Maximizing Efficiency in the Tower Environment

Shane VanHoven

  • Mentors
  • 120
    • View Profile
Maximizing Efficiency in the Tower Environment
« on: June 20, 2020, 09:19:40 AM »
The ground and tower positions on VATSIM are typically a student's first exposure to working airplanes. So, as expected the controllers you usually encounter on these positions when flying through are slightly less versed in air traffic control concepts than the radar controllers. I'd like to break down some rules that can be used by tower controllers during high volume situations so that newer controllers might be able to more easily understand their options and can be well equipped to excel during events at the tower level.

I remember when I was first starting out on the network, it sometimes felt like ground and tower were just fillers, and that approach and center were the actual "money makers." During events, especially with high departure counts, a ground controller has the power to make or break the airport... being the difference between a line at the runway 20 deep, or no line or delays at all. So to all the S1s and S2s out there: take pride in your job as a ground and tower controller. Make the effort now and that work ethic will flow into your higher level training in the radar environment.

What do I mean ground can make or break an airport? I'm talking about the departure sequence. Many major airports have multiple "tracks" that aircraft use to exit the terminal airspace. This is usually in the form of multiple SIDs, but sometimes they are just fixes or canned routes that aircraft are assigned to fly before joining their filed route.

Familiarity with your SIDs as a ground controller is important not only for runway assignments, but also for your sequence that you deliver to your tower controller. If you taxi two airplanes out to one runway that are departing on the same SID or route, the tower controller must provide adequate in-trail radar separation to the departure controller. Normally 3 miles. If you taxi two airplanes to the same runway that are filed over two separate SIDs, the tower controller can launch the airplanes with minimum runway separation, which I'll probably detail in a separate post (but the jist is READ THE 7110.65!! ;)). Essentially, having back to back departures over the same fix is less efficient than alternating SIDs between successive departures.

How can this be easily applied? At the beginning of events, make a plan. Decide what destinations or SIDs will have the highest volume. That is, if you're working a departure airport for a crossfire event, you'll know that you'll have an extreme number of departures on that route compared to other random ones. So, the good technique in situations like this is to "feed" the departure runway in more than one way. Line up all your event/crossfire departures at one taxiway, and all the other randoms at another taxiway, maybe a parallel, or on the other side of the runway. Most major airports have the pavement to apply this technique.

For the purpose of example, look at O'Hare. https://aeronav.faa.gov/d-tpp/2007/00166AD.PDF

If departing 10L, the ground controller could line up the event aircraft on L, and the random departures on N. That way the tower controller can put non-event departures in between the event aircraft in order to not delay random flights outside of the event, and to keep the airport moving without just completely sitting still while waiting for whatever intrail restrictions are in place for the event airports.

All too often I see event airports with a line of 25 airplanes at the runway and a departure is rolling every 2 minutes. So, let's work to improve and learn all the options we have as ground controllers to help excel at events!

Discuss below your facilities' ground plans for events I'd love to hear them!
Shane VanHoven
Minneapolis ARTCC
Private pilot, Instrument, ASEL
FAA Air Traffic Control, ORD ATCT

Nolan Danziger

  • Mentors
  • 112
    • View Profile
Re: Maximizing Efficiency in the Tower Environment
« Reply #1 on: June 20, 2020, 09:04:21 PM »
Good post, Shane. The cab guys often get overlooked when it comes time to plan a big event, but sequencing into the destination starts from the moment an aircraft pushes back from the gate.

We're pretty spoiled with DFW (https://skyvector.com/files/tpp/2007/pdf/06039AD.PDF). Not only do we have parallel runways so we don't have to worry about any complicated squeeze plays, we also have plenty of holding/sequencing pads that we can use. The layout of DFW is pretty awesome too. We utilize directional departures so that westbound departures use the westbound runways and vice versa for eastbound guys. It's always interesting to learn how other facilities deal with their situations!
Zoomer 😎

Jeremy Peterson

  • Members
  • 250
    • View Profile
Re: Maximizing Efficiency in the Tower Environment
« Reply #2 on: June 23, 2020, 08:16:46 AM »
Operationally, the cab (including the delivery, ground, tower, and ramp controllers if we have one) works best as a team. This means proper controller handoff briefings are vitally important—the last thing a controller needs to do is go, “oh wait, was this aircraft given clearance already?” Another important aspect, which is applicable for any controlling position, is awareness of one’s own volume threshold. Not everyone has the same controlling abilities; some have been controlling for decades, and some just got their certification. The implications for these differences in controlling ability is two-fold: on the one hand, there is responsibility by the Events Coordinator to ensure controllers are not being overwhelmed with traffic; and, there is responsibility by controllers to advocate when they are or anticipate being overwhelmed.

