Falcon 9’s return to flight reminds spaceflight industry of past anomalies

by Aaron McCrea

Nearly every rocket that has flown has had an anomaly. With thousands of components that all need to work in perfect harmony to have a successful mission, there is likely to be one component that is unable to perform every once in a while.

While witnessing an anomaly in spaceflight is not as improbable as it may seem, it is becoming unusual to see with the cadence of flights increasing. It can be almost expected that the first launch of a new rocket will end up with a failure somewhere in flight. However, when a rocket that has been rapidly flying without any issues for over seven years is not able to deliver its payload, the spaceflight industry takes note. 

On Thursday, July 11, at 10:35 PM EDT (Friday, July 12, at 2:35 UTC), Falcon 9 was grounded after a failure on the upper stage caused the loss of the Merlin Vacuum (MVac) engine. The anomaly also resulted in the loss of the Starlink satellites onboard after they fell short of reaching a stable orbit.

Falcon 9 has been an extremely reliable launch vehicle, having only four anomalies to date. Most of these failures occurred while it was in the early stages of the Block 2 and Block 3 vehicle design. Falcon 9 Block 5 has had 297 successful launches until Saturday’s 298th launch, making it one of the most reliable rockets ever to fly.

Falcon 9’s anomaly history

Falcon 9’s first in-flight failure happened on June 28, 2015. Following a nominal liftoff for Falcon 9 Block 2 on the CRS-7 mission, the second-stage liquid oxygen (LOX) tank was overpressurized and ruptured 139 seconds into flight. The tank’s rupture led to the rocket being destroyed 11 seconds later. This mission was planned to resupply the International Space Station using SpaceX’s uncrewed Dragon capsule, which was separated from the rocket before it crashed into the ocean and was lost. 

Launch of CRS-7. (Credit: NASA/Charles Babir)

This failure of the vehicle and NASA payload led to NASA’s involvement in the investigation. It was found that a strut inside the second stage LOX tank failed to manage the forces it was expected to withstand. SpaceX solved this issue by replacing the faulty strut with a more robust version and by requiring more audits to be done on the quality of the vehicle before leaving for flight. After the failure, Falcon 9 had a 95 percent success rate, and it took nearly six months for the next mission to fly on Dec. 21, 2015.

After the failure of CRS-7, routine launches returned, and landings began to succeed for SpaceX. The first successful landing of an orbital rocket took place on the next launch when booster B1019 landed back at SpaceX’s Landing Zone 1. Then, four launches after the failure, the first successful booster landing atop droneship Of Course I Still Love You occurred. SpaceX was gathering more launch and landing successes until September 1, 2016.

Nine months after CRS-7, Falcon 9 experienced another anomaly. AMOS-6 was planned to be SpaceX’s 29th Falcon 9 launch. During propellant loading for a preflight static fire test, a liner inside the composite overwrapped pressure vessel (COPV) tank buckled, causing propellant to collect underneath the liner, igniting a spark caused by friction. This created an explosion that destroyed the entire vehicle. The AMOS-6 anomaly brought Falcon 9’s reliability down to just under 90 percent. On Jan. 14, 2017, SpaceX returned to flight just over four months after AMOS-6. 

The AMOS-6 static fire anomaly. (Credit: SpaceX)

Recent anomalies in the spaceflight industry

Early into the development of a brand new rocket and even new iterations of operating rockets, it is almost expected to witness an anomaly. Many up-and-coming startups and even government agencies have had failures come on the first few launches of a brand new system. SpaceX has even begun to take the approach of “fail fast but iterate faster” for rapid design acceleration. No rocket has as much of this idea behind it as SpaceX’s Starship.

Every single one of Starship’s test flights, from Starhopper to the high-altitude suborbital Ship hops to today’s near-orbital flights, has come at a rapid pace. For example, it took five attempts for teams to learn how to land the Ship. However, as the flight tests have progressed, SpaceX has learned from mistakes and has rapidly improved the vehicle in ways that would have never been foreseen if it was just a blueprint. Today, Starship is not a reliable vehicle, but every integrated flight test has completed more of the system’s requirements and succeeded in more and more of SpaceX’s goals. It might not be much longer before Starship is seen as a reliable satellite launcher. 

Starship’s maiden flight. (Credit: Max Evans for NSF)

There are many examples of startup space companies working to develop a rocket to get to the success of a profitable aerospace company. A company that is currently in that process is Firefly Aerospace with its Alpha rocket. Alpha has completed five flights and does not have the best track record. There has been a failure on the first flight, a partial failure on the second and fourth flight, and two successes on the third and fifth flight. Holding strong to the recent success, Firefly looks to continue with another clean launch in October later this year. If all goes well, Firefly may have already passed the inflection point of success.

