Shuttle launch called off due to faulty fuel tank sensor
Wednesday, July 13, 2005
The U.S. space agency NASA called off the launch of Space shuttle Discovery today after a problem with a fuel sensor in the external tank used to detect fuel exhaustion. According to the agency-run NASA TV, the low-fuel sensor was either malfunctioning or damaged. The launch was already facing the threat of a scrub due to thunderstorms in the area.
The sensor is one of four used to trigger the engine cutout after launch. Although only two are required for normal operation, and the Shuttle can be flown with one, NASA elected to maintain full redundancy. Should more of the sensors fail, the engine might burn out due to lack of fuel, a situation that has not been tested.
The problem was detected during a simulation of an empty tank. When placed in a mode simulating an empty tank, three of the sensors correctly registered that the tank was empty, while the faulty sensor stayed in the “full” state. NASA is currently unsure whether the problem relates to the sensor, the instrumentation circuits reporting the sensor’s state, or the simulation circuits.
The problem comes after a separate incident yesterday when a cockpit window cover fell from the Orbiter, damaging thermal protection tiles. A similar problem caused the replacement of the fuel tank in June. NASA described the problem as an “intermittent fault”.
The launch, which was scheduled for 3:51 ET (20:51 UTC), would have been the first launch of a shuttle since Columbia‘s February 2003 crash which killed all crew members aboard.
It is still unknown to NASA officials what caused the sensor to become defective. It is also unknown at the moment whether the issue will be fixed on the launch pad, or in the Vehicle Assembly Building – in which it takes close to a full day’s time to transport a shuttle between the two areas.
In the interim, the crew of the shuttle will stay at KSC unless there are further delays, in which case the crew might be transported back to JSC in Houston for refresher training.