The Wall Street Journal
Why Even Sunny Days Can Ground Airplanes
Flight 88 Got Stuck On a Congested Route; A Legacy of the 1920s
By PAULO PRADA and SCOTT MCCARTNEY
September 28, 2007; Page A1
On Aug. 23, a sunny day in Fort Lauderdale, Fla., Delta Air Lines Inc.'s Flight 88 pushed back from the gate on schedule, awaiting clearance for its 12:40 p.m. takeoff to New York's John F. Kennedy International Airport. The weather was cloudless in New York, too. But for the next three and a half hours, the aircraft, with 135 passengers, sat on the ground, taxiing back and forth between the gate and the runway.
Why? Thunderstorms over Rochester, N.Y. — 260 miles northwest of Kennedy Airport and nowhere near Flight 88's flight plan.
"There was good weather in Fort Lauderdale, good weather in New York, and most importantly, good weather en route," recalls Gary Edwards, Delta's chief dispatcher. "But we were still delayed."
That's because weather wasn't really the culprit. Passengers of Flight 88 — like legions of other hapless travelers this summer — were really delayed by the constraints of an antiquated and inefficient system for managing U.S. airspace. A lack of capacity in a seemingly boundless sky translates to massive delays on the ground.
After this summer's travel troubles, much attention has been paid to the myriad hitches that contribute to airline delays: from old air-traffic-control technology, to the growing number of corporate and regional jets, to labor tensions among air-traffic controllers, and overscheduling by airlines. Yesterday, following two days of congressional hearings on airline delays, President Bush met with Transportation Secretary Mary E. Peters and pressed her and legislators to find solutions to the problems.
Yet one fundamental shortcoming in the nation's air-traffic system has gone little discussed: the federal map of routes, largely unchanged since the 1950s, that airplanes are required to follow.
Just like rush-hour freeways on the ground, the nation's airways, particularly on the East Coast, have become choked with traffic. Block one with a small thunderstorm and jets sit on the ground waiting for hours because there's no room for them on other routes.
"The limiting constraint is the airspace as opposed to the concrete," said W. Douglas Parker, chief executive of US Airways Group Inc.
Never have such constraints been as apparent as this summer, during which fewer than 70% of commercial flights arrived on time, according to figures compiled by the Department of Transportation through July. (On-time arrivals for each of the past three years have hovered at about 77%.) Fixing the problem requires addressing issues that involve a roster of players including airports, airlines, air-traffic controllers, and the Federal Aviation Administration. (See related article.1)
"What's hard for everyone to understand is no one really controls the system, not even the FAA," says Russ Chew, president at JetBlue Airways Corp. and until earlier this year chief operating officer at the FAA. "Airlines can't tell one another not to schedule flights. The FAA can't tell airlines how to schedule flights. It's like saying, 'There's too much traffic on a California freeway at 9 a.m. — fix it.'"
Before it could depart for New York, Flight 88 had to wait as air-traffic controllers rerouted planes flying east toward JFK that were affected by bad weather on their route. But by pushing those planes south, some as far as Atlanta, controllers then had to squeeze them through one of the country's worst airspace bottlenecks: around Washington, D.C. The airspace there is particularly tight because blocks of nearby sky are used sporadically for military training and often unavailable to commercial flights.
Waiting in Fort Lauderdale, Flight 88 wanted to get on a highway in the sky labeled as "J121." Its entrance ramp was a particular navigational point — a radio beacon on the ground — in Charleston, S.C. But before Flight 88 could even get off the ground and head for J121, all manner of airlines and corporate jets were also converging for their turn to enter the same airway. Flight 88 found itself grounded, waiting to enter the queue. Delta dispatchers tried to get the flight on other airways; those were jampacked too.
Part of the blame lies with airlines, which continue to schedule more flights than during the boom of the late 1990s, even though they are using fewer aircraft and employees. The cutbacks give airlines less flexibility to reschedule flights or replace aircraft in the event of delays or cancellations. Further, airlines have increasingly filled their schedules with multiple trips on small, regional jets. Three regional jets on those airways carry as many people as one large jet.
