The agency also hasn't set deadlines for when key aspects of the new air traffic control system will be in place, Scovel said. Nor has FAA made clear to airlines and other air traffic system users exactly what benefits they can expect and when they'll be achieved, he said.
As a result, airlines and others are being discouraged from spending money on cockpit equipment necessary to take advantage of the new air traffic system, Scovel said. Many of the new system's benefits hinge on airlines equipping their planes with expensive new equipment to communicate with air traffic controllers and broadcast their location to other planes and controllers.
The Associated Press was provided with a copy of Scovel's testimony to the House Transportation and Infrastructure Committee. FAA officials had no immediate comment when contacted late Tuesday.
FAA officials have predicted the agency's NextGen modernization program will be as revolutionary for civil aviation as was the advent of radar six decades ago. It's actually a collection of new programs aimed at moving planes faster and more efficiently that will markedly change almost every major aspect of today's air traffic system. Those changes are considered critical to enabling the system to absorb substantial predicted increases in air traffic without becoming paralyzed by congestion.
The troubled $2.1 billion software program is the main tool air traffic controllers will use to identify and track aircraft, except when planes are immediately approaching and departing airports. It was supposed to have been completed by the end of last year, but the FAA now doesn't expect to be finished until 2014, at a cost overrun of $330 million, Scovel said.
Those estimates may be optimistic. A Mitre Corp. study and an analysis by the inspector general's office estimate the added cost of the computer system, called ERAM, could be as much as $500 million, with potential delays stretching to 2016, Scovel said.
David Grizzle, the head of the FAA's Air Traffic Organization, described ERAM to an air traffic control conference earlier this week as "the chassis on which all the NextGen functionality will be bolted." He said the system is expected to increase the number of planes controllers can handle by nearly two-thirds.
"We've got to get it right, and we've encountered some significant challenges in delivery," Grizzle said.
Some of ERAM's problems have previously been made public, including glitches that incorrectly identify planes and interfere with the ability of controllers to pass along responsibility for tracking a plane from one control center to another.
But the problem may be more extensive than the FAA has previously acknowledged. Scovel said his office has found similar problems in another critical FAA computer system that shares the same aircraft tracking software. That system is used by controllers to track planes as they approach and depart airports.
"ERAM's persistent problems have raised concerns about the overall design of the system," he said in his testimony.
Cost overruns in ERAM will affect the FAA's budget for other major elements of NextGen and could "crowd out other critical programs," Scovel said.
The FAA has focused much of its initial NextGen efforts on improving the flow of air traffic at congested airports in 21 major metropolitan areas. However, the agency has been slow in developing the flight procedures that will allow airlines to save fuel and time by flying shorter, more direct routes, Scovel said. The FAA did a study that identified ways to streamline the process for deploying new procedures, but agency officials estimate it would take five years just to put the streamlining initiatives in place, he said.
In the meantime, many of the new flight procedures the FAA has approved are merely "overlays of existing routes" rather than the fuel-saving procedures that require the more advanced navigation capabilities airlines had expected, he said. Airlines are anxious to reduce their jet fuel cost, which vies with labor as their No. 1 expense.