This is regarding: "SpaceX Chief Executive Elon Musk says a propellant dump caused the destruction of the Starship upper stage" "Musk said the failure was linked to venting liquid oxygen propellant near the end of the burn. That venting, he said, was needed only because the vehicle was not carrying any payload" https://spacenews.com/spacex-says-propellant-venting-caused-loss-of-second-starship/

enter image description here

The main issue identified so far is the engines overheating when dumped LOX mixes with unburned fuel in the running main engine's exhaust.

As Organic Marble correctly noticed, Space Shuttle routinely dumped LOX through the main engine nozzles while Orbital Maneuvering Engines (located next two main engines) were running.

I can only guess that Hydrazine burning engines will be less susceptible to overheating while their exhaust is exposed to outside LOX.

  • 1
    $\begingroup$ "The main issue identified so far is the engines overheating when dumped LOX mixes with unburned fuel in the running main engine's exhaust" In the exhaust shouldn't be a problem. But the engine bay is a dynamic environment during launch, with a lot of crazy turbulence going on there. Although once in vacuum it's different, but still likely that gases are moving in and out of that area. And it's a very hot area (from radiation) so any unburnt fuel and oxygen mixing together there will likely ignite. For the booster they have CO2 fire suppression in that area. I don't think the ship has that. $\endgroup$ Commented Jan 18 at 15:52

2 Answers 2


Note: Original Question:

Is it ever normal to vent LOX from spacecraft with active engines?

First, to establish the ground truth, on January 13th, 2024 at an address at Starbase in Texas, Elon Musk said...

So flight Flight 2 actually almost made it to orbit. So in fact ironically if it had a payload it would have made it to orbit because the reason that it actually didn't quite make it to orbit was we vented the liquid oxygen, and the liquid oxygen ultimately led to fire and an explosion, because we wanted to vent the liquid oxygen because we normally wouldn't have that liquid oxygen if we had a payload. So, ironically if it had a payload it would have reached orbit.

We can see from the second part of his statement...

we normally wouldn't have that liquid oxygen if we had a payload

... that venting O2 is "not normal".

To understand if venting O2 would have been a normal or "necessary procedure" for this Starship test, we have to engage in some speculation. We can speculate that it would be desirable for Starship to reach orbit with no excess propellant in its tanks, since it would need to be as light as possible for reentry. We can also speculate that it would be better to dump O2 while under power because dumping it after engine cutoff might destabilize the orbiting ship. In that case, the act of dumping propellant could be explained as a normal reaction to flight control projecting that Starship would reach orbit with too much propellant remaining in its tanks.

To validate this theory, we can do some napkin-math analysis of the OFT2 broadcast flight telemetry.

For example, we can base our analysis on this chart posed on Twitter/X and use it to extrapolate the velocity that Starship might have reached if it continued its burn without the LO2 dump.

Extrapolation of Flight Test 2

This extrapolation indicates that Starship was on track to reach a speed of 27700 km/h or 7694 m/s. Let's assume, for the moment, that this is accurate (it might not be).

Is this enough? To achieve, for example, a 100/250 elliptical orbit, the velocity at perigee would need to be 7893 m/s. Therefore, it appears that the version of Starship used for OFT2 was 7893-7694=199 m/s short of reaching the perigee velocity for a 100/250 initial elliptical orbit.

As a point of interest, following the initial launch burn, a fully operational Starship would need to do a circularization burn at the elliptical orbit's apogee to achieve a circular orbit at 250km. Delta-v required for that burn is about 46 m/s. Then it would need to do a similar burn to deorbit. Finally, it might need around 100 m/s of delta-v to decelerate and hover briefly for the tower catch. $199+46+46+100 = 391 m/s$. 391 m/s is a rough estimate of the improvement in delta-v needed for Starship to be able to complete, for example, a complete refilling mission with zero payload (according to these back-of-the-envelope quality calculations).

To conclude, this napkin-math analysis indicates that the version of Starship that was used for OFT2 had insufficient delta-v, even without any payload, to reach a 100/250 elliptical orbit. Therefore, it would not have made sense for Starship to dump propellant to prevent it from being too heavy during reentry.

Disclaimer: No one should place their blind trust in this analysis or in the chart that was posted on Twitter/X. Anyone who wants to have an accurate understanding of Starship's current level of progress is strongly encouraged to do their own independent analysis.