Airlines also point to the rapid growth of private-jet travel as another clogger of airborne arteries. A corporate jet with only a couple of people on board uses the same routes, effectively taking up as much space on route J121 as does Flight 88, with scores of passengers.
Airlines, airports, and even administrators say the nation's air-traffic control system, which technologically lags behind systems already rolled out in other parts of the world, is too slow to handle even today's volume. The stress on the system first became apparent a decade ago, when the Internet boom and a strong economy fueled the last big air-travel bonanza. But the urgency to address it disappeared after the technology bubble burst and in the wake of the Sept. 11, 2001, terrorist attacks.
As air-traffic levels recovered in recent years, the system has struggled to keep up. "This is a very old ATC system that can't handle traffic and weather as well as it should," says Marion Blakey, administrator of the FAA for the past five years until her term expired earlier this month.
The nation's airways evolved from air-traffic routes established in the 1920s when the government was developing airmail service. Pilots followed established ground routes, generally flying low enough to trace actual roads and spot one geographic landmark, then another. In 1926, the Air Commerce Act authorized the government to build a network of other navigational aids, beginning with bonfires that were later replaced by illuminated towers and, eventually, radio beacons and radar.
Even now, the concept remains the same: Today's pilots, while flying at much higher altitudes, still follow the same routes, flying from one radio transmitter to another. As they move along, aircraft get passed from one air-traffic controller to another. Each controller, in turn, is responsible for a specific chunk, or sector, of airspace.
But with so many planes in the sky, controllers don't have as many holes in the conga line to move planes from one airway to another if a particular flight path gets congested. Add thunderstorms or other disruptions, and the problem gets worse. Airways are so close together, especially in the Northeast, that it's hard for jets to deviate from the route without butting into another airway. Another problem is that airways are extremely wide — eight miles across — because of the inaccuracies of radar.
"There's so much demand today, you can't reroute on other freeways," says Bob Reding, head of operations at AMR Corp.'s American Airlines.
So planes sit, and sit. For hours.
Consider JFK airport, which has grown quickly over the past decade due to greater use by traditional carriers such as Delta and American, which use it as an international gateway, and the rise of JetBlue, a discount carrier that chose JFK as its primary base when it began flying in 2000. In July, the latest month for which figures are available, JFK reported 39,351 takeoffs and landings, compared with 31,849 in July 1999.
Delays have grown along with the number of flights, as controllers struggle to get them all airborne.
Data compiled by the DOT's Bureau of Transportation Statistics ranks JFK 30th out of 32 major airports in terms of on-time departures for July, with only 62% of flights taking off on schedule. (Dallas/Fort Worth and Atlanta ranked lower.)During the same month in 1999, 81% of JFK's flights departed on schedule.
What's more, JFK airport leads the country in so-called "taxi out" delays, according to data the DOT recently began reporting for flights that take more than three hours to depart once they have pushed back from airport gates.
Air-traffic controllers say JFK should be able to handle about 85 takeoffs and landings an hour on a normal day. Yet because of the crowded airways leading into and out of the region, the airport rarely runs at full capacity. This summer, for instance, during the busy travel times between 3 p.m. and 10 p.m., the airport was handling an average of 74 flights an hour, according to FAA figures.
"JFK has the physical runway capacity," says Mr. Reding. "It should be able to handle more flights."
Aviation authorities say airlines should be more realistic when making schedules. "Their hopes and expectations may be boundless," says Ms. Blakey, the recent FAA administrator. But "neither the airport nor the airspace will be able to handle all the traffic that is planned and hoped for."
Some problems around New York could be alleviated by a plan the FAA recently authorized to redesign airspace over the region. It aims to make it easier for planes to enter and exit airways by allowing pilots to depart in a greater number of directions upon takeoff. The so-called "fanning" of departures, the FAA says, will ease the flow of aircraft once they have left the runway.
