Seeing as I'm supposed to be on vacation this week, I thought I'd write about a subject that I enjoy: vectoring. I guess that doesn't make you (the pilot) feel all warm and fuzzy inside, does it? Knowing that controllers enjoy vectoring you all over the sky for spacing when you just want to go straight-in to your favorite hub? Oh well. What can I say? Most of us (controllers) do enjoy it and Mama said honesty is always the best policy. So, honestly, I enjoy it.
The place I get to practice vectoring the most is on the SHINE arrival into CLT (Charlotte, N.C.).
Charlotte SHINE 5 STAR -- simplified
(Click here for large version -- 320 Kb PDF.)
Looking at the chart you'll notice that the SHINE STAR has two transitions. From the HMV (Holston Mountain ) VOR you fly the 156 radial to join the CLT 314 radial at the SHINE intersection. From the VXV (Volunteer) VOR you fly the 094 radial to join the CLT 314 radial at the BURLS intersection. If CLT is landing south, the jets go in (enter CLT Approach airspace) at 11,000, the turboprops go in at 9,000 and the props go in at 7,000. It isn't depicted on the chart but CLT Approach's airspace starts just inside of SHINE, between SHINE and JOHNS. That means that all the airplanes should be lined up in-trail (the right distance between them) and at the correct altitude by SHINE. Notice I said "should be." It doesn't always work out that way.
A controller once had a bumper sticker made that said, "Air Traffic Control: Combining the Science of Physics with the Art of Ballet." Or something like that. It never really caught on because ... well, it just doesn't "sing." It was probably the best description I've ever seen of Air Traffic Control, though. At least a description that would fit on a bumper sticker, anyway.
The reason I mention it is that vectoring airplanes into a hub like Charlotte really is an art. You can learn to do it reasonably well in five years but you'll spend the rest of your career trying to master it. Don't get me wrong: The physics behind it are impressive. The wind constantly changes as the airplanes descend and change headings. The turn rates vary depending on the pilot flying, the speed, the altitude and the type of aircraft. I guess the best way of putting it is that the physics are so complicated and variable it becomes an art trying to control them.
There's one other point I'm trying make: Vectoring for in-trail is a very personal thing. I've never seen two controllers do it exactly the same way. I've never talked to two controllers who even think about it in exactly the same way. What follows is the way I think about it. I hope it will help you understand the process but it probably won't help you understand the way another controller runs his arrivals.
You can separate controllers into two camps when it comes to vectoring arrivals: those who like to use speed control and those who don't. Everyone has to use it to some degree, but I actually like using it. I feel it gives me better control of the situation and it's more efficient. Speed control isn't without its drawbacks, though.
A major factor on the SHINE arrival is that the airplanes are coming from the west and northwest. That means that they normally have the wind on their tails. If you slow a jet to 250 kts., it makes it tougher to get the airplane down. There's always a trade off to consider. Slow a jet to 250 kts and you increase the time it remains in the higher altitudes (and in the stronger winds), which decreases the effectiveness of the speed in relation to the other inbounds.
In the distant past, I ran arrivals into CLT from the northwest, the southwest and the northeast. Speed control is far more effective when airplanes are flying into the wind. If you slow an arrival from the northeast to 250 kts., at FL220 its ground speed will actually be slower than at 11,000, assuming a strong wind from the west. All thing being equal, the airplane in front starts its descent first and starts gaining ground speed in relation to the airplane behind. It just doesn't get any better than that (increasing separation) when you're running in-trail.
On the other hand, running arrivals in-trail on the SHINE arrival is entirely different. Just the other day I had a SHINE arrival from over HMV (the northwest) assigned 300 kts., and another SHINE arrival from over VXV (the west) assigned 250 kts. Despite the 50 kt. difference in indicated airspeed, the one from the west was overtaking the one from the northwest by about 100 kts.
