Say Again? #25:
Beating the System
The more pilots try to beat the system, the more work controllers have to do to keep everybody safe. AVweb's Don Brown had a recent difficult week when Atlanta ARTCC was understaffed, under-radared, and overstressed. This kind of week will only become more common as both traffic levels and ATC retirements increase, unless everybody goes back to following the rules.
Most days when I come home from work, no matter how bad things are (and they usually are), I take a few moments to calm down and take stock of my life. Invariably, I come to the conclusion that I've got it good. I've got a good home, a good wife (a great one actually), good kids. I've even got a good dog.
I'm living the American Dream: a wife, two kids, a dog and a two-car garage. Never mind that one of the cars has 120,000 miles on it. Just the mere fact that I'm able to afford all this without my wife having to work puts me way ahead of most Americans. Any man in my position ought to consider himself lucky.
Not this week. I don't think you can pay me enough for this week.
It all started out pretty good, actually. I went in Monday at 3:10 p.m. on the late shift like normal. As soon as I signed in I was bombarded with, "You should have been here this morning. It was a zoo." Yeah, yeah. It's always a zoo in here. It turns out that wave after wave of thunderstorms had been moving across the Atlanta airport (ATL) all morning. Every sector that works ATL inbounds had to hold. Not just the interior sectors designed to hold the arrivals ... every sector. We even shut off Indianapolis Center (ZID) and Washington Center (ZDC). We probably shut off the other Centers too but I didn't get to talk to the other Areas that night.
Been there, done that. So what? Well, Maiden was broke again, that's what. That's certainly not new. The Maiden radar is always going down. I've even written an article about it.
That's about as far as we got in the conversation before the day shift went home. By this time, all the storms that had been wreaking havoc at ATL were approaching Charlotte (CLT) and it was our turn to go into holding. Been there, done that too.
As a matter of fact, I kind of like holding. It's something different from the ordinary grind. At least in my mind, it's much more orderly than trying to "run the snake." The "snake" is where you slow everybody up and start giving all the inbounds "S-turns" back and forth across the arrival. You take the first guy and vector him 45 degrees off the arrival. When he's whatever distance off the arrival that you need you turn him back 90 degrees across the arrival for "X" number of miles and then another 90-degree turn back toward the arrival.
You do that however many times you need to for your spacing. Then you vector everybody else right in behind him. Instead of having a straight line of arrivals pointing toward the airport like normal, you wind up a "snake" wiggling its way toward the airport.
Now bear in mind that the head of the "snake" is descending toward the airport. So the "head" is down around 11,000 feet and the "tail" is up in the flight levels, depending on how long the "snake" is of course. I don't like snakes. They scare me. If you insist on playing around with them, don't be surprised when they bite you.
If I've got to handle one I'd just as soon get it into a box and keep it there until it's time to send it on its way. It can coil up as high as it wants but all I have to worry about is staying away from the box (i.e., the holding pattern). It's safer, more orderly and more efficient, in my opinion.
So there I am with airplanes stacked up in the holding pattern from 13,000 to FL210. For those that remember me talking about sector stratums you'll realize that I've left myself an "out." The sector owns 11,000 to FL230. That means I can go under the holding pattern at 11,000 and 12,000 or I can go over the holding pattern at FL220 and FL230.
Because I was expecting to be in the holding pattern for some time I wanted the option of running some en route traffic without having to take every airplane around the holding pattern. Yes, there is a method to our madness. Always leave yourself an "out."
Anyway, the thunderstorm finally passes CLT and we start running the arrivals again. Even that was relatively painless. Drawing from previous lessons learned, the first time I went through the stack, I added "say altitude leaving" to every descent clearance I issued. That seemed to clue everyone in that I wanted an "altitude leaving" report every time I issued a new altitude. It worked like a charm and we sent all the arrivals in. No fuss, no muss.
