February 22, 2009

I've been Irked

I received a few kind emails recently filled with both concern and curiosity regarding the comments I made at the end of my last picto-blog. All has not been as perfect and peachy as it normally is inside my head in regards to air traffic, and now that I've stabilized my emotions a bit, I'm ready to tell those who feigned interest about it without naming too many names.

First things first. Please recall, or be reminded of, some basic fundamentals:

- The only thing in air traffic you can count on being consistent is things changing.
- This job is complex, and it's easy to forget that.
- For all the "support specialists" around the building and the agency, we have surprisingly little support.

Second things second.

I, like most controllers, just want to be left alone so I can keep planes separated. When others interfere, its a safety issue. Duh.

I'll try to keep this as simple as I can. I will fail at this, I'm sure. Be prepared for a long post. I could maybe split this up, but I'd lose my train of thought.


Currently, in the FAA, there are three ways that we sequence airplanes into airports. First, just let um go direct and descend them to the lowest altitude you own and hand them off to approach control. That works for places like ALB or SYR. Second, Miles in Trail (MIT). This is used most often to places with busy traffic, mostly the Class B airports (JFK, LGA, PHL, ATL, ORD, DTW, etc). Third, is "Metering". I mentioned this a few posts back about seeing someone Metering to Boston. Allow me to explain.

If you go WAY back, you'll find a post or two about departure delay programs that are designed to help controllers meet our MIT restrictions to major airports. Traffic Management Unit (TMU, which is basically a separate area in the center that worries about the "big picture") figures out how long it will take planes to fly from their departure airport to the point where we need to deliver our airplanes "in-trail" to another sector or center. These departure times help space the planes apart or else the airline's schedules will cause 10 planes to show up in the same sector from different directions all at the same time, creating a complicated and potentially unsafe situation. Airplanes can't be at the same place at the same time and live to tell about it very well.

The Metering program is a similar concept, but at the arrival end of the system. As the airplane approaches a set distance from the arrival airport, the flight is processed into the Metering computer. This computer estimates the time it will take, using an aircraft's normal speed and the winds aloft, to reach the runway in use at the arrival airport. It then displays the call signs of all the incoming flights on a timeline, showing where there is higher demand at certain times. Based on the arrival rate of that airport and runway, the airplanes are virtually sequenced farther apart, time wise, and the computer then computes how much time differs between the scheduled arrival time and the meter time based on realistic sequencing using standard separation on final approach. If a delay is needed, the time is shown, in minutes that need to be lost/wasted/delayed upon, to the controllers working the airplane as a little number next to the datablock on the scope. Theoretically, if a center controller can somehow make that number a 0, then the airplane will not encounter any delays being vectored onto final approach. They won't have to hold on the arrival because the new sequence time doesn't overwhelm the approach controllers. Please note the use of "theoretically" and "estimated".

This program has been in limited use over the past few years, and has been relatively successful within the scope of its use. The only airports that have Metered have been ones that only involve one approach control and one center. Denver Center meters to Denver approach. Los Angeles Center meters for LA approach. Boston Center meters for Boston approach. They kept it simple. And for the most part, it works out OK. Metering would assigned delays evenly over three or four arrival routes.

We began Metering to Newark in October of 2008. This is/was much more complicated. Cleveland, Boston, New York, and Washington Centers, plus New York approach, were all involved in one meter list on three different arrival routes. The metering program established crossing arcs at the boundary of each sector in all these centers that meter. The controller's goal at each sector is to use vectors and/or speed control so the airplane crosses the meter arc as close to the time assigned by the metering computer as possible. Sectors are expected to impose a certain amount of delay based on the size and complexity of said sector.

A main issue with Metering is that no one really understands how the computers calculate the times. Not even the folks at TMU have answers. So, controllers received a half hour briefing on WHY we are metering, which seems great, and the basic concept was thought to be understood. But no one gave us any clues on how to actually USE the program, or why it might not make any sense. For example, if you take the handoff on an airplane that has a little 3 next to the datablock, your sector is supposed to lose at least 2 minutes, making that number a 1 or 0. The next sector may have a bigger number, but that isn't your responsibility. So you turn the plane left and slow him down. The number goes to 0, so you turn the airplane back on course. "Uh, now that I have a 0, what speed should I assign?" "Uh, normal speed...I think." So you tell the aircraft to resume his normal speed... you handoff the aircraft to the next sector... and bam! you get a 2 again. WHAT!?