Amazing isn't it? Not the speed but the fact that I'm talking about vectoring and I haven't even issued the first turn yet. Let's get to it. Let's say we have one arrival to CLT coming over HMV and another one coming over VXV. They are going to be at SHINE at exactly the same time. My first instinct is to clear the one on the VXV transition direct (or vector him) to the SHINE intersection. That cuts out the "dog leg" on that transition and it shortens the route by a few miles. It's usually enough to make the situation work without any other action. Well, that was simple wasn't it? My job is done. See you next month.
Oh, all right. We'll say there is another arrival 10 miles behind the first one on the VXV transition. Got that? Two from the west, 10 MIT (miles-in-trail) and one from the northwest over HMV. The one from the northwest is tied up with the first one from the west. The solution still looks simple: Shortcut the first one from the west direct SHINE. That cuts about 5 miles off his route and puts him in front of the one over HMV. If those two were tied at SHINE, that means the second one from over VXV should wind up 10 MIT behind the one from over HMV. We're done.
If it was that simple it wouldn't be any fun, would it? What I'm leaving out is the compression factor, among other things. If you have two aircraft 10 MIT in the flight levels, by the time they get to 11,000 and the first one starts slowing to 250 kts. to meet the speed restriction at SHINE, you'll wind up with about five MIT. Remember, as the airplanes descend from the west, the winds start diminishing and the ground speed starts bleeding off. The airplane in front is (supposedly) lower, so the next airplane in line starts gaining on the first one, even if they're assigned the same speed.
In addition, there's an axiom among controllers that says, "Never split a pair." In the above scenario, the "pair" is the two aircraft on the VXV transition that are 10 MIT. In other words, why mess up a good thing? You've got what you need on those two so don't spilt them up. In this case, you'd turn both of them direct SHINE and slow up the guy over HMV. The guy over HMV was tied for first but now he becomes airplane number 3. No, it doesn't seem fair but that's the way it works. We've got a lot of mileage to "lose" so I'd slow him to 250 kts. Yes, that will make it harder for him to get down. There is a method to my madness. Read on.
You can basically forget about the two from over VXV now. They're set up and we won't have to touch them again. The guy over HMV is the problem now and I can devote my full attention to him. He's slowed down now, which means he is descending slow and this limits the effectiveness of the speed assignment. I'll watch it for a couple of minutes and give the computer time to recognize that his speed has changed. At the Centers, the computer displays the ground speed based on the average history of the last six updates from the radar. This presents all sorts of technical complications but let's keep it simple: The computer is lying to you about the ground speed right now. You learn to live with it and make it work.
Now it's getting down to the nitty-gritty of the dance and it's decision time. You guesstimate that you're only going to have 3 miles between airplane number 2 and airplane number 3, so it's time to do something about it. You need to keep five miles between the two airplanes (or 1,000 ft. of altitude) because it's an Operational Error -- a "deal" -- if you don't. Airplanes number 1 and number 2 are on about a 110 heading to join the SHINE arrival at SHINE. Airplane number 3 is still on the HMV 156 radial so his heading will be about 160 or so to correct for the wind. Turn him right to a heading of 200 degrees. Now you have him at a right angle to airplane number 2 (110 vs. 200 degrees). In addition, that turns him into the wind so that his 250-kt. airspeed is slower than the other guy's (who is now slowing to meet the speed restriction at SHINE) 250-kt. airspeed. You also have the altitude working for you. Because airplane number 3 was slowed, he should be higher than airplane number 2. That gives you even more effect from the wind. And last, but certainly not least, you should have a couple thousand feet of altitude to work with just in case you guessed wrong on the vector. In other words, you can level him off a thousand feet above airplane number 2 if you see you won't have 5 miles between the two airplanes.
Simple huh? You ought to try it with 20 airplanes instead of just three. It gets a lot more complicated. It is fun though. As long as you're on your game, the pilots are on theirs and the equipment works, that is. Some days I can "choreograph" a masterpiece. Some days it looks like I need to go back to dance school.
