KENNEDY SPACE CENTER, Fla.--Engineers have traced an electrical problem blamed for grounding the shuttle Endeavour Friday to a power distribution box in the ship's engine compartment, officials said today. Replacing the box will delay launch until at least May 8--Mother's Day--and possibly later.

"I'm here to disappoint everybody by saying I'm not going to tell you what the new launch date is because I have no idea," Mike Moses, chairman of NASA's Mission Management Team, told reporters after engineers decided on a course of action. "We have a lot to evaluate, both the work to do, the R & R (removal and replacement), the retest that has to be done, how we work all that schedule in.

A technician works in the shuttle Endeavour&39's cramped engine compartment where a suspect power distribution box is located. Work to replace the box will delay another attempt to launch Endeavour to at least May 8 and possibly longer.

"But we can tell you pretty much it's not going to be any earlier than the 8th. That doesn't mean we're going to go launch on the 8th, that just means we know right now the 8th is our next available opening," he said.

Launch Director Mike Leinbach said engineers plan to remove the suspect aft load control assembly--ALCA-2--box from Endeavour's cramped engine compartment tomorrow, install a replacement Tuesday and get into a complex re-test procedure Tuesday night or early Wednesday.

To make a launch at 12:09:17 p.m. EDT on May 8, NASA would have to start a fresh three-day countdown around 10:30 a.m. Thursday. Whether the team can complete the ALCA-2 swap-out and re-test in time remains to be seen.

But if Endeavour does not make May 8, launch likely would move to May 10. A launch on May 9 could result in the shuttle undocking from the International Space Station on May 23, the same day a Russian Soyuz crew ferry craft is scheduled to depart. There are no known conflicts for subsequent launch opportunities.

In the meantime, "the team is upbeat," Leinbach said. "A little disappointed, of course, that we couldn't launch. But responding to problems is one of the things we do best around here and the team always likes a good challenge. I'm sure we're going to be really glad when Endeavour's finally on orbit but right now, the team is upbeat and ready to execute the plan that we've laid out."

The aft load control assembly electrical switching box is located on the right side of the shuttle&39's engine compartment.

Endeavour commander Mark Kelly and his crewmates--pilot Gregory H. Johnson, Michael Fincke, Gregory Chamitoff, Andrew Feustel, and European Space Agency astronaut Roberto Vittori--flew back to Houston early Sunday aboard a shuttle training aircraft to participate in additional ascent simulations later this week.

"Things happen fast," Johnson said in a Twitter update. "We are now all aboard an STA for return to Houston. Be back in a few days. More to follow."

Endeavour was grounded Friday during the final hours of the countdown because of telemetry indicating multiple fuel line heaters used by one of the shuttle's three hydraulic power units were not activating normally. The heaters are needed to keep the lines from freezing and possibly rupturing in flight.

The shuttle is equipped with three auxiliary power units, providing the hydraulic muscle to move the ship's engine nozzles, wing elevons, rudder, tail fin speed brake, body flap, landing gear brakes and nose wheel steering system. The shuttle can safely fly with a single APU, but flight rules require full redundancy for a countdown to proceed.

Likewise, each of the shuttle's three APUs is equipped with redundant heater "strings" and only one channel is required for normal operation. But again, the flight rules require redundancy to protect against a subsequent failure that could knock the system out of action.

Early Saturday, engineers ruled out a problem with the fuse panel in the shuttle's cockpit that routes power to the APU circuitry. That left two possible culprits: one or more faulty heater control thermostats or the aft load controller assembly the heater circuitry runs through.

To find out if a faulty thermostat was to blame, engineers working in Endeavour's cramped engine compartment Saturday afternoon sprayed compressed air on APU No. 1's B-channel heater thermostats to lower their temperature enough to find out whether they would cycle on or not. They did not, but that could have been the result of a wiring problem or a bad connector. Additional tests were carried out overnight and no such problems were found.

Engineers met early Sunday and recommended replacing the ALCA-2 box.

"The box will come out tomorrow and we'll send it down to our malfunction lab for a detailed inspection," Leinbach said. "The new box goes in on Tuesday...And then after that, we get into the re-test Tuesday night, Wednesday, that kind of timeframe. It's going to be a full two days of re-test."

The shuttle's electrical system features three main circuits, or buses, for redundancy. As a result, three aft load control assemblies are present in the engine compartment.

Each 50-pound box includes dozens of power switches that route electricity to components in nine major systems, including the auxiliary power units, the environmental control and life support system, solid-fuel booster electronics, the shuttle's main engines, its orbital maneuvering system rockets and flight control systems.

The box is located just forward of a right-side engine compartment access door and Leinbach said the replacement operation was not particularly difficult. An ALCA was changed out during a 1995 shuttle launch campaign and engineers will use the same procedures for Endeavour.

The issue for NASA is the time needed to complete testing to make sure the myriad subsystems downstream of the box are receiving power as required.

"Anytime you break connection to a box like this, you essentially invalidate all the testing we did up to that point," Leinbach said. "You could take the tack of saying all you're doing is replacing the box and everything downstream of that box should be fine. Well, that's true, But our requirements, the way we do business is whenever we break a connection we go back and retest it.

"That's just the prudent thing to do and the way our requirements are set. So we have to retest every one of those nine systems. The details within those systems, you could probably write a thesis on how many individual tests there are within those nine systems. And that's why it takes so long."

Assuming NASA sticks with a May 8 launch, Endeavour would dock with the International Space Station around 9 a.m. on May 10 and the mission's primary payload, a $2 billion particle physics detector, would be attached to the lab complex the next day.

The mission's four spacewalks would be carried out May 12, 14, 16 and 18, before undocking around 2 a.m. on May 20. If that schedule holds up, landing back at the Kennedy Space Center would be expected around 6:30 a.m. on May 22.

But NASA managers plan to extend Endeavour's mission by two days, if possible, to give the shuttle crew time to help their space station counterparts with needed internal maintenance. In that case, undocking would slip to May 22 and landing would be expected before dawn on May 24.