After a break, I spent the rest of the night working non-radar at the Wilkes sector. Again, being the twisted individual that I am, I actually like working non-radar. It's something different. A break from the ordinary. As long as the session is safe and orderly, it can be fun. Yes, my fellow controllers, I said fun.
During my very first session that night I had an inbound to UKF, one to HKY and another to SVH all at the same time. The approaches at SVH and UKF conflict with the approach to HKY. All three approaches have transitions off of BZM VOR. The SVH arrival was coming up from the south so I put him on the 7-mile DME arc for the VOR/DME RWY10 to SVH, put the UKF arrival on the BZM transition ILS RWY01 UKF and held the HKY arrival at BZM, above the other two airplanes.
Zip! Zing! Zoom! This is a blast. The UKF arrival is slightly ahead of everyone so he gets cleared for the approach first. The SVH arrival and the HKY arrival are basically on top of each other. I talk the SVH arrival into taking the arc for the VOR/DME RWY10 SVH so that saves me the extra time it would take for the procedure turn on the LOC. With any luck, the UKF arrival will be canceling just about the time that the SVH arrival is turning away from BZM VOR and I'll be able to clear the HKY arrival for the approach as he's turning back toward the BZM VOR on his first inbound leg of the holding pattern.
At about this point, another UKF arrival shows up from the east. No sweat. I put him in holding at BURCH intersection at 6,000. There's a transition to UKF ILS RWY01 from BURCH. As soon as the HKY arrival reports procedure turn inbound I'll be able to run the next UKF arrival in. I just love it when a plan comes together.
Best Laid Plans
"Atlanta Center, Frieghtdog five is missed approach UKF." Huh?!!! Have I mentioned I hate not knowing what the weather is like? I admit AWOS and ASOS have saved us a lot of frequency time and that's a good thing. But when we had to read the weather to every arrival, we were "in the loop." It forced any changes in the weather into our consciousness. I knew the weather was IFR but I didn't know it was that bad.
Oh well. I've got the missed approach protected just like I'm supposed to so it's no sweat. Well, OK, maybe I was starting to sweat a little but it's still safe and orderly even if the expeditious part just went out the window. As I set about trying to untangle the knot the D-side trainer (oh yeah, I forgot to mention a trainee was on the D-side) said, "This is like a &^%#$ chess game!" Exactly. A good game of chess is still fun, even if it makes you sweat a little.
The Beatings Will Continue ...
That was Monday. Tuesday was a horse of a different color. I can't put my finger on exactly how things started changing but they did. The traffic count on Monday was 8,307. Tuesday it was down to 7,718. On the surface, you'd think Tuesday was an easier day. But Tuesdays are always busier than Mondays. I think things were so bad on Tuesday that Central Flow (a.k.a., the Air Traffic Control System Command Center) took pity on us and started rerouting some airplanes around us. At least I hope that is what happened. Because we were taking a beating.
The same weather pattern was still around Wednesday. Thunderstorm after thunderstorm was moving through our airspace causing the usual havoc. The whole area was still IFR. I was working non-radar at the Wilkes sector again for most of the day. Although I was starting to get mad about the situation (I'll get back to that later) I was actually better off than most of the controllers in my Area. The high-altitude sectors were just being beaten senseless.
I should take a "before and after" picture of a controller that's been sentenced to the D-side at the Pulaski (PSK) High sector. They come back from break laughing and joking around like controllers always do, right until they turn the corner. Then they see the guy sitting at the PSK sector by himself with three bays of strips. They know it's coming and the supervisor says, "Take the D-side at Pulaski." They look like the judge just sentenced them to 20 years of hard labor.
They might as well. The position is unworkable. It's been that way for years. It was probably a decade ago that we instituted a "temporary" fix and split the sector in two, vertically. We split FL310 and FL330 off and made that a new sector, Salem. We knew it wasn't the best idea but it would have to do until we could redesign the sector and traffic flow. Now PSK only works FL240 through FL290. It didn't improve the traffic flow, but it did make the volume manageable.
