot (a fundamental part of metering).

The advantage is, if the weather drops, this plan is unaffected.

The disadvantage is that, in periods of good weather, you’re sacrificing higher potential arrival rates.

 

FAA Practice – Severe Clear Day:

No rate is set. Based off of the nominal single runway stats for VMC, the runway will see an average of about 30 aircraft per hour (though DCA commonly pushes 40/hr). This is just a nominal statistic, however, and a rate is not set to meet this. As such, there are no slots, and no delay associated with matching flights to those slots. If three aircraft had an ETA of 1200, all three would be allowed to depart on time on the assumption that it would either work itself out, or be handled by vectors in the TRACON. Given the ability to clear pilots to essentially self-separate, based on the nominal rate, the controller would have to strategically issue delay vectors to establish a sequence, amounting to 2 minutes of delay. In actuality, airports like DCA can commonly push an aircraft per 1.5 minutes. That’s 1/3 to 1/2 of the scheduled delay as part of a scheduled/metered approach. Additionally, since this is done tactically, and flights do not all arrive and depart exactly to the schedule, the delay vectors that are tactically applied are generally not the full theoretical delay because an implicit delay has already been incurred due to variation in the individual flight ATDs, wind, and so on. Using the same three hypothetical flights as above, the average delay is 2 minutes.

 

The advantage is that the plan operates at a higher efficiency than scheduling off of a lower rate (because there is no schedule – sequencing is all tactical).

The disadvantage is that, if the weather drops, controllers shoulder the initial load (but again, the weather doesn’t often suddenly drop, and programs would be put into effect in anticipation of forecast weather).

 

Now, multiply that 1 minute difference in the average delay by the number of flights going into an airport. The number will probably be pretty significant.

 

The US is one big agency with a plethora of airspace, Europe is made of tens of ANSPs all with minute airspace in comparision.

I've mentioned a couple of times why I understand why it works for Europe. What you brought up is essentially why. It working doesn’t necessarily mean that it’s as efficient as it could be, though (granted, “as efficient as it could be, given all the red tape” may be valid).

 

I guess having the worlds busiest single and duel runway airports we must be doing something bordering on efficient .

:rolleyes:

 

And there it is!!! The never ending UK Calling Card of "your TM arguments of efficiency are invalid because we push a ton of traffic into two airports that we never bothered adding runways to." * + #

 

*EGKK has two runways, technically, though they ignore the closest because they put the runways too close together for their own rules of standard separation to be useful so they stack departures on it. So, functionally, it's the busiest single runway airport.

+Another airport is being evaluated as a replacement for EGLL at an alternate location specifically because of the limited space, limiting the ability to add a reliever runway.

#

 

 

But I like to think that since JFK has adobted the European way, as in metering/slots for departure, things are much much better there!

 

Right, but that's definitely not the norm for here, and only affects departures.  Arrivals are not metered unless some type of constraint is expected (like right now, N90 is getting hammered by low ceilings - note the lack of holding/diversions if you look at FlightAware right now)