From my experience with lift station pumps there are a number of possibilities. Like any forensic evaluation, you need to determine the first point of failure, and keep in mind that does not mean it is the ONLY point of failure.
Contrary to what was implied above, contactors are specifically designed to interrupt locked rotor current for their rated load, so don't start there. This eliminates a jammed pump because if that happened, the OL would trip and the contactor would open normally; that's what it is designed to do! Turning on into a back-spinning load may damage the pump, but the motor cannot pull more than locked rotor current, so the above still applies.
Contacts weld for a number of reasons, but NEW contacts weld for far fewer. Here is my short list:
1) Lightning. You would probably know if that was the cause from other obvious evidence however.
2) The contactor was seriously undersized. This has an entire new list of subcategories, one of which was already mentioned, i.e. wrong voltage? If you supplied a 480V rated contactor and they connected to 208 or 230V, the contactor was grossly undersized. The motor would not have developed sufficient torque to accelerate however, so if it were a centrifugal pump it would not have any flow, hence low current as well and it would have just overloaded. So the welded contacts would only be caused by this if it is a PD lift pump, which is rare.
3) Rapid cycling. If the start/stop control system was not properly set or designed, the starter could have been rapidly cycling on and off, even chattering. This welds contacts faster than anything else normal. Check how they are calling for the pump. If they are relying only on one float, it MUST have a hysteresis in its operation; a significant difference in where the contacts close to call for the pump and where they open again to stop it. Otherwise, turbulence in the wet well will chatter the contactor. If they have 2 floats, a start and a stop, there must still be sufficient distance between them to prevent this from happening.
4) A mechanical jam of the contactor assembly itself. Debris from installation (which explains why it worked before it left) such as drill tailings or wire insulation ends from stripping, can fall into the armature of the contactor and jam into it, preventing it from opening. If that happened, the well would run dry, the motor would eventually seize and the overload relay would trip long after the damage was done.
5) Attempted manual operation. A lot of contactors operate on a horizontal plane, so there is a plunger-like indicator on the front that sucks in when the contactor coil is energized. A LOT of inexperienced people will check for rotation by thinking of this as a "manual override" button for the control circuit. What they don't realize is that on something as big as a 60HP contactor, a human finger cannot exert enough pressure on that "go button" to overcome the magnetic forces that are trying to push the contacts apart. The result is, the contacts are chattering, even if not audibly or visually, and they weld VERY rapidly. They realize right away what they have done, but they NEVER admit it!
If I were to bet, I'd go for #4 or 5.
