The answer I have received from Fluke in case anyone interested:
The DC voltage function is a single 600.0V range specified at ±(1% of reading + 5 counts).
1% of 1.5 V is below the resolution of the range, leaving only 5 counts which is equal to ±0.5. ie the DC voltage function is accurate to ±0.5V
One possible explanation is that the Range has a small negative offset, Does the meter read 0.0 when the test leads are shorted ? (it does)
The instrument is reading within the published specification (±0.5V at 1.5V), there are no user adjustments to remove an offset.