Like I said, that was probably 10 years ago. Just how long is a "temporary" fix supposed to last? How long are we supposed to wait for a sector/airspace redesign? Just in case you think this is the only airspace problem, think again.
Let's Make a Deal
I'm on the Operational Error Reduction Team at Atlanta Center (ZTL) . The Quality Assurance Office gave us some interesting data to work with. The "top ten" sectors for Operational Errors at ZTL read like a who's who of sectors begging for redesign. Starting on the northwest side of Atlanta and working clockwise to the east you have Rocket High (#6), Crossville High (#8), Burne High (#3), Blue Ridge Ultra High (#9), Pulaski High (#5), Lainer High (#1 on the (near) hit parade), Spartanburg High (#7 and another FL240-FL290 "temporary" fix) and then Dublin High (#2).
That's 8 out of 10 problem sectors at ZTL that are in the high altitudes. If you put the pieces of the puzzle together you see that each one of them butts up against the other to form a wide arc from the northwest side of Atlanta circling around to the east and ending on the southeast side of Atlanta. It's not necessary but if you want to see all the pieces you can go here and click on the "Sector Maps" link. To see just the high altitude map click on the blue "High (240-330)" link.
If you look at the "top three" sectors (Lainer, Dublin and Burne) you'll notice that they have something else in common. Not only do they work the ATL traffic but they work the CLT traffic too. Lainer High (#1) works the northeast arrivals into ATL and the CLT west departures (after they pass through a small section of the #7 problem sector, Spartanburg). The Dublin High sector (#2) works the ATL arrivals from the southeast and feeds the CLT arrivals from the south to Spartanburg (#7). The Burne sector (#3) works the ATL north departures and the CLT arrivals from the west, which they feed to Pulaski (#5).
The Dalas Sector (#4) is a low altitude sector that runs the arrivals from the northwest into ATL. As far as I can tell it's only reason for its existence is to poke a hole in my theory because they don't have anything to do with CLT. Which brings us back to Pulaski (#5).
One Plus One Still Equals TwoAll that just to get me back to familiar territory. At Pulaski, we run two streams of CLT arrivals (the SHINE STAR), three streams of the ATL arrivals (the MACEY STAR), the CLT north departures, the GSO west departures and one stream of the GSO arrivals (the BROOK STAR.) It also works one stream of the Cincinnati (CVG) arrivals (SWEED STAR). Toss in assorted arrivals and departures at GSP (Greenville-Spartanburg, S.C.) and AVL (Asheville, N.C.) and you've got one complex sector. Oh, and don't forget all the en route traffic. Keep in mind that we've already spilt this sector vertically because of the volume.
Are you ready for my big theory? Here it is.
Volume + Complexity = Operational Errors.
Every controller out there is now going, "Big DUH dummy!" and they're right. It's just that simple. Every controller out there knows it so why doesn't the FAA, as an institution, seem to recognize it? It might be because the solution to complexity and volume isn't popular. Or cheap.
The solution to volume is More. More sectors. More frequencies. More controllers. The only "more" that the FAA (or their political bosses) wants to hear is "more with less." That's been the prevailing motto at the FAA for more years than I care to remember: "Do more with less." OK, we did that. Now what?
It's simple. Cut a sector in half and you cut the volume in half. Oh, all right. Maybe not exactly in half but you get the idea. In order to do that, though, you need a new radar scope, a new frequency and at least one new controller. None of which are cheap. That's assuming you can even buy one.
The other side of the equation, the one I want to concentrate on, is complexity. The way to limit complexity is to cut down on the number of "jobs" a controller has to do. To put it another way, you need to limit the different types of operations a sector has to handle. A sector can work a tremendous number of arrivals to one airport (like ATL or CLT) very efficiently, if the controllers aren't distracted by working the departures too. If, due to whatever limitations, the controllers are forced to work the arrivals and the departures in the same airspace the efficiency suffers. Add in volume and the safety starts going downhill too.
