Well, an orbiting planet can possibly travel in orbit faster than a mineral packet.
Take Earth. In Aurora, its orbital distance is 150m km, meaning its orbital circumference (distance traveled every revolution) is 150m X 2PI. Its year duration is precisely 1 Earth year or 365 Earth days. So the Earth orbital speed is [ (150m X 2PI) m km / (365 X 24 X 3600) sec], which is equal to 29. 9 km/s. This matches real world, as expected.
If you had a spaceship, such as an orbital habitat with engines, traveling at less than 29. 9 km/s, it could possibly be unable to reach Earth except with an intercepting course.
Mineral packets travel just like spaceships, always straight to their target. If said target moves, the packet adjusts its bearing accordingly.
Also, mineral packets always travel at 1000 km/s when fully loaded. If partly loaded, they travel faster following some equation close to E=mv^2, with E the fixed energy installed mass drivers give to objects (might not be linear).
With a planet far enough and having a year short enough, it should be possible that its speed is above 1000 km/s. It can even be a binary system with the 2nd star traveling very fast, or some other nonsense.
Have you got that system data for us? I find it quite funny.
The solution would be to use way more mass drivers than necessary to send faster packets, or change its target to a more stationary "relay" body. As for the current stuck packets. . . dunno.