Verifying Fatigue Stresses in Leap Bridge Steel


I have a bridge model that is consistently telling me there are fatigue stress issues with my plate girder section.

I have exported the live load range that the model has created for a given girder and while comparing it to the section modulus at the locations reported to have issues I am coming up with a stress that is significantly less that what is reported. 

When I use just the bottom fiber Sx for the areas, I compute the stress for point 42.18 as follows:

When I add in the load factor of 1.75 x 2 ksi I get the stress to be 3.52 ksi at point 42.18 for example. The fatigue report is telling me this stress is being computed for the 12 ksi allowed is 1.75 x 6.089 = 10.66 ksi.  This leads to the the report of 88% of capacity being returned.

I am not sure how to reconcile these stresses, the model I have is a bridge that is skewed in excess of 20 degrees, so perhaps there is a deflection base later flange stress being considered, but the reports for this in the analysis do not return a lateral bending moment in the flange.  If anyone can provide some direction it would be greatly appreciated, thank you in advance.