See picture at post 101. It appears that the arch is the middle of the span bearing on dual concrete columns tied together with concrete with a half arch on either side from the piles.
The picture shows at least one set of piles in the water. Scour could be a factor here.
If one set of piles went under due to scour, the bridge would tend to tear away, bringing other sections down with it. The wreckage shows that different things were happening in different sections. Some came down straight. Other sections tore and collapsed on their sides.
A failure in one of the beams at the top of the bridge would not have had this kind of wreckage as a result, due to redundancy of design and safety factors in loading.
This kind of catatrosphic failure indicates a cause at the lowest part of the bridge—in the very foundation of it.
Too early to tell—much investigation has to be done.
The details will take months or even years, but eventually, it will agree with what was figured out within the first day or two.
In that case, a structual failure due to metal fatigue and corrosion caused by the harsh Minnesota weather.
Failure of DOT inspectors to identify the structual problems which caused this specific failure.
Lack of State and Federal funding, to correct the problems which were already identified.