Question-and-Answer Resource for the Building Energy Modeling Community
 | 1 | initial version |
Radiance uses a stochastic algorithm to calculate the light transport, which will result in slightly different illuminance values from run to run. The way to get more reliable results is to increase the sampling density -- in terms of number of sampling points and number of rays spawned from each point -- as well as the number of "bounces" followed for the ambient, or global illumination. In OpenStudio under the Radiance Parameters I recommend you select the "Fine" preset, which will increase many of these parameters for you in a coordinated way: 
These settings should give you more consistent results from run to run, though there will still be some variability in the illuminance results. These results directly inform the lighting load schedule, which presumably directly relates to your Zone Lights Total Heating Energy.
Lastly I have to admit to never having used PAT and Radiance much beyond some simple workflow testing. We'll need to take a look and make sure there isn't some other issue at work here; the strange result in your progressive analysis makes me think perhaps there's something else going on here. Care to share your model? I could take a look.
