Which failures on STS missions were the most impactful?

Please exclude the o-ring issue on STS-51-L, as well as the tile damage caused by insulative ET foam on STS-107.
Unrelated failures that happened on the pad or during the STS-107 mission before reentry may qualify.

This question was prompted by a comment on Which STS mission had the most accidents or failures?

Bonus points for failures that would have been the most impactful, had they not been fixed mid-mission, excluding those that were documented and prepared for ahead of launch.

  • 4
    $\begingroup$ Note, it wasn't tile damage that caused the STS-107 reentry failure. It was the damaged leading edge of the wing (made out of large reinforced carbon-carbon panels) and not one of the famous silica tiles that caused the fatal accident. $\endgroup$ Commented Nov 22, 2022 at 22:47

2 Answers 2


The failures in this answer will be described at a very high level. A long write-up could be done on any of them. References are provided for further reading.

Failures Occurring Prelaunch

  • There were 5 "pad aborts" in the shuttle program - incidents when at least one of the Space Shuttle Main Engines (SSMEs) ignited for liftoff, but then were shut down due to the detection of a failure. The most dangerous happened on STS-041D. During the startup of SSME 3 (the first to start) its Main Fuel Valve moved sluggishly. When the engine controller detected this and shut the engine down, SSME 2 had begun its startup sequence and was also shut down. Hydrogen leaked through the sluggish valve on SSME 3 and caused a fire on the pad. Confusion ensued about whether it was safe to get the crew out of the vehicle.

  • On STS-061C, during External Tank propellant loading, the operators were working around a problem with a valve in the O2 loading system, and managed to inadvertently drain 14,000 lbm (6350 kg) of O2 out of the tank. The countdown proceeded to T-30 seconds at which point some slightly "off" indications in the O2 system caused the team to recycle the countdown to T-20 minutes, at which point the O2 drain was detected. If the vehicle had launched that low on O2, it could not have made it to orbit and an ascent abort would have resulted.

Failures Occurring During Ascent

  • STS-001 was impacted by several failures of preflight analysis. The overpressure wave resulting from ignition of the Solid Rocket Boosters was vastly under-predicted. The impact on the vehicle shook it to the extent that one of the struts holding a Forward Reaction Control System propellant tank buckled. Also, the base force on the vehicle caused by propulsion plume effects was under-predicted, resulting in a trajectory quite a bit higher than expected.

  • On STS-051F a design problem in turbine outlet temperature sensors caused the SSME 1 controller to shut the engine down at 343 seconds after liftoff. This was late enough in the flight that the vehicle could continue to orbit...but then the sensors on SSME 3 started to drift. Booster flight controller Jenny Howard, unsung hero of the shuttle program, diagnosed the problem and made the correct calls that enabled the crew to keep SSME 3 running till orbit.

  • On STS-079, shortly after reaching orbit, one of the 3 Auxiliary Power Units (which despite the name, were the only source of hydraulic power on the Orbiters) mysteriously shut itself off. Analysis on the ground failed to reach a consensus on the cause of the failure. Flight rules at the time called for a shortened mission due to the failure of the units, but it was decided to change the rules and allow the mission to go full duration. Entry was performed using only 2 APUs.

  • STS-093 was the ascent most like a sim. One of the SSMEs developed a hydrogen leak in its nozzle cooling circuit. Due to the control system logic used on the engine, this resulted in over-consumption of oxygen and an early depletion of the external tank. There was also an AC short during ascent which cleared, but set off multiple alarms in the crew station, and knocked out redundant controllers on the SSMEs, taking critical insight away from the controllers trying to diagnose the engine problem.

Failures Occurring On-orbit

  • On STS-002 and STS-083 one of the 3 fuel cells which generated all electrical power for the Orbiter failed. In these cases the flight rules were followed and the missions were cut short.

  • On STS-044 one of the 3 Inertial Measurement Units used in the guidance, navigation, and control system failed. Once again the flight rules were followed and the mission was cut short.

  • On STS-051 an incorrectly wired pyrotechnic system blasted the Orbiter payload bay with shrapnel when a satellite was deployed. The Orbiter sustained damage to payload bay hardware including thermal blankets, wire tray covers, the aft bulkhead, and thermal protection system tiles.

Failures Occurring on Entry

  • STS-001 was also impacted on entry by a failure of preflight analysis. The hypersonic pitch trim was incorrectly predicted. Fortunately the Orbiter's control system was able to automatically compensate for the errors.

  • Finally my vote for the closest call of all. On STS-009 two of the hydrazine-powered Auxiliary Power Units caught on fire during entry and exploded after landing on the runway. Also on this flight, one of the Orbiter's four redundant primary flight control system computers had failed early on landing day and the systems were configured to not use it. At nosewheel touchdown on landing, a second primary computer failed, and the crew worked the failure response procedure incorrectly - they did not take the prior failure into account. Had this happened in the air, a loss of vehicle control was likely.


  • $\begingroup$ Comments are not for extended discussion; this conversation has been moved to chat. $\endgroup$
    – Rory Alsop
    Commented Nov 25, 2022 at 11:12

In addition to those pointed out by @Organic Marble, STS-27 is another notable example of a near LOCV scenario. About 85 seconds into the mission, the shuttle's heat shield was struck by a piece of ablative insulation from one of the solid rocket boosters. This resulted in one tile from the heat shield becoming completely separated from the shuttle, and damaging possibly hundreds more. Replaying the launch footage, engineers on the ground noticed the debris strike, but were unsure of the extent of damage to the shuttle, and requested the crew to investigate the damage using the shuttle's robotic arm. To the crew, the severity of the damage was immediately clear. However, since the mission was carrying a classified payload, and therefore could only send footage back to the ground via a very slow, encrypted channel, the engineers were unable to get a clear picture of the damage, and as a result dismissed the crew's concerns.

Despite the crew believing that the damage was not survivable, the shuttle re-entered safely. Inspections of the shuttle on the ground confirmed the damage, and it is believed that the shuttle survived only because the missing heat shield tile was located directly under a steel mounting plate for an UHF antenna, which was able to withstand more heat during reentry than the aluminum structural components which would have been exposed in many other locations. Also worth mentioning is that fact that STS-27 was only the second shuttle mission following the Challenger accident. Had it too led to the loss of a crew and orbiter, it would have likely led to the cancellation of the shuttle program.

  • $\begingroup$ I've never noticed it before (I'm here because me & my buddies are arguing about it while speculating about what's going to happen to Starship), but that source actually says the bracket was aluminum. @organic-marble any help here? $\endgroup$ Commented Apr 15, 2023 at 17:30
  • 2
    $\begingroup$ @AntonHengst Not steel, not a bracket, and not a UHF antenna. An aluminum door over a TACAN antenna. See space.stackexchange.com/a/63337/6944 $\endgroup$ Commented Apr 16, 2023 at 2:08

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.