CAPE CANAVERAL, Fla. - Because of an engineering mystery, NASA will have to wait until Sunday at the earliest to try launching the space shuttle Discovery — and probably later than that, officials said Thursday.
The timing of the next launch attempt depends on what engineers find out about the fuel-tank sensor problem that forced a postponement of Discovery's launch on Wednesday.
"If we were to get extremely lucky, it is theoretically possible that we could still launch on Sunday," deputy shuttle program manager Wayne Hale said. But he added that "this represents a really optimistic, good-luck scenario, which I think is not very credible."
It's more likely that the "no earlier than" launch date would be pushed ahead, day by day, until NASA gets a better handle on the problem. Hale said NASA had hundreds of engineers around the country trying to figure out the "unexplained anomaly," but no one yet knew why the sensor wasn't working properly.
"I wish I had more answers for you,” Hale said.
Uncertainties now surround the schedule for Discovery's mission, which is to mark NASA's return to flight after more than two years of investigations, upgrades and reforms in the wake of the shuttle Columbia's catastrophic breakup in February 2003. The flight's main aims are to resupply the international space station — and more importantly, demonstrate all the safety measures that NASA has instituted since Columbia's loss.
Because of Columbia, NASA has focused intently on fully resolving any issues of concern. Those issues range from the shedding of foam insulation from the shuttle's external fuel tank, which was implicated in the Columbia accident, to the current unresolved glitch.
Intermittent glitch
The glitch involves a low-level sensor in the liquid-hydrogen compartment of Discovery's external fuel tank. During a routine test in the final hours of Wednesday's countdown, the launch team commanded the sensors to show an empty, or "dry," reading. Three of the sensors did just that, passing the test. But one flunked, continuing to show a "wet" reading.
Unfortunately, all four sensors have to pass the test, and so managers stopped the countdown and brought Discovery's crew back out of the shuttle orbiter.
On Wednesday night, the fuel tank was drained of its liquid hydrogen and liquid oxygen, and the sensors were checked again. For about three hours afterward, the troublesome sensor continued to indicate there was hydrogen left in the tank, Hale said. But then it began showing correctly that the tank was dry, he said.
"Right now, the sensor indication is as it should be," he said, "which represents something of a problem for our troubleshooting team."
Full-scale plan
The mission management team is drawing up a full-scale troubleshooting plan, which involves poring over the test data as well as the documentation for past tests and the origins of the sensor system components. The problem could lie in the sensor itself, or in the electronic circuits that interpret sensor readings, or in the connective wiring.
Hale said a Sunday launch would be possible only "if we go in and wiggle some of the wires and find a loose connection."
Mike Wetmore, NASA's space shuttle processing director, said he was waiting for the go-ahead to unload fuel-generating supplies and make the shuttle safe for detailed troubleshooting. The "break point" for getting out of the current 48-hour countdown mode would come early next week, he said.
Under some scenarios, Discovery might have to be rolled off the launch pad and into NASA's 52-story-high Vehicle Assembly Building, Hale said. That could cause the shuttle to miss the current launch opportunity, which extends until July 31. In that case, liftoff would have to be rescheduled for no earlier than Sept. 9. The timing of the launch opportunities is determined by the space station's orbital position as well as NASA's desire to launch at least the first two shuttle missions during daylight hours.
Because the troublesome sensor currently appears to be in working order, some reporters wondered whether mission managers might consider fueling up the shuttle again and just letting it launch. Hale replied rhetorically: "Could we talk ourselves into going after all of this, without doing anything?"
Why the sensors are important
The low-level sensors are supposed to signal the fuel system to get ready for shutting down the shuttle's main engines if propellant levels get too low. For that reason, they're also known as engine-cutoff sensors or ECO sensors.
If the sensor system fails, the shuttle's propulsion system might not be able to adjust the fuel system correctly during its ascent to orbit. A false "dry" reading could lead the system to shut down the engines too soon, leaving the shuttle too low to reach orbit. A false "wet" reading could keep the engines running at a high rate when they should be in shutdown mode. Tests in the early days of the shuttle program indicated that such a condition could result in "serious, uncontained damage" to the propulsion system, Hale said.
"It's a good thing to shut your engines down in a controlled manner before they actually run dry," Hale said.
Problems with the fuel-tank sensor system have bedeviled NASA for months: They first popped up while testing a different fuel tank back in April, but engineers could not reproduce the glitch or track down its source. Instead, they replaced components of the system with approved spares. The system checked out OK in May, and Discovery was mated to a newer, heater-equipped fuel tank in June.
Even though another full-scale tanking test was not conducted, managers felt reassured that the problem had gone away.
After Wednesday's launch scrub, NASA Administrator Mike Griffin said the off-and-on nature of the problem was frustrating. "This is an intermittent anomaly, and the hardest of all electronic problems is an intermittent one," he told reporters.
Weather was 'no-go'
The sensor glitch came on top of at least two other problems in the last 24 hours of the countdown, including the fall of a cockpit window cover that damaged two tiles on one of Discovery's engine pods, and the failure of a launch-pad heater for the tanking system. Those earlier problems were quickly fixed.
Hale told reporters that even if there were no sensor glitch, NASA would not have been able to launch Discovery on Wednesday because there were thunderstorms too close to Kennedy Space Center at the appointed time for liftoff.
"We would have been no-go for launch," he said.
Hale said Discovery's seven-astronaut crew, headed by commander Eileen Collins, was staying in the Cape Canaveral area, continuing their training and getting in some "relaxation time."
Mission focuses on safety
Discovery's 12-day mission is aimed at testing safety measures developed in the past two years — including the addition of cameras to monitor the launch, changes in the design of the fuel tank and the orbiter, and the use of a new 50-foot (15-meter) boom to inspect critical parts of the shuttle's protective skin while in orbit.
The shuttle is due to take tons of supplies up to the international space station, and bring tons of equipment and trash back down.
While docked to the station, Discovery's spacewalkers are to test some of the tools that have been developed to repair minor damage in the shuttle's tiles or reinforce panels. During a second spacewalk, they will replace a balky guidance gyroscope on the station — and a storage platform will be installed on the station's exterior during a third outing.
Collins has said she and her crew are planning to pay tribute to Columbia and its crew, but she has held back the details to build suspense.