Satellite-based (aka, NextGen) technologies have been in use for decades, and at most airports they have enabled minimization of distance flown and fuel burned. In fact, at the very few airports where NextGen is failing, the problem is not the technologies: it is too many flights, and FAA’s lazy refusal to impose more restrictive airport flow rates.
If you spend any time studying today’s routes and flight profiles for U.S. commercial passenger flights (and it is REALLY easy to do, with FlightAware, FlightRadar24, and other websites that present FAA’s ATC data), you will see that all flights are already capable of and actually flying optimized routes: long, direct flights from origin airport to destination airport, with smooth and continuous climbouts and descents. But, for a small handful of airports, you will also see that ATC ends up creating long conga lines of low, slow and loud arrivals (the Long Island Arc of Doom is the classic example) … simply because there are too many flights arriving in too small a time window.
The root problem is the hub system, and FAA’s policy of enabling undisciplined hub scheduling by the dominant airline. FAA does this to maximize a theoretical number called ‘runway throughput’, and thus to help the airlines to maximize their profits. In simplest terms, a hub airline can tweak their profits upward a percentage point or two, if they can process say a dozen simultaneous arrivals, sorting the passengers quickly between gates, then send all those flights outbound at exactly the same moment.
Obviously, this is only theoretically possible. Because of limited runway capacity, each arrival and each departure needs roughly a one-minute window where the runway is theirs alone, so the scheduled ‘banks’ of a dozen ‘simultaneous arrivals’ and ‘simultaneous departures’ get spread out over two 12-20 minute windows. To safely handle the arrival banks, ATC has to level off the arrivals and extend the arrival pattern to long final legs, spacing the flights at roughly one-minute intervals; to process the departure banks, ATC issues immediate turns on departure (with terrible impacts in places like Phoenix), so that takeoff clearances can be issued in rapid succession.
The reality that FAA and Bill Shuster refuse to accept is this: runway capacity is limited, and we can pretend to be creating new technological solutions, but so long as there are only so many arrivals that a key hub airport can handle per hour, it is folly for FAA to let hub airlines schedule in excess. It only guarantees delays, which then cascade into other airports that otherwise would never see delays. Also, it is important to note that hourly flow rates do not address the problem. Delays happen every time, when just two arrivals aim to use one runway at the same minute. So, if FAA is to work with the airlines to design delay-free arrivals, the schedule needs to look at small time increments, even how many arrivals every 5-minutes. Fortunately, this finer data granularity is easily studied with todays digital processing capabilities.
The solution is obvious: we need Congress to change the laws, so as to disincentivize excessive hub scheduling; and, we need FAA to aggressively restrict airport flow rates at key delay-plagued hub airports, so that the conga lines never need to happen.
An Example: Seattle Arrivals
Here’s an example of what happens at an airport, when just one more flight creates enough traffic, to necessitate ATC stretching the arrival pattern. Seattle is a great example, because it is a major hub airport but [KSEA] is far from other major airports, thus flight patterns are not made more complicated by airport proximity issues. The dominant airline is Alaska (including its feeder, Horizon), but Delta began aggressive hub growth in 2012. The airport has triple-parallel north-south runways; a south flow is by far the dominant airport flow configuration. Whenever ATC has enough arrivals to reduce spacing to less than two minutes apart, the arrivals are extended downwind, turning base abeam Ballard (12nm), abeam Northgate Mall (14nm), abeam Edmonds (20nm), or even further north (see this graphic that shows distances on final from the runway approach ends).
The scrollable PDF below has sample arrivals on December 29th, with altitudes added to the screencaps, to illustrate level-offs and descent profiles. Five sample arrivals are included:
- Horizon #2052 vs Horizon #2162 vs Horizon #2405: all are Dash-8s, from KPDX. Horizon #2052 has no traffic and is able to use the preferred noise abatement arrival route over Elliott Bay; the other two flights both have to extend to well north of Green Lake, including a long level-off at 4,000ft.
- Alaska #449 vs Alaska #479: both are from KLAX. Alaska #449 has no traffic and is able to use the preferred noise abatement arrival route over Elliott Bay; Alaska #479 has to extend to well north of Green Lake, including a long level-off at 3,800ft, starting to the west of Alki Point.