As far as the traffic management perspective, often enough, the most impactful area in the terminal environment is the runway. Depending on your airspace, changes in runway configuration can have a range of effects from a simple switch of a SID/STAR to a whole reconfiguration of the TRACON airspace. Take, for instance, JFK. In an ILS 22L/R, depart runway 22R and 31L@KE, the aircraft waiting to depart on 22R can use 31R and taxiway Charlie to queue (this is a LOT of space). For traffic departing 31L@KE, there’s only space on taxiway Bravo (and some on taxiway Alpha north of taxiway Juliet) for queueing meaning that area is very sensitive to surface congestion. Furthermore, the availability of runway crossings for 22R arrivals means taxiway Kilo must be protected so that area is very much a hotspot. Alternatively, on a ILS 31R, visual 31L, depart 31L operation, we use runway 22R for queueing if needed and all the taxiways on the east side of runway 22R. The hotspot here is usually aircraft exiting runway 31R not turning properly or holding short of taxiway Bravo with their tail over the runway, creating a hazard for upstream arrivals. This is where the efficiency of your tower operations team is most evident: an efficient and well coordinated tower team can recognize the potential surface constraints, anticipate pilot actions, and positively/proactively position aircraft to prevent backups or chokepoints.

As far as ramp goes, the execution of it is highly dependent on the skill of the ramp controllers, knowledge of existing real-world operations, and the application of this to the VATSIM environments. An example of this is how JFK manages Cross the Pond. Internally, we (ZNY) have considerable real-world knowledge and experience working with the JFK ramp controllers and we have come up with a system that works for the VATSIM environment and is highly flexible, allowing for one ramp controller to up to nine. Usually, we set it up like this: aircraft calls delivery for clearance, clearance hands off to ramp for pushback, aircraft calls ramp, ramp tells aircraft to pull up to a certain point and then call ground metering, ground metering informs aircraft of sequence or when to call the ground controller, ground gives taxi instructions, then ground hands off to tower, and that’s that. We’ve found that ramp works best for departure operations and NOT arrival ones (aircraft can just disconnect if there is no gate available).

My last point will be delegation of responsibility: in my experience, having an identifiable chain of command improves the communication process for the whole ARTCC. This means delegating supervisory functions to “Tower Supervisor” or “TRACON Supervisor” (or in the case of N90, “XXX Area TRACON Supervisor”). This allows for decision-making to be clear and concise, giving authority to one person instead of distributing it to all controllers. This doesn’t mean input isn’t accepted, it just means in the event that a decision needs to be made but there are conflicting opinions, the SUP can make the decision then and there. For TMU positions—and I realize that some facility’s don’t even use a TMU position at all, let alone multiple—for high impact events like FNO or CTP, having multiple TMU positions can really help the event run smoothly. For N90, we can split TMU functions into 4 positions, PHL TRACON into 3, and ZNY into 4 plus 6 departure PIT (works on issuing reroutes) positions and a supervisory TMC (STMC). Now, we have only used multiple ZNY, N90, and PHL TMU positions a few times but it considerably helps make the workload easier on the EC (who also has to deal with real-time staffing concerns and other managerial business) or whoever the traffic management controller in charge (TMCIC) is.

Anyways, that’s my essay, if you wanna know more, you probably know how to reach me, or at least someone does!
Jeremy Peterson (HP)

Brad Littlejohn

  • Members
  • 152
    • View Profile
Re: Maximizing Efficiency in the Tower Environment
« Reply #3 on: June 23, 2020, 12:22:38 PM »
Really good post here, Shane!

One thing I will say, however, and this is completely tongue in cheek, is that at ORD, you have it easy!  :P Let me explain.

With how ORD has changed so much to keep nearly every runway parallel in an east/west configuration (I think barely the 14/32s and 4/22s are still there), you've increased effeciency for the number of arrivals and departures that can come in, as they are all based in a single flow. The downside to that, as you've mentioned, is the in-trail separation needed for departures with the same SID.

When you have simultaneous converging runway operations with parallel runways, you can get that separation implicitly applied, while splitting the cramming of so many arrivals down the same runway by offloading them to the other set. For example, KLAS.

