Question-and-Answer Resource for the Building Energy Modeling Community
 | 1 | initial version |
The difference will depend a lot on the geometry of your model and whether you are defining view factors between surfaces when using ScriptF.
The two methods should be identical (or close to identical) for six sided rectangular prisms. If you are not defining view factors the biggest difference between the default view factors and what is assumed in the CarrollMRT method is that ScriptF will assume that surfaces facing the same direction (or within 10 degrees of the same direction) cannot see each other. With the Carroll method, surfaces can exchange radiant heat with other surfaces facing the same direction through the mean radiant temperature node. Differences in results will be exacerbated when surfaces facing the same direction are very different temperatures.
If you can share a visualization of your geometry and/or hourly time step data, that might help better understanding why there you are seeing differences in your specific case.
Hope that helps!
