KENNEDY SPACE CENTER, Fla.--After three launch delays due to technical problems and bad weather, the shuttle Discovery was grounded again Friday. Its final launch is delayed until at least November 30 because of a hydrogen leak in a vent line attached to the ship's external tank.

After the scrub was declared, engineers discovered a large crack in the protective foam insulation on the tank that likely would have caused a launch delay, even if the leak had not developed. Engineers have not yet evaluated the crack to know what sort of repairs, if any, might be needed or even whether the damage can be fixed at the launch pad.

But engineers are hopeful that the problems can be resolved by the end of the month for another launch attempt at 4:05:46 a.m. EST.

"We will come back in late November and give it another shot," Launch Director Michael Leinbach said. "I'm disappointed for the team today, for sure. But as we always say, and it's absolutely the truth, we're going to fly when we're ready, and clearly, we were not ready to fly today. So we'll come back another day and try it again."

Asked if Discovery was "jinxed," given the problems that have repeatedly delayed the veteran orbiter this week, Leinbach said "it's a machine, and every now and then, machines break, and right now, our machine is broken, and we need to go fix it."

"No, we're not jinxed at all," he said. "We're just dealing with one problem after another. Does it get frustrating It's frustrating, because we'd rather be launching and getting up to the space station, sure."

Hydrogen gas spewing from a 7-inch vent line attached to the shuttle Discovery&39's external tank triggered another launch delay Friday. The next launch window opens November 30.

The leak developed during fueling for a planned 3:04 p.m. EDT launch. Sensors near the vent line indicated a "significant" leak with gaseous hydrogen concentrations of more than 60,000 parts per million. Leinbach said the concentration actually was higher than that, but the sensors go up only to 6 percent. Launches are not allowed to proceed with concentrations higher than 4 percent.

Leaks in the vent line attachment fitting--called a ground umbilical carrier plate, or GUCP--delayed two earlier shuttle missions in 2009, but extensive corrective actions were taken, and engineers believed that the problem was resolved.

The earlier leaks began when the tank was full, and the hardware in the vent line had been chilled to cryogenic conditions. The leak today developed much earlier, while the tank was being filled, and the leak rate was much higher.

Engineers cycled a valve inside the quick-disconnect fitting in a bid to clear out any ice or other contamination that might have caused the leakage. But the system remained out of limits, violating both NASA's launch commit criteria and ground safety rules.

Leinbach and Mike Moses, chairman of NASA's Mission Management Team, then decided to call off the countdown, delaying launch.

Engineers quickly began assessing possible repairs in hopes of making another launch attempt Monday, the last day in Discovery's current launch window. But the timeline was tight, and engineers will not even get their hands on the hardware to find out what went wrong until Saturday.

Complicating the picture, a launch on Monday would have required changes to the shuttle's space station resupply mission because of temperature issues related to the lab's orbit.

Instead, Moses said, the management team decided to give up on Monday and to press instead toward at launch at the opening of the year's final shuttle launch window, on November 30.

At about the same time, engineers inspecting Discovery's external tank spotted a 7-inch crack in the foam insulation near the top of the intertank region on the side of the tank facing the shuttle.

It's not yet clear what caused the crack or what, if anything, might be needed to repair it. But given that it's on the same side of the tank as the shuttle, repairs likely will be required. Whether that work can be done at the pad, where access is difficult, remains to be seen.

Discovery's crew, meanwhile--commander Steven Lindsey, pilot Eric Boe, Michael Barratt, Nicole Stott, and spacewalkers Timothy Kopra Alvin Drew--broke quarantine and flew back to the Johnson Space Center in Houston after the delay was announced.

Discovery commander Steven Lindsey and Nicole Stott, along with the rest of the shuttle&39's crew, flew back to Houston today after a launch scrub.

The vent line is used to carry excess hydrogen gas away from the shuttle when the tank is filled with super-cold propellant. A valve used to route hydrogen to the vent line is closed a few minutes before launch, when the tank is pressurized for flight.

In the most recent previous GUCP leak, the shuttle Endeavour was grounded June 13 and 17, 2009, when sensors near the umbilical attachment plate detected hydrogen concentrations of more than 6 percent.

After the second launch scrub, engineers collected detailed measurements and concluded that the problem was caused by an alignment issue between the hydrogen vent port on the tank and the vent line interface. To ensure a tight fit, engineers replaced a rigid Teflon seal with a more flexible design, modified the umbilical plate mounting pins, and installed washer-like shims to counteract the alignment issue.

The shims provided additional pressure on one side of the attachment fitting that was believed to be pulling away slightly as the hardware contracted under cryogenic conditions.

The fix worked, Endeavour was safely launched in July 2009, and no other GUCP problems occurred before Discovery's fueling Friday.

Discovery's launch originally was scheduled for Monday, but the flight was delayed to Tuesday and then Wednesday by work to replace suspect quick-disconnect fittings in the pressurization system of Discovery's right-side orbital maneuvering system rocket pod. Then, during checkout of the avionics used to control Discovery's three main engines, an electrical glitch was encountered that prompted a 24-hour delay to Thursday.

After a detailed engineering review, NASA managers decided that the unexpected electrical "signature" was not a threat, and Discovery was cleared for launch. But the weather refused to cooperate, and NASA's Mission Management Team decided early yesterday to order another 24-hour delay.

The next launch window opens November 30 and closes December 5 or 6. If Discovery isn't off the ground by then, the flight likely will slip to the end of February, pushing launch of the shuttle Endeavour into late April or early May. NASA's final shuttle flight, a station resupply mission with Atlantis, is targeted for next summer.