Well, I finally figured out how it works. The program bases this delay number off of an arc time at your boundary, and that time isn't supposed to change. The number next to the datablock assumes the aircraft will fly at his present speed direct to the next fix in the flight plan from his present location. If you instruct the plane any differently, the number is bound to get confused and change. So, I consider this a small victory. Just when I get things figured out....

Our story begins at the Delancey (DNY) sector with mostly light traffic. I take a handoff from Cleveland Center on a prop at FL270 (its a nice prop) going eastbound. That aircraft checks in and all is well. Then, suddenly, the same sector in Cleveland flashes me another airplane, also at FL270. This new flash is the A340 flight from Singapore nonstop. We work this plane everyday, and it saddens me if we every have to delay it. Its an 18 hour flight. Literally. They left Singapore before I got to work yesterday. Well, Cleveland Center doesn't like Metering, most likely because they don't understand it. So, they ignore the delays that are supposed to be incurred on the airplane. The first sector ignores it, so the 5 minute delay becomes a 7. The next sector ignores it, so it becomes a 9. Then, the last sector ignores it and tries to hand me this flight from Singapore with a 10 or 11 minute delay that I'm supposed to give this flight. This basically amounts to a holding pattern at HNK. I wouldn't have to delay the flight at all if the last three sectors had just slowed the plane down when it was over Michigan. Yeah, we still have to delay the airplane, but slowing the plane down a little far out actually saves gas, and no one really notices the difference in speed.

So, they're offloading a 10 minute delay on me....and they're giving me a deal at the same time. It's not going to be a weak deal, either. The targets are going to merge at Fl270 in about 3 minutes if I don't do something. So, I get on the landline and call the Geneseo sector. "I show a 10 minute meter delay on Singapore 22, give him a spin and we'll save this other guy at FL270" "uh..oh...rgr"

To make this situation more entertaining, my D-side is a rowdy, boisterous controller who thinks refusing this handoff is the greatest thing anyone has every done in a decade. He predicts that within minutes, our supervisor will get a phone call complaining about this "bold" move- refusing a handoff. No more than 10 seconds pass before the phone starts ringing, and my D-side starts wringing his hands anxious for our supervisor to tell Cleveland to do their job and get over it.

My Irk'ness lies in what really happened. Instead of setting the phone down, walking over to our sector and asking us why we spun Cleveland, *OUR* supervisor comes running over in a flurry, practically yelling at the both of us "Why aren't you taking handoffs? What the hell is going on?" I calmly respond "There was a 10 for a delay, and they were giving me a-" "Who cares, just take it, they said thier TMU worked it all out, it'll be fine!". "Uh, they were also giving me a deal at the same time, I was killing to birds with one stone (ok maybe not the best phrase...)" "What! Just take the plane on a 090 heading, just take the handoff!"

So, my D-side starts yelling at the supervisor to get on the phone and actually defend our area for once. I'm appalled that I have to defend my obviously correct action to prevent a mid-air collision over Ithaca, NY. And another controller pipes in how ridiculous the whole situation is that we have to add 10 minutes of delay to the last 30 minutes of an 18 hour flight.

The situation was almost funny, if it wasn't such a systemic problem. Cleveland had been protesting the whole Metering thing from the beginning by telling their controllers not to delay aircraft, just pass it on to Boston. I was naive expecting Cleveland would do their jobs, and in the end, my supervisor just didn't like being bothered.

Just when got it figured out....
As a result of this and many other similar instances where Cleveland didn't really do their job, and as a result the program didn't really work as deisgned, and thus many controllers opposed the Metering program as a whole, Newark Metering was put on hold for a while. (Edit: On Feb 23rd, we "tested" it out with slow traffic, and it didn't go very well)

I didn't get too flustered by that incident at first. However, there were a few other little things that added to my frustration towards the people I work for, culminating in a flat denial of a facility tour for a friend that flies jets for a major airline. Apparently, its a security issue. We're not allowed in cockpits and pilots aren't allowed in the Center. Thank god we're all safe.

Thanks for putting up with my little ranting craziness. My attitude has been slowly improving this week, but hopefully I can at least partially recover from any indignation I have towards my mostly incompetent employer.

Till next time...



Level 7,000 said...

