Starliner is going to the International Space Station next time, but it is not a delivery run, it is more a hardware interrogation. This reached a breaking point following a crewed test which ceased to resemble a standard checkout and began to resemble a risk management boundary case. In June 2024, the CST-100 Starliner, the spaceship made by Boeing, was flown with astronauts Butch Wilmore and Suni Williams to ISS in what was planned as an eight-day mission. Helium leakages and thruster issues piled up until NASA decided to take the safe option namely: leave the spaceship at dock, take the capsule back to Earth, and relax the program schedule about whatever the data was attempting to tell them.
Now that unwind is being reflected in the paperwork as well as it is reflected in test plans. Boeing and NASA agreed to change the Commercial Crew contract between the two, such that the next flight, Starliner-1, becomes an uncrewed cargo flight with no earlier than April 2026. This final order consisted of four instead of six missions, two additional missions remain as options and the total contract price reduced to 3.732 billion by 768 million. “This modification allows NASA and Boeing to focus on safely certifying the system in 2026,” said Steve Stich, manager of NASA’s Commercial Crew Program.
In the case of the engineers, the attraction and the uneasiness lie within the service module. The architecture of Starliner propulsion has maneuvering authority clustered: 28 reaction control system thrusters to point and translate, 20 larger OMAC thrusters to maneuver the orbital path, and 4 RS-88 launch abort engines, which never have a second opportunity to start. The crew module has its own RCS thrusters to control the vehicle in reentry, but the crucial work of the mission, such as rendezvous, docking, deorbit targeting, is dependent on plumbing and valves in the service module which cannot be inspected once it has been discarded.
The Crew Flight Test identified two problems interacting in uncomfortable manners. Five of the 28 service-module RCS thrusters failed during the approach to station; four of them came back after reset. Ground work followed up the immediate mechanism to an overheating which disfigured small Teflon poppet seals within thruster valves, limiting propellant flow and cutting off the thrust. White Sands testing demonstrated that the seals were recoverable in case of a cooled situation, and the vulnerability was evident in case of thermal conditions and firing patterns coinciding. Simultaneously Starliner suffered five helium leakages in the service module pressurization system, one of them noticed prior to launch and four of them revealed in orbit, dwindling the buffer that keeps propellants flowing steadily under pressure.
The latter pairing contributes to the explanation of the need to frame Starliner-1 as not a celebration, but validation. NASA and Boeing have been constructing test campaigns based on thermal cycling, integrated firings, and leak-rate characterization, such as one at White Sands Test Facility that tests the thrusters in a single “doghouse” to test detailed thermal models. The described list of mitigations is publicly stated and encompasses the addition of thermal barriers to the doghouse, changing the profiles of pulse thrusters to prevent overheating, and considering new seals with helium that should resist leaks.
In the background of the propulsion description is a more lengthy motif: Starliner has found verification lapses too many times, when it is too late to make changes cheaply and when the time schedule is no longer elastic. The 2019 Orbital Flight Test-1 software issues revealed a non-end-to-end combined avionics and software tests; subsequently corroded oxidizer valves required a redesign; and parachute link-strength issues and flammable taped wiring harnesses were found during inspection. There are warnings of “process escapes” in test and verification issued by NASA on many occasions, and what occurred with Starliner proved to be the propulsion surprises, which fit that pattern.
This is not just a postmortem of a single vehicle due to the realities of NASA operation. This agency continues to desire “dissimilar redundancy” in order to ensure the US does not have functional dependence on access to ISS crew by a single family of spacecraft. This station also is on the road to retirement about 2030 with a finite number of crew-rotation slots in which Starliner can demonstrate reliability after certification. The comparator set by the record of Crew Dragon: 11 successful NASA crew flights since 2020, with no major propulsion problems, is clear, and Starliner has an expendable service module, so every flight establishes evidence which would otherwise focus the postflight diagnosis.
Starliner-1, in its turn, is the NASA project of making the next flight a controlled experiment: will the changes made to the thruster assemblies, and the changes introduced to the helium-systems work in the orbit as ground data claims it will? Until that solution comes, Starliner will be a show that is less about its capsule and more about the temperatures and the pressure within its valves that were never intended to be the tale.
