Stacked holding vs. vectors

Cal Goat

Prestige Worldwide™
Going into SFO this morning, we were the victims of a runway swap right as we started the arrival. This was after sitting for 90 minutes on the ground at 6am for flow control.

Initially, we were giving holding instructions at PYE, but it never materialized and we were given about 45 minutes worth of delay vectors. It created an annoying problem in that we were considering a diversion for fuel, but we had no clear picture of where we would be when we hit bingo fuel nor how much longer we could expect to wait at any given time.

Why can't ATC hold incoming airliners in a stack at a fix outside of the FAF so that 1, crews can make educated guesses about how long it will take them to divert and 2, we can be confident that we are being served in the order in which we arrive?

This was especially maddening today as SFO was VFR by a wide margin (broke out above 3000' and saw the runway from 15+miles with no clouds in between) and they were only landing on 1 runway. I don't pretend to know what controllers are dealing with on their end, but as a pilot, I'd wager to say that this ambiguous delay vectoring is actually a threat to safe decision making since it leaves crews more in the dark about their place in the sequence.
 
In my part of the country I cannot fathom 45 minutes of delay vectors. Holds, sure, delay vectors though, no. I'll gladly butcher praseology on a holding instruction with an EFC time. Having said that if SFO was VFR you don't get a hold around these parts while they change ends in nice weather.

My only contributions to this are A) At my TRACON (and we have a busy class B airspace design), I could stack like 6 planes at the FAF in holding 1,000 lower or higher and I'm in another controller's sector assuming 5 mile legs or less. 6 isn't that many honestly. and B) If you're seriously on delay vectors vice a hold for 45 minutes, ask for an estimate. Seriously, if some controller has time to play the delay vector game for 45 minutes they also have time to, order you a pizza, play you in a game of radio chess or read you 4 chapters of Moby Dick.
 
It depends on airspace configurations, traffic already in the terminal environment, et cetera. SFO always has ridiculously long delays every single day. There is no way that we could vector guys for 45 minutes inside our airspace or stack many (20+ aircraft) in our terminal area and still get the last departing aircraft, last arriving aircraft, satellite traffic out of the picture so we rely on the center to hold aircraft 40+ miles from the airport. Center is exceptionally good at getting into the hold, but it takes them a long time to get out of it. Ideally, we want the last aircraft touching down as the first arrivals are on a ten mile final opposite direction.
 
Thanks for the replies. Here's our track. They had us pretty slowed down by the time we hit PYE. They kept telling us they would have us on downwind "soon" and then saying "sorry, I need to turn you around one more time. Just a couple more minutes." I've been through this a few times with NorCal and it can create fuel stress simply because of the uncertainty with the clock.

Next time I'll keep pressing for updates. The companies I've worked for have always implied that telling ATC you have "minimum fuel" is a worthless radio call. You guys really only care when it's emergency fuel?
 

Attachments

  • image.jpg
    image.jpg
    179.1 KB · Views: 371
That track seems crazy to me, but I don't work there and have no idea what was occuring during that time period. We care about minimum fuel, but we're extremely concerned when you hit emergency fuel.
 
I don't work that airspace, but I can at least tell you that you would be dismayed by the lack of information the controller working you has about downstream activities. When center gets told "hold everyone for (airport) starting with (callsign)," that's about all we know for a while. It is kind of absurd how many parties are involved between that point and you touching down.
 
That's an awesome track. I used to live near PYE and have definitely seen them stack hold people on the Golden Gate 6 STAR at PYE and LOZIT. I'm not a controller (but I did get an awesome tour of Norcal Tracon once, and stayed at the Holiday Inn Express last night).

It looks like they swung you around PYE wanting to merge you into the right downwind for the ILS 19L. When that didn't work they sent you south to try and merge you into the HADLY2 STAR. That didn't work either so they sent you around again before finally getting you onto downwind.

While it doesn't excuse keeping you in the dark, I see some logic to those vectors. When SFO changes from west flow to southeast flow it's not just a runway change, all the STARS change, and even the directions those STARS come from. So it turns into a total s---show trying to merge the people who were on the original STARS (i.e. you) into the new ones.

That's my best guess anyway. YMMV, you got what you paid for, etc. :)

Edit: The controller who works that NW corner sector also isn't the final approach controller, so they could have been waiting for a hole in the people coming from the east to get you in. As I said, when SFO loses the ability to run parallel simultaneous charted visual flight procedures (to 28L/R), things go downhill quickly.
 
Last edited:
Going into SFO this morning, we were the victims of a runway swap right as we started the arrival. This was after sitting for 90 minutes on the ground at 6am for flow control.

Initially, we were giving holding instructions at PYE, but it never materialized and we were given about 45 minutes worth of delay vectors. It created an annoying problem in that we were considering a diversion for fuel, but we had no clear picture of where we would be when we hit bingo fuel nor how much longer we could expect to wait at any given time.

Why can't ATC hold incoming airliners in a stack at a fix outside of the FAF so that 1, crews can make educated guesses about how long it will take them to divert and 2, we can be confident that we are being served in the order in which we arrive?

This was especially maddening today as SFO was VFR by a wide margin (broke out above 3000' and saw the runway from 15+miles with no clouds in between) and they were only landing on 1 runway. I don't pretend to know what controllers are dealing with on their end, but as a pilot, I'd wager to say that this ambiguous delay vectoring is actually a threat to safe decision making since it leaves crews more in the dark about their place in the sequence.

Seems like you were just the guy in the wrong spot at the wrong time.... as far as minimum fuel I would hope most certainly change their plan to accommodate as best they can, but I have seen those who feel its just a "boy who cried wolf" situation.

As far as the above in bold.... holding stacks at least in a center environment don't necessarily mean you are going in the order you came. Its really where you are at the moment they say "GO!" Those who empty a stack based on altitude end up with a painfully inefficient sequence.

Just my 2 cents as a fairly new center controller. Opinions may vary widely.
 
I've been based at SFO for almost 3 years now. In that time I have completed one (1) actual hold. I've started 4 and ended up with vectors before getting to the back of the race track. Every other time it's the vectors game.

You make a really good point about the fuel planning and lack of information and this is one of the main reasons we generally take release fuel plus another 15-20 minutes (on a good day).
 
Heck, I'd probably always call for the maximum amount of fuel the plane could hold accounting for passenger weight and runway length.
 
I don't know the SFO operation, so I wont comment on the specifics of this particular case, but here (N90) our airspace is not designed to hold, so usually if we have to hold, we'll tell the feeder centers (ZNY, ZBW, and ZDC) to go in a hold, and they're the ones that have the holding patterns with airspace designed to protect for those. Any aircraft within our airspace that needs to hold will probably get vectored around if its not a very lenghty delay. Having said that, if the delay is going to be significant, then more than likely that we'll make up some holding point somewhere.
 
Back
Top