Due to torsion or twisting of the supports -- presumably from the river flow -- there was a tremendous sideways spring force on the road bed. I'm guessing the bolts or welds holding the two parts of the bridge together at the south end just popped due to this sideways force. The release may have been triggered by the the construction work or the unusual loading due to lane shutdowns, but it was going to happen at some point.
The question to me is why wasn't this kind of stress picked up during any of the inspections that were held during the last decade. I presume that the deformation was so gradual that the minor cracks that would show up on the roadbed each year would just be tarred or paved over. The inspections that I heard about were looking for microscropic stress fractures, but perhaps they were only looking for fractures in the direction of the roadbed and not laterally. But you'd think if any responsible engineer saw the satellite view they'd have some serious questions.
By “different explanation”, I meant that this particular incident wasn’t due to inadequate money being spent on maintenance, but that whoever did the maintenance was perhaps performing microscopic inspections without taking a macroscopic view (as provided by the satellite photo).
For hopefully the last time, there were no pilings in the river so hence there was no side forces from river current on the bridge.
It is hard to tell, but it looks like the bend could also be a result of the camera not being straight above the bridge and the bridge having a rise (be curved vertically) over the deepest portion of the river.
Do you know if the bridge was flat or if it had a definite rise over the center of the river?