No Launch Today

Image credit: NASA/George Shelton

The caption reads: Space shuttle Atlantis stands on Launch Pad 39A at NASA's Kennedy Space Center in Florida.

NASA announced this morning that it is postponing the launch of the space shuttle Atlantis until it can sort out a problem with some of the shuttle's instruments:

NASA postponed the launch of the shuttle Atlantis to January as the agency tries to figure out why an engine-cutoff sensor failed today when the external fuel tank was being filled.

NASA Postpones Launch of Shuttle Atlantis to January (Update1)

As the Bloomberg article mentions, the problem is that a fuel sensor on the external fuel tank failed. The external tank is the big orange thing in the photo. It stores the liquid hydrogen (LH2) and liquid oxygen (LO2) that the shuttle's engines burn. The fuel sensor is part of an engine cutoff system, which is designed to stop the engine from running before it has run out of fuel. From the diagram, it appears that the sensor is designed to trip when the fuel level drops below five percent. I'm not sure what the consequences of letting the engine run out of fuel would be, but if they're going to so much trouble to avoid it, it must be pretty bad. As this press briefing explains, NASA won't launch unless all four of these sensors are operating correctly.

This isn't the first time there have been problems with this system. A report following STS-114 in November, 2005 states:

Although anomaly troubleshooting was extensive, the root cause of the failures observed during the first STS-114 tanking test and launch scrub was not identified. The most probable causes include an intermittent electrical connection within the Orbiter or ET wiring from the ECO PSB to the sensor and return, an intermittent high resistance or open circuit in the senor itself, or a thermally-induced intermittent failure internal to the PSB.

STS-114 Engine Cut-off Sensor Anomaly Technical Consultation Report (PDF)

STS-114 was flown by Discovery.

STS-115, which was also flown by Atlantis, was also delayed due to a malfunction of these sensors.

Later, NASA believed that it had isolated the problem to a manufacturing defect, and had the sensors replaced before this mission:

Engineers, last year, switched to new engine cut-off sensors and attached additional instruments to monitor their performance, only to see the issues resurface aboard Atlantis.

"I think this has been sort a cloud that has always been over us," said William Gerstenmaier, NASA's associate administrator for space operations. "We thought we had it fixed when we changed out the engine cut-off sensors before."

NASA Renews Hunt for Shuttle Fuel Sensor Glitch

These sensors must endure extremes of heat, cold, vibration, and shock and yet still operate correctly. They must also function while immersed in highly reactive chemicals. As this diagram shows, the LO2 and LH2 sensors must be coated differently to function properly. They're a challenging engineering problem.

While it's certainly the norm in software engineering, trying different solutions until one is found to work isn't unknown in other branches, either. NASA's engineers and program managers have proceeded cautiously on this issue, investigating, trying new approaches, and testing them. Eventually, a solution or design correction will be found.

These sensors are part of a backup system that is there in case the primary system for shutting off the engines fails. Postponing the mission reflects an increased attention to safety that was brought about by the Challenger and Columbia disasters. As such, it represents a promising trend at NASA. It will also be a way to learn more about engineering systems for space travel, and as long as such experience isn't bought with the needless deaths of flightcrew, that can only be a good thing, too.

UPDATE: Added a paragraph about the process of isolating the problem.