I learned of a number of situations when venting is done:

  • Rockets vent before launch to remove any excess gases or fuel vapors that may have built up in the fuel and oxidizer tanks

  • Propulsive venting in flight is used to support ullage (positioning of propellants towards the bottom of the stage in preparation for engine start/restart). (ref)

  • Rockets vent when the vehicle is in space to prevent pressure build up within the tanks due to boil off.

Second Follow-up

SpaceX issued an Feb 26, 2024 update that stated

At vehicle separation, Starship’s upper stage successfully lit all six Raptor engines and flew a normal ascent until approximately seven minutes into the flight, when a planned vent of excess liquid oxygen propellant began. Additional propellant had been loaded on the spacecraft before launch in order to gather data representative of future payload deploy missions and needed to be disposed of prior to reentry to meet required propellant mass targets at splashdown.

A leak in the aft section of the spacecraft that developed when the liquid oxygen vent was initiated resulted in a combustion event and subsequent fires that led to a loss of communication between the spacecraft’s flight computers. This resulted in a commanded shut down of all six engines prior to completion of the ascent burn, followed by the Autonomous Flight Safety System detecting a mission rule violation and activating the flight termination system, leading to vehicle breakup. The flight test’s conclusion came when the spacecraft was as at an altitude of ~150 km and a velocity of ~24,000 km/h, becoming the first Starship to reach outer space.

Additionally, on Feb 26th the Federal Aviation Administration (FAA) closed its investigation into the second Starship/Super Heavy launch in November. The FAA stated in its letter to SpaceX several corrective actions including

...operational changes eliminating pre-second engine cutoff propellant dumps...

  • $\begingroup$ Thank you, Phil. You've touched on a point that has bothered me for a long time. Was Tsiolkovsky wrong about the 3-stage rocket design? Is the two-stage Starship design awfully underpowered? $\endgroup$ Commented Jan 15 at 1:27
  • $\begingroup$ But this question is about dumping oxidizer next to burning engines. Is it ever done? $\endgroup$ Commented Jan 15 at 1:33
  • 1
    $\begingroup$ I'm not convinced that it would be a fire hazard, since you need both fuel and oxygen to create a fire or an explosion. Dumping while thrusting could still be dangerous, especially with an experimental rocket, for a different reason. The pressure inside the tanks is needed to maintain structural integrity. If the LO2 dump led to too much loss of pressure, then the rocket could buckle. Venting would be done to prevent boiloff from causing too much pressure to build up inside the rocket, but I doubt that this would be done by design during launch as such a process would be wasteful. $\endgroup$
    – phil1008
    Commented Jan 15 at 2:49
  • 2
    $\begingroup$ phil1008 -"you need both fuel and oxygen to create a fire or an explosion", you need heat, fuel, and oxygen. Plenty of heat in the engine compartment, as well as possibly some unburnt fuel. LOX making it into that area might cause higher temperatures than designed for. I think this may have been what Elon was alluding to even though he didn't give much in the way of details. $\endgroup$ Commented Jan 15 at 4:17
  • 6
    $\begingroup$ A very well-respected scientist gave some sage advice during an ISDC keynote a couple of years ago. He said, "What Elon says is only interesting to you, in that it gives you a signal, that there is a reason why Elon thinks my believing this might be useful to him. What might that be? The content of the statement itself is utterly irrelevant for figuring out what Elon might actually be up to." $\endgroup$
    – phil1008
    Commented Jan 15 at 7:21

Until the OMS-1 burn was made unnecessary by tweaking the MECO targets, the Space Shuttle did this all the time. Residual LOX was dumped during the OMS-1 burn.

LO2 propellant residuals are dumped through the SSMEs during OMS-1 burn in order to take advantage of the available settling thrust, and to reduce the OMS propellant requirements by providing useful thrust/impulse to the Orbiter.

See this answer for more on the Space Shuttle LOX dump.

From Integrated Main Propulsion System Final Report, STS-51B, page 17

  • 1
    $\begingroup$ Thank you. This is a very legitimate point. But in terms of apples and oranges, I am not sure if we can conclusively compare LOX dump during main engines burning (Starship) vs small Orbital Maneuvering engines burning (Shuttle). AJ10 (OMS) - Hydrazine burning engine en.wikipedia.org/wiki/AJ10. $\endgroup$ Commented Jan 18 at 0:05

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.