The European Space Agency (ESA) is the most recent government agency to build and launch a brand new rocket with Ariane 6. The inaugural flight of Ariane 6 went off without a hitch until an hour and 14 minutes into the flight. Ariane 6 successfully entered a circular orbit and deployed a majority of the 11 payloads on board. Then, an auxiliary propulsion system (APU) malfunctioned, causing the planned third deorbit burn to fail. Even with many years and billions of dollars put into development, it is common for one small anomaly to cause the entire mission to end up as a partial failure.

Ariane 6’s inaugural launch (Credit: ESA)

Another government agency that has gone through this problem and ultimately succeeded is the Japanese Aerospace Exploration Agency (JAXA), Japan’s space agency. The agency has had a few issues with its new H3 rocket that debuted last year. After the first test flight of TF1 in March 2023 failed due to the second stage engine failing to ignite, it took JAXA just under a full year to prepare for H3’s next flight. On the second flight, H3 successfully reached the planned Sun-synchronous orbit and deployed its satellites. Then, just over four months later, the first-ever designated flight was a complete success.

Rocket Lab follows a similar path to the other aforementioned private companies. On Rocket Lab’s first flight of Electron, it successfully survived stage separation and faring separation. However, due to a software failure that caused the telemetry feed to be lost, the flight termination system was activated — ending the mission. After this launch, Electron successfully flew until its 13th flight. This flight failed during the second stage burn due to a faulty wire that shut down the electric connection to the engine’s turbopumps.

Visible sparks at the point of anomaly during Rocket Lab’s 41st launch. (Credit: Rocket Lab)

Then, on Electron’s 20th mission, the second stage shut down early due to a fault with the igniter. The igniter failure caused the thrust vector control system to malfunction, straying the rocket off its planned trajectory. 20 more missions went by without any problems until Electron’s 41st mission. Electron was unable to reach orbit due to an arc that induced a short, causing the upper stage Rutherford engine to lose thrust. After the failure, it took Rocket Lab just under three months to get Electron back in the sky. Rocket Lab is now up to 50 launches and has a success rate of 92 percent. 

Blue Origin’s New Shepard suborbital rocket has launched successfully 24 times. However, on the NS-23 mission, a failure was caused by the BE-3 main engine, which activated the launch escape system on the capsule. Blue Origin’s booster, Tail 3, which was flying on its ninth mission, structurally failed due to the temperature increases of the BE-3PM engine that was not accounted for on the booster. Fortunately, there were no crew members and only commercial payloads onboard the capsule. The launch escape system worked as intended and brought the payloads back to Earth safely. New Shepard has gained a success rate of 96 percent over 25 missions with only a single failure. It would take over 15 months for New Shepard to return to flight after NS-23 halted Blue Origin’s operations.

New Shepard anomaly on the NS-23 mission. (Credit: Blue Origin)

Falcon 9 returns to flight

Falcon 9 returned from its short hiatus with the Starlink Group 10-9 mission. Launch occurred from LC-39A at the Kennedy Space Center on Saturday, July 27, at 1:45 AM EDT (5:45 UTC). The FAA announced on July 25 that no public safety issues were involved in the anomaly. This does not mean that the investigation is over, but it does mean that SpaceX can resume Falcon 9 launches.

SpaceX revealed that the issue was most likely caused by a crack that developed in a “sense line” during the first second stage burn. The sense line is responsible for checking the pressure in the second stage LOX tank. When the second stage burn began, the MVac engine had too much cryogenic LOX flowing through it and had become overcooled so much so that the engine was overpressurized. This likely led to an explosion, ending the burn and causing the loss of attitude control. SpaceX mentioned that the sensor connected to the sense line is not used by the flight safety system and other systems make the sensor redundant. Therefore, SpaceX’s plan moving forward is to remove the failed sense line and sensor from the vehicle.

It is an anomaly in itself that SpaceX can get Falcon 9 back online in less than 14 days after a second stage failure. With 355 missions and only four failures over 14 years of operation, Falcon 9 is one of the most reliable rockets to ever fly with a 99 percent success rate.

 (Lead image: Falcon 9 launch from SLC-40 at the Cape Canaveral Space Force Station. Credit: Max Evans for NSF)

Related Articles