But the airways themselves will remain just as congested, prompting a growing call for the FAA to accelerate plans for a wholesale upgrade of the air-traffic-control system.
Experts say the solution is a satellite-based navigation system that would allow planes to abandon the highway maps and fly freely since a computerized system can check for conflicting flight paths. Known as NextGen, short for Next Generation Air Transportation System, such a system would allow navigation points to be moved to avoid storms. Planes could fly closer together since they would broadcast their exact position using satellites for accuracy rather than relying on slower and less-accurate radar hits used now.
But creating that infrastructure will be complex. The FAA last month awarded the first contract, valued at $1.8 billion, to ITT Corp., for the initial phase of upgrading equipment to a satellite-based system. To replace the entire domestic radar system with NextGen, taxpayers will be expected to foot a bill expected to run as high as $22 billion by 2025.
Most airlines, as they upgrade their fleets, are already installing equipment needed to fly within the NextGen system. But there is only so much they can do. Many of the exact specifications won't be known until future contracts wend their way through government bureaucracy. "We'll invest in the technology, but we need to know what it's going to be," says Joe Kolshak, chief of operations for Delta.
Meanwhile, problems unrelated to technology will continue to affect how efficiently airways are managed. Labor issues, for instance, have plagued the FAA as it sought in recent years to cut costs, paying new hires less since September 2006 while at the same time facing a recruitment crunch to replace retiring controllers, many of whom have worked for the agency since they were brought in to replace strikers fired by President Reagan in 1981.
Privately, executives at several airlines say they believe labor disputes are leading air-traffic controllers to reduce how many planes they handle at one time, creating delays. One day this summer, for example, the FAA's Washington Center insisted that traffic be spaced out every 15 to 20 miles as jets entered its airspace. Normally, five miles is the minimum. Frustrated with the delays caused by the spacing, "We called the command center," says Delta's Mr. Kolshak, "and immediately it dropped to 10 miles with no explanation."
Controllers and FAA managers deny that labor issues are causing any disruptions to the flow of aircraft. They recognize, however, that the greater number of planes flying this summer at times has pushed the limits of individual controllers and operations centers. "We don't always get it right," says Michael Cirillo, vice president of system operations services at the FAA. "There may be spacing at times that we may not have needed."
Out of the hands of both controllers and airlines is yet another issue that complicates travel: "special use" airspace restricted for military or security purposes. Airways around Washington, D.C., for instance, must avoid sensitive government buildings and airspace used for training by the military. Training areas blanket the East Coast, from Palm Beach, Fla., to Nantucket, Mass., some beginning as close as a mile or two offshore and extending out more than 150 miles.
If a storm arises to block a critical north-south route, planes on the coast often can't veer offshore because they'll enter military practice areas — even if there is no military activity going on. They can't go inland without entering other airways that are already full. That forces controllers to put jets into holding patterns and stop departures. Airlines suggest the military could be more flexible with the airspace and go further ashore to practice, but the Pentagon says military pilots are constrained by the system too, and that offshore flight zones are critical to national security.
At Delta's massive operations control center near Hartsfield-Jackson Atlanta International Airport, the airline's flight dispatchers recently watched on computer monitors as aircraft traversed the East Coast after a morning of heavy rains in the Northeast. On a close-up of airways leading to and from the airspace around JFK, 11 blue blips indicated departing aircraft while seven green blips represented arrivals.
Though the morning storms had cleared, managers were closely monitoring the day's weather to anticipate problems and the potential for blocked airways. A storm doesn't even need to materialize before the FAA sometimes grounds planes before traffic tie-ups begin in the sky.
"We can gridlock LaGuardia in a heartbeat because of rain in western Pennsylvania," says Neil Stronach, Delta's vice president of operations planning, control, and reliability. "And that's just impossible to explain to customers."
Write to Paulo Prada at paulo.prada@wsj.com and Scott McCartney at middleseat@wsj.com
URL for this article:
http://online.wsj.com/article/SB119093585795041925.html