When in configuration #1 (landing 26L, departing 26L/R and 19L/R; we have 4 different setups there) there is one particular SID that only departs 26R and requires a right turnout (STAAV8), whereas everything else departing 26R requires a left turnout. That means you can implicitly get that in-trail separation for two STAAV8 departures by having a departure launch from the 19s after the first STAAV8 departure is airborne. It takes sequencing this at the ground level for this to be accomplished, otherwise every other aircraft would be waiting for that same in-trail separation to take off, creating that delay. So the sequencing at that level is very important. Also, the good part about that, is that while that departure off the 19s is on its takeoff roll, any arrivals that just landed on 26L can safely cross 26R, so they aren't waiting either. Now, this isn't to say to just send all departures except STAAV8 to 26R and all other departures to the 19s, as you'll still have in-trail separation on the other SIDs, wake turbulence, arrivals crossing, etc., to deal with.

Another example. In Configuration #3 (landing 1L/R, 26L, departing 1L/1R; 26R is closed/taxi only, owned by GND), we're limited the most by the airport layout for most taxi instructions. If you're on the GA side of the field (west of 1L), you'll have easy access to the runways, but the main terminals are east of 1R, so they will all have nearly the same taxi instructions, including crossing 26R. But because of this configuration, you're going to have more in-trail separation to deal with, as well as the parallel runway ops we see at JFK, ORD, LAX, and the like. So what we can do in this case is separate the departures by the turnout they have off of the departure end. For example, all of our southwest/northwest departures (BOACH,SHEAD,MCCRN) can be given 1L for departures, while everything southeast/northeast/east (TRALR, HOOVR,COWBY/LAS) can be given 1R, and both 1L and 1R departures (especially if RNAV) can be simultaneously launched. This way, you're maximizing runway efficiency for both arrivals and departures, instead of having only one runway for arrivals and one  for departures.

It's in a setup like this that I'm actually happy that LAS isn't going to a single flow setup for traffic, because we wouldn't have all of these different ways to organize traffic (I'll save Configs #2 and #4 for another post).

BL.

Joseph Kaiser

  • ZTL Staff
  • 8
    • View Profile
Re: Maximizing Efficiency in the Tower Environment
« Reply #4 on: June 25, 2020, 02:53:56 AM »
At ATL, all five runways are parallel and typically 2 of the 5 are used for departures (the two ones closest to the ramps on the north/center complexes, 27R and 26L). There's also the option to do "full triple departures" and depart off runway 28 as well.

The RNAV departure setup is extremely efficient and can be configured in multiple ways. The most common "departure split" involves south and east departures departing 27R and the north/wests off 26L. This allows for each direction to have a dedicated path immediately off the ground, with sufficient divergence to launch aircraft simultaneously/with just same runway separation. Bad IMC makes things less efficient, due to the lack of tower-applied visual between arrivals and departures on the same complex.

The ground sequence is accomplished by alternating the direction of departure - a south, then an east, then another south. If all departures are going the same direction, then they can be sequenced by individual departure gates/fixes. (There's four SIDs per direction.) The parallel taxiways can be used to accomplish this, with one direction on one and the other on the other, and then ground just alternates each into into one line to the runway. This is more difficult in east configuration, because as I explain below you can't really have departures going the same way toward the runway on the parallel taxiways with arrivals coming. As such, people might be held at the ramp for the sequence or might even be taxied eastbound on one taxiway then westbound on the other.

In west flow, for arrivals:

On the north complex we have the taxiway V loop, so arrivals taxi there "around" the departure end of the runway, simultaneously as aircraft takeoff, and then eastbound on the parallel taxiway to the ramp. Since all departures will also be taxiing eastbound to the runway, and the V loop spits the arrivals out at the westernmost ramp, regardless of parking there's not gonna be a nose to nose conflict.

On the center complex, it's the same deal with arrivals crossing the departure runway (27R) at a few taxiways that will result in them only going east to the ramps.

In east flow:

On the north complex, arrivals still taxi on V but it gets more complicated. There are two parallel taxiways between the ramp and departure runway, and usually the farther one is used as a westbound taxiway and the closer one as an eastbound taxiway to alleviate conflict, since the arrivals are going east to parking and the departures are going west to the runway.

Since there's no taxiway V equivalent on the center complex, we depart airplanes at intersection M2 and arrivals cross the runway behind the  departures uninterrupted. However, the same issue with east flow comes into play as with the north complex, with one taxiway treated as westbound (departures) and the other eastbound (arrivals).
« Last Edit: June 25, 2020, 06:27:36 AM by Joseph Kaiser »
ZTL Instructor