I added Beam 1234, from Nodes 34 to 91, to brace the beam in the X direction and it reduced the utilization ratio to 0.259. If you can add kickers to the others it should resolve your issues.
Is a bolted splice out of the question? You may also consider using round bar, this will make the welding easier and will be easier for QC. Maybe even a combo of web plates & round bar.
I believe it is simply due to no lateral support in the X direction. Those Beams are just relying on the torsional resistance of the Beams about the Z axis and getting thrown all around.
Not sure which beams were failing but I noticed that the highlighted beams intersect and don't share a common node, the are essential cantilevered beams according to STAAD.
I found the following in ACI 350-06:
ACI 350-06 [Appendix C-ALTERNATIVE LOAD FACTORS, STRENGTH REDUCTION FACTORS, AND DISTRIBUTION OF FLEXURAL REINFORCEMENT]
1.4D+1.7L (C-1)
1.4D+1.7L+1.7H (C-6)
1.4D+1.7L+1.7F (in RC 9.2.5)
0.9D+1.7H (in RC 9.2.4)
0.9D+1.7F (in RC 9.2.5)
ACI 350-06 [R9.2.1]
U...
