The failure of a power unit on the Orion capsule slated to fly on SLS’s first test flight late in ’21 could delay that test flight by as much as an additional year.

Replacing the PDU isn’t easy. The component is difficult to reach: it’s located inside an adapter that connects Orion to its service module — a cylindrical trunk that provides support, propulsion, and power for the capsule during its trip through space. To get to the PDU, Lockheed Martin could remove the Orion crew capsule from its service module, but it’s a lengthy process that could take up to a year. As many as nine months would be needed to take the vehicle apart and put it back together again, in addition to three months for subsequent testing, according to the presentation.

Lockheed has another option, but it’s never been done before and may carry extra risks, Lockheed Martin engineers acknowledge in their presentation. To do it, engineers would have to tunnel through the adapter’s exterior by removing some of the outer panels of the adapter to get to the PDU. The panels weren’t designed to be removed this way, but this scenario may only take up to four months to complete if engineers figure out a way to do it.

A third option is that Lockheed Martin and NASA could fly the Orion capsule as is. The PDU failed in such a way that it lost redundancy within the unit, so it can still function. But at a risk-averse agency like NASA, flying a vehicle without a backup plan is not exactly an attractive option. It’s still not clear what went wrong inside the unit, which was tested before it was installed on the spacecraft, according to a person familiar with the matter.

None of these options are good. The first two will certainly delay the planned November 2021 launch, which by the way is already four years behind schedule. The third will risk a failure of the mission, which though unmanned would certainly lead to further delays in the manned mission expected one or two years later.

That they don’t know why the unit failed and cannot fix it easily speaks very badly to the design of Orion and SLS. Compare this with SpaceX, which in the past month has demonstrated it can in only days switch out engines on both its commercial Falcon 9 rocket and its new next generation Starship rocket. Moreover, SpaceX has demonstrated repeatedly that once they identify an issue they move immediately to understand it and fix it.

With NASA, Orion, and Orion’s contractor Lockheed Martin, such flexibility and agility appears all but impossible. They have designed a monster that cannot be fixed easily, cannot launched quickly, and costs an ungodly amount of money.

I increasingly believe that Starship will reach orbit before SLS/Orion, even though the latter has been in development for almost three times longer, and will cost 25 times more.