I cloned my design, and now some of the loss values changed, resulting in differences on the Production Report. Why is that?
This is because of mismatch values and the random values generated in the simulation, which show as shade losses on the Production Report. For module mismatch, Irradiance variation (i.e. differences in sunlight), Temperature differences, and Module nameplate power differences (i.e. module binning) are all applied statistically across the array: each module is randomly sampled from each of the defined ranges – and that module will then have those deltas (in irradiance, temperature, and nameplate power) applied for all calculations each hour.
The discrepancy between the different design versions is caused by the statistical randomization values for irradiance, temperature, and module tolerance (caused by the Irradiation Variance category in mismatch). This randomization happens when the simulation is run (statistical distributions applied module-by-module used to define module binning and other mismatch parameters). The randomization values live in the Condition Set.
There are three drivers of mismatch, each of which creates a random distribution between modules in the array (which can cause a difference in production, usually under 2% if identical designs are re-simulated or cloned):
A: Difference in POA irradiance (normally distributed)
B: Temperature differences (uniform distribution)
C: Module manufacturing tolerance (uniform distribution). Note that this parameter has two inputs to account for positive-tolerance module binning.
You can learn more about mismatch works and is calculated in HelioScope here.