Again, I'm not telling you anything that any experienced controller doesn't already know. The solution is as simple as it is unpopular. You have to segregate the traffic. There has to be structure. The concept of "Free Flight" flies right out the window. That thought certainly won't earn me any Brownie points will it?
If you'll remember, I'm sitting at the Wilkes sector without any radar. I'm not really interested in winning any Brownie points at the moment. I'm just interested in surviving another day and making sure that you do too. We all survived Tuesday. Wednesday made me think we were pressing our luck.
Just last month I was telling you how I got the NOTAM process changed to include ARSR (Air Route Surveillance Radar) outages. By Wednesday I've figured out that it wasn't doing me (or anyone else) a heck of a lot of good. I hadn't found one pilot who'd seen the NOTAM.
One of the great things about writing this column is that I have a lot of new friends that help me figure out what is going wrong. By Thursday, one of my readers had figured out part of it and sent me a helpful hint. 25 pages worth. As in 25 pages worth of NOTAMs for his single flight. I had to search through them three times (using a word processor to help) just to make sure that the NOTAM wasn't in there. It wasn't.
I had another friend file a flight plan, using a different method, to see if he had any luck. Nope. No mention of the radar outage. I know it's out there. I spent most of my breaks this week verifying that the NOTAM was indeed in the system. By Friday I still hadn't figured out where it was buried though. I guess I'll have to make some new friends in the Flight Service option so they can explain the NOTAM system to me. That's assuming anyone actually can explain that mess.
But speaking of messes, let's get back to mine. Just last month I was saying thank you for all the things I've learned since I started writing this column. Here's one of the most important things I've learned.
AIM, Chapter 5-1-7. Flight Plan- IFR Flights
d. Area Navigation (RNAV)
1. Random RNAV routes can only be approved in a radar environment. Factors that will be considered by ATC in approving random RNAV routes include the capability to provide radar monitoring and compatibility with traffic volume and flow. ATC will radar monitor each flight, however, navigation on the random RNAV route is the responsibility of the pilot.
As some of my readers like to point out, the AIM is not regulatory. A few of them like to hold up the FARs (like a shield) and remind me that they only need to comply with the FARs to be legal. An airline pilot I met at Communicating for Safety gave me the best comeback to that I've heard. "Legal does not equal safe."
But it was too late. I'd already switched into the "legal" mode and made an adventure into the FARs. I wasn't really surprised to find that the FARs didn't have much to say about filing a flight plan. You can read about it here if you feel so inclined. Yes ... I know that's the VFR section (91.153), but if you look under the IFR Flight Plan section (91.169) it's just going to refer you back to 91.153.
Perhaps I should just say the FARs will tell you what to file but the AIM will tell you how to file. Regardless, the damage was done. Now I was reading the AIM in my "legal mode." Let's go back and read Chapter 5-1-7.d. of the AIM with a legal eye shall we? "Factors that will be considered by ATC in approving ..." Approving? You mean I (ATC) have the authority to "approve"? I guess that means I have the authority to withhold my approval too, doesn't it? "Factors that will be considered ..." Will be considered, huh? That's right up there with the word "shall" for controllers. What were those "factors" again? "... the capability to provide radar monitoring and compatibility with traffic volume and flow."
Well there's a no-brainer for this week anyway. I don't have any radar this week so I won't be "approving" any RNAV routes. But even when I get the radar back I "will be" still "considering" your "proposed" route of flight's "compatibility with traffic volume and flow." Does anyone (besides me) want to get back to the safety mode instead of the legal mode? That's what I thought.
Here's the way the system was designed to work. Back in the brown-shoe days, before the advent of computer data processing, when an aircraft called for clearance, the clearance request was routed through the Center. The D-side (Data Controller) picked up the strip and said, "N12345 is cleared from the GSO airport to the ATL airport via direct ... aw, that guy knows he can't go direct ... make it GSO, Victor 310, join Victor 20, BZM, Victor 222, LOGEN direct ATL."