All this isn't nearly as much fun as vectoring for approaches, though. I don't get to practice vectoring for multiple approaches as often as I run arrivals on the SHINE STAR, but it's a lot more fun. Some people think I have a twisted idea of what fun is but that's their problem.
The only place I get to practice vectoring for approaches is at the good-old Wilkes sector. For the new folks, Wilkes is my favorite sector and it's really nothing but an Approach Control for Hickory, N.C. (HKY) run by Atlanta Center (ZTL). The sector "owns" 10,000 feet MSL and below. While running multiple inbounds to HKY can be fun in and of itself, the real fun is when you have approaches going on at HKY, SVH (Statesville, N.C.), UKF (Wilkesboro, N.C.) and MRN (Morganton, N.C.) at the same time. The traffic is moving a lot slower but it's vastly more complicated.
All of the approaches to those airports overlap to some degree; at least using en route (or Center) air traffic control rules they do. I guess I should explain that. The first difference between a Center and an Approach Control is the radar systems we use. Centers use an Air Route Surveillance Radar (ARSR) and Approach Controls use an Airport Surveillance Radar (ASR). Just as the names imply, those radars were designed for two different functions. Center's radars (ARSRs or Long Range Radars) were designed to cover a lot of territory. Approach Control's radars (ASRs) only cover a small area. The tradeoff is the rotation speed of the radar dish. A fast rotation speed (like an ASR) means the radar beam can only go out so far and bounce back before the dish has moved (or rotated) too far to receive the reply (the bounce-back from the radar beam.) A slow rotation speed means the radar beam and reply can travel much further distances and still be received by the radar dish.
ARSRs rotate slower to see further. ASRs rotate faster to provide more (or faster) updates on the radar target. Therefore, an ASR is considered more accurate than an ARSR. That isn't technically correct but to save me from delving into areas beyond my understanding that's what we're going to go with. The gist of all this is, after the FAA's technical wizards played around with their slide rulers awhile, the lateral separation standards for Approach Controls were set at three miles and for the Centers the minimum is five miles. In other words, Center controllers can't run airplanes as close together as Approach controllers can.
(Note to nerds: Yes, I know all of the above is a gross oversimplification.)
Where were we? Oh yes, vectoring for approaches. I know two extra miles (five vs. three) doesn't sound like a lot but you'd be surprised how fast it adds up. We've had some Approach controllers transfer into Atlanta Center over the years and it just makes them crazy -- how slow it is -- trying to vector for approaches in the Center. They get used to it.
Despite the fact that I use a lot of speed control to run arrivals into CLT, I find I rarely use any form of speed control when I'm vectoring for approaches. It's a contradiction that I can't explain so I'm not going to try. I know Approach controllers use it constantly but I can't think of any Center controllers who do. We just vector.
I know pilots are concerned with how close in they are vectored to join the final approach course and the angle of intercept, but what I'm concerned about (as a controller) is timing. Let's look at the ILS RWY24 HKY approach and the MRN LOC RWY 03 approach.
Morganton (MRN) LOC 03
(click for larger version -- 245 Kb PDF)
Hickory (HKY) ILS 24
(click for larger version -- 300 Kb PDF)
The first thing you'll notice is that the approaches themselves don't overlap. But the missed approaches do. Remember this rule from the controller's "bible" - FAA 6110.25?
4-8-9. MISSED APPROACH
Except in the case of a VFR aircraft practicing an instrument approach, an approach clearance automatically authorizes the aircraft to execute the missed approach procedure depicted for the instrument approach being flown ...
That means I can't have an aircraft cleared for both of those approaches at the same time, because they will have a conflict if they both fly the missed approach. What's a controller to do? Vector of course.
For various reasons, I can't vector an aircraft for the final-approach course at MRN. I can vector aircraft onto the localizer at HKY. Does this mean I'll just work around the guy going into MRN by vectoring the guy going into HKY all over the place? Nope.