After getting a tour of ZME one day I truly respect what you have to do to get traffic from point A to point B. I have seen what the metering data on the radar is displayed like and it is pretty interesting because I fly a BE58 and I still get metered by ZME, I just started to ask them how long and I slow up a bit. I would rather that then a turn 90 degrees off heading just to burn another 5 gallons of fuel. I appreciate everything that you do even if I do not directly fly through the Northeast corridor. Time will tell when the managers have to actually do their job. Good luck, many are watching, I promise!

Geoff Arnold said...

Thanks for the explanation. I was wondering how metering into BOS worked, and why it wasn't more widely used.

Steve said...

Excellent explanation of how the heck it all works. And I completely agree with the ridiculousness of all the "security" regulations. I might only be a Private Pilot, but it's still odd that I can't get a tour of the facility where the helpful ATC folks I talk to over the radio all the time work.

Jimh. said...

Too bad there are not more like you on that side of the tower!

Don Brown said...

" My attitude has been slowly improving this week,..."

Not to worry...the FAA will 'fix" that attitude soon enough. :)

Hang tough, DM. 25 years is a loooong time to work for the FAA.


ZOB Controller said...

As a ZOB controller, from the area your speaking of, i have to correct a few things. First off, metering times for the EWR arrivals, especially the aboved mentioned Singapore flight, doesn't start until they cross the border from toronto. so the first and only sector that will get a time on that flight is the GEE sector. So slowing him over michigan, not gonna happen. second, before you slam your brother/sister for not meeting his/her time, maybe you should research the problems that meter has within the northeast. it's not the controllers that are broken, it's the system. in 70 miles i've seen the metering times go from +30 to -5 because someone is adjusting times. so in that small distance, i've turned, spun and speed controlled planes to death only to watch all that work go down the shitter when someone changes the arrival order. i've shown a time of +3, slowed the aircraft down and had ZBW controllers tell me that they show a -3. the system is the problem.

You are right on several things though, one, you should have never gotten those two planes at 270. you should never get handed a deal. two, cleveland center controllers hate metering, because its broken.

i know it's easy to blame that anonomous guy in the OTHER center/approach/tower. but you should really dig deeper into the metering program before you blame controllers for its mistakes.

Anonymous said...

So you've been in the FAA for four whole years and you will blog about Cleveland not doing their job? It's nice to know that you have all the answers.

How about you look at all the problems that metering is full of in regard to EWR and the actual makeup of the sectors in ZOB and Toronto before you start blogging about something you know very, very little about.

The metering times are a work in progress and it's been documented time and again that Cleveland times and Boston times do not coincide most of the time. Even if we show zeros, you probably show a time and we're told to stick to the times.

If we(ZOB) get a time that doesn't seem right, they will make a call and, in the case you brought up, they had done just that and it had been worked out. Not ZOB's problems that communication on your end is that poor.

In your blog, you state several times about Cleveland not doing their job, but do you realize that a/c comes through Toronto Center and is worked by only one Cleveland sector, which typically doesn't display a time until the a/c is on the northeast side of Buffalo? Did you know that, or just showing how inexperienced you really are?

As for two at 27 coming together, I can't address that since I have no exact knowledge of the incident, but to be fair to you, lets say it happened how you say(which based on most of your blog I doubt), but let's say it happened that way, but we aren't infallible and do make mistakes. To spin an a/c to make a point when a call to say those two might be tight, lets do something. Our controller may have dropped the ball, but your smug reaction to show how big you are is typical small man syndrome and not how a truly professional controller acts.

We have enough of those on our side of fence, trust me when I say that, but since you're young and learning, take some friendly advice when offered.

Learn more before you start slamming your fellow controllers.

deltamike172 said...

I appreciate all your comments and I take anything an FPL suggests to heart to better myself. I have posted some clarification on the metering aspect of this post, but I think the ZOB controllers may be overreacting how I view them.

"My Irk'ness lies in what really happened. Instead of setting the phone down, walking over to our sector and asking us why we spun Cleveland, *OUR* supervisor comes running over in a flurry, practically yelling at the both of us"

Mistakes happen, and sometimes we don't catch everything, and that is by no means what my issue is. I felt that the best move was the spin the SIA22 since I would have had to anyways. A heading to miss the prop at FL270 was definitely an option, especially if I didn't have to hold the A340 down the road.

I really didn't think anything of the situation until my supervisor started in on me and my D-side.