CAPE CANAVERAL, Fla. - NASA engineers are attempting to root out the cause of a sensor glitch that scrubbed Wednesday's launch of the space shuttle Discovery.

Pad engineers drained Discovery's fuel-filled external tank of its super-chilled liquid propellant late Wednesday after a sensor sent errant signals to flight controllers during a countdown check.

"All I can say is shucks," said Wayne Hale, NASA's deputy shuttle program manager, during a press briefing after the failed launch. "We came out here all ready to go here today...and we incurred a problem."

Discovery's launch, slated for today at 3:50:53 p.m. EDT (1950:53 GMT), is now delayed until Saturday, July 16 at 2:40 p.m. EDT (1840 GMT) at the earliest, NASA officials said.

"That's probably the very best case scenario," Hale said.

Earlier, NASA chief Michael Griffin told reporters that Discovery's STS-114 spaceflight - the space agency's first attempt to resume shuttle operations since the Columbia disaster - could slip to Monday if not later.

Unsolved anomaly

At 1:30 p.m. EDT (1730 GMT) today, launch controllers conducted a test of four external tank fuel sensors required to function properly before the orbiter can lift off.

The sensors are designed to monitor the fuel levels inside Discovery's external tank during launch in order to shut down the shuttle's main engines before propellant runs out. If one sensor fails, the shuttle could keep firing its engines on an empty tank, which is a hazardous situation, NASA officials said.

During Discovery's countdown, launch controllers directed the sensors - which read 'wet' with a full fuel tank - to read 'dry' as if the tank was empty. But one sensor inside the liquid hydrogen section continued to read 'wet,' violating launch restrictions that call for four, fully functioning sensors.

"It was clearly a violation of our launch commit criteria," Hale said. "It took us five minutes to decide [to scrub]."

NASA has had problems with Discovery's tank sensors before during an April test on a completely different external tank, an anomaly which engineers still cannot fully explain.

With a second occurrence, which cropped up long after Discovery's external tank was filled with fuel, shuttle engineers hope to put the glitch to rest.

"This is an intermittent problem," Griffin said in the later briefing. "When we can explain it, we will."

The fault, shuttle officials said, could lie anywhere between the sensor electronics box aboard Discovery, which processes the data, and the actual tank-bound sensor itself.

"It could be an open circuit beyond the box in the external tank," said Steve Poulos, orbiter project manager, during a press briefing on the failed launch.

Launch director Michael Leinbach said pad engineers will shroud Discovery with its protective rotating service structure late tonight or early Thursday morning to gain access to the orbiter.

Safeguarding Discovery's external tank to allow the work will take about 21 hours, meaning pad crews will be able to begin work on the shuttle by 3:00 p.m. (1900 GMT) Thursday.

Downtime for crew

Astronaut David Wolf, a veteran of three shuttle launches and three launch scrubs, said that while it's disappointing that Discovery remains Earthbound, there is one benefit from the scrubbed launch.

"Once you come down from the excited state of launch, you get relaxed and look around," Wolf told "At that time you could really use a break that you couldn't get planned."

Discovery's STS-114 crew, commanded by veteran astronaut Eileen Collins, is remaining at KSC for the time being, though could fly back to Johnson Space Center (JSC) in Houston, Texas if their mission suffers an extended delay, NASA officials said.

  • Fixing NASA: Complete Coverage of Space Shuttle Return to Flight