Damage from hurricanes Katrina and Rita has cost NASA three months' work in getting the space shuttle ready for its next flight, but planners said yesterday that they should be able to launch next May.

NASA officials also said they have found no single cause for the loss of insulating foam from the external fuel tank during Discovery's July launch, but they expressed confidence that technicians could prevent a recurrence by changing the way the foam is applied.

"We found several potential contributing factors," said Richard Gilbrech, leader of the NASA "tiger team" studying the foam loss. "We do not believe any one of these would have caused the release, so we are attacking as many potential contributors as we can."

NASA grounded the shuttle fleet after several pieces of foam broke away from Discovery's external tank, including a 0.9-pound chunk from the protuberance air load, or PAL, ramp, a ridge of insulation designed to protect cables running along the outside of the tank.

The debris did not hit Discovery, but the incident embarrassed NASA, which redesigned the tank after the loss of the shuttle Columbia, which disintegrated during reentry in 2003 after a chunk of tank insulation breached the orbiter's heat shielding.

Gilbrech's team began its investigation of the Discovery foam loss shortly after the July 26 launch. The shuttle program manager, N. Wayne Hale, said in yesterday's telephone news conference that the team has spent the past three months doing "analytical work" to map the tests that will be needed before the shuttle flies again.

Hale said the work was significantly hampered by Katrina and Rita. NASA's Michoud Assembly Facility, which builds the external tanks, is in New Orleans, and while employees kept the installation free of flooding, the work schedule has been thrown into disarray.

"The plant came through very well," Hale said. "Our biggest problem is infrastructure." He said the hurricanes washed out roads into the facility and wiped out the homes of many Michoud employees.

Five hundred employees currently are working at Michoud, he said -- about one-quarter of the total. "We lost the equivalent of three months because of the hurricanes," Hale added, but the May launch window "is something we can work toward."

For upcoming shuttle flights, NASA has insisted on launching in daytime so a full range of imaging devices can examine the orbiter for damage. The number of days this can be done in any one period, however, is limited by the need for the shuttle to rendezvous with the international space station without exhausting its auxiliary fuel. The May window runs from May 3 to May 23.

Gilbrech said engineers working on the tank insulation at Michoud are focusing on three factors as the collective cause of Discovery's PAL ramp mishap. One was a section of foam that had been inadvertently "crushed" during preparations for the mission and then repaired.

Also, he added, engineers were examining the "knit lines" that mark the edges of different foam layers, which are sprayed by hand on the PAL ramp. These are roughly equivalent to "curtains" that develop in spray painting and may enhance the possibility that layers of foam could be peeled away by the stresses of launch.

Finally, Gilbrech said, engineers will examine air pockets in the foam, known as "voids," which can expand and perhaps burst when the insulation expands and contracts because of differences between the air temperature at Florida's Cape Canaveral and the super-cold liquid fuel inside the external tank.

Early in the investigation, engineers had suggested they might eliminate the PAL ramp altogether. Hale said this is a future possibility, but the current fix calls only for improved procedures: using newer foam and applying it with better techniques, doing more imaging of the foam's interior structure, and keeping people away from the insulation so it does not get damaged.

The space shuttle's external fuel tank leaves Kennedy Space Center in Florida en route to a barge that will take it to NASA's Michoud Assembly Facility in Louisiana for modifications.

Richard Gilbrech, left, with shuttle proram manager N. Wayne Hale, discusses his team's findings on possible causes for foam loss.