That still happens in some limited airspace (like mine) where there isn't any Approach Control overlying the departure point. The majority of traffic, however, departs from airports within Approach Control airspace. When data processing became available, the flight plans still went to the Centers but they were also sent to the Approach Controls and Towers. The authority (and responsibility) to "approve" the route of flight still remained in the Centers.
Be aware that I'm glossing over some stuff (like the fact no one had RNAV back in the brown-shoe days) so this article doesn't go on forever. The next thing to come along was "silent departures." The Tower didn't call for a clearance or a release. They just launched them. The responsibility to review the flight plan for "compatibility with traffic volume and flow" remained with the Center but the "approval" became automatic. And it started to become a distant memory that anyone was supposed to be approving anything at all. Toss in RNAV, short staffing, lousy training and a few other issues and this is what you wind up with.
A flight plan I saw:
C182/U GGE 3322/8417 0I8 at 6,000
Make Me File Bad Flight Plans
Does anyone honestly think a Center controller reviewed that flight plan and "approved" it? I hope not. I mean, I really hope you don't believe they did and I really, really hope one didn't. Do I really need to point out how many things are wrong with this flight plan?
A "/U" that filed direct? Putting "VFR GPS" in the remarks section doesn't make it legal you know. Much less smart. A Lat/Long? I think I've beaten that horse to death already. A random route through a sector without radar? I guess he didn't get the NOTAM either.
Of course, if the pilot had filed in accordance with the AIM, the fact that I didn't have any radar coverage that day wouldn't have affected his route that much, if at all. He would have been on the airways. That's the thing about breaking the rules. Once you break one, it becomes harder and harder to comply with the other rules.
It Ain't a Pretty Picture
I guess it's time to wrap this thing up. As always, there are dozens of details that I'd love to go into but I just don't have the space and time. I'd love to walk the Center controllers through the steps of fixing flight plans that you don't "approve" of while they're still on the ground. I'd love to explain to pilots how following the rules and guidance the FAA publishes helps us, the controllers, and in turn gives us enough time to help you, the pilots. I'd love to explain exactly how direct routings, at any altitude, dramatically increases the complexity of air traffic control.
What I'm left with instead is giving you a snapshot of what it's like at work these days. I believe it was Thursday afternoon. I was sitting at the Wilkes sector again, sans radar, and sinking fast. Virtually every airplane entering the sector needed to be rerouted onto the airways. I didn't have a D-side to do it for me so I was trying to keep the ones I already had separated and give myself a fighting chance to keep the ones coming separated. There's that complexity thing again. I was losing.
I turned around to get some help and saw a whole herd of controllers plugged in at the Pulaski High and Shine Low sectors. They were losing too. Spring Low and Bristol Low were combined up like normal. They shouldn't have been, but they were. The Moped Low sector had to shut down the CLT departures at FL230 or below to keep them out of Pulaski High. That was causing them to sink and they were taking Bristol/Spring Low down with them. The only sector that was less busy than mine was Salem High. That was because the rides were so bumpy that no one wanted to fly at FL310 and FL330. Which was the reason they were all trying to get into the Pulaski sector of course.
That's when I realized that there wasn't any help to be had. I didn't have time to verify it but I believe every controller in my Area was plugged in. No one was on break. There weren't any reserves left. And that's an all too common occurrence these days.
... Until Morale Improves
As I said at the beginning of this article, I don't think you can pay me enough for a week like this. I don't have any choice, of course: I'm just over three years away from retirement and I can't afford to walk away from that. There are over 40 controllers at ZTL who can, though. They're eligible right now. If they walked, we'd be out about 15% of our active, journeymen controllers. And it takes five years to train a controller, but everyone is already too busy to train them.
Do me a favor will you? Help us out. Help us so we can help you. Stick to the rules. Read the AIM and try to comply with what it says. Then maybe, just maybe, some of those folks eligible for retirement will hang around a little longer. If you keep trying to "beat" the system don't be surprised when the people running the system decline to accept any more beatings.
Have a safe flight.