Hi Ronald,
I think these results have to do with the data in your weather file.
Here are results from a parametric simulation I ran using the "S49 BiFa" case in your file with a weather file for a location in Alaska with a similar latitude to your weather file. It shows a similar bifacial gain vs tilt angle for azimuth = 90 and azimuth = 270:
Attachment bifacial-gain-vs-tilt-az-high-latitude-tmy2.png not found
The results with your EPW file confirm your observation below:
Attachment bifacial-gain-vs-tilt-az-high-latitude-epw.png not found
When I compare the DNI data for your file with the Alaska file, I see that the DNI data shows a tendency to be higher in the morning hours. I think that is consistent with the results showing that the irradiance on the back side of a west-facing array would increase more with the array tilt angle than for an east-facing array:
Attachment dni-daily-average-profiles-epw.png not found
The Alaska data does not have that pattern:
Attachment dni-daily-average-profiles-tmy2.png not found
Best regards,
Paul.