We still work on the "first come, first served" principal, so let's say the guy landing HKY is first. Let's also say that if I just let the situation unfold without controlling it, the aircraft landing MRN would be halfway through his approach before the aircraft landed at HKY. The only solution (without vectoring) would be to clear the MRN arrival to the FIQ NDB to hold. In that the holding pattern looks a whole lot like a procedure turn (that's a hint there, folks), I could make it work that way. But I have some other considerations and I'm going to vector the MRN arrival.
Can you follow all this? I'm going to vector a guy -- at an airport where I can't vector to the final approach course -- for spacing with an airplane that's on approach to a different airport. I sure get some puzzled responses from pilots when I do it for real. Let me show you how it works.
The guy landing HKY is now over MIRTY and inbound to the airport. That puts him 10 miles away from the airport. The guy landing at MRN is only five miles southeast of FIQ (Fiddlers NDB.) In other words, he'll be at FIQ before the guy landing at HKY either lands or misses. Due to some other considerations (other traffic), I want to control the timing of the situation. I want the guy at MRN on the approach the minute I learn the status of the guy landing HKY. What I'll do to accomplish that is vector the MRN arrival away from FIQ.
Let's look at the approach itself for a second so you can see the point I'm aiming for with the vector. You see that the inbound bearing from FIQ to MRN is 029 degrees. The pilot will have to execute a procedure turn, so I'm looking at the reciprocal bearing, 209 degrees outbound. I'm aiming for a point about five miles northeast of the FIQ. When the HKY arrival is on about a three-mile final, I want the MRN arrival to be about 5 miles northeast of FIQ.
"Cherokee two three four five six, turn right heading zero three zero, vector for spacing."
There are two more things to notice and then we'll see if we can pull this stunt off. Note that the initial-approach altitude for the LOC RWY 3 MRN is 5,000. The missed approach procedure for the ILS RWY 24 HKY is a climb to 4,000. 5,000 minus 4,000 still gives me 1,000 feet of separation. Convenient huh?
"Cherokee two three four five six, turn left heading two seven zero, expect direct Fiddlers in one minute."
I'm just vectoring him a little left so the turn won't be too long to get back to FIQ. In other words, I'm just fine-tuning the spacing a little and working on the timing.
"Cherokee two three four five six, five miles northeast of Fiddlers, turn left direct Fiddlers, expect approach clearance prior to Fiddlers."
You have to take the relative speeds into consideration, but with the turn it generally takes the MRN arrival longer to fly the five miles to FIQ than it takes the HKY arrival to fly the three miles to the runway.
"Atlanta Center, Cessna one two three four five, missed approach Hickory, request radar vectors ILS runway two four Hickory."
"Cessna one two three four five, Atlanta Center roger, radar contact."
"Cherokee two three four five six, two miles from Fiddlers, maintain five thousand until Fiddlers outbound, cleared localizer runway three approach Morganton."
Pretty slick huh? The MRN arrival doesn't have to fuss around with getting set-up on the outbound leg -- he'll be lined up on it the second he crosses FIQ. If I do it right, he's already on a heading very close to the 209 degree outbound leg for the procedure turn. I'm telling you, when things are going right, this is more fun than you ought to be allowed to have at work. As I told a fellow controller during a good session the other day, "If they put a coin slot on this radar scope and put it in a video arcade, people would feed it quarters all day just to play this game.
ATC can be a lot of fun. But it isn't a game. Unfortunately, it seems as if things "go right" less and less often these days. Between short staffing, aging radars, faulty frequencies, lack of training, record traffic -- the list seems endless -- I often wonder if things will ever seem "right" again.
But I thought it was important for you to know that controllers like what we do. Just like pilots who turn their love of flying into a career -- even if they grow to hate their job -- still love flying, controllers love controlling. God help us, we love it so.
Have a safe flight.
Want to read more from air traffic controller Don Brown? Check out the rest of his columns.