HB Solar radiation analysis for part of year


I have a question about the different skies used for HB solar radiation simulations and different values I’ve been obtaining using two different methods to simulate similar things.

I’m trying to use HB to simulate solar radiation entering a building through glazing for different time periods:

  1. for all sunlit hours in a single day, and
  2. at each hour on the same days.

For 1. I used the genCumSky component, while for 2. I used the genClimateBasedSky.

Summing hourly results from CB skies, I found that the radiation results from genCumSky are lower than CB sky results. In a test scenario with a 1x1m unobstructed horizontal surface, CB skies for 21 Dec from 12-15 produces ~3800 Wh/m2, while the results for genCumSky were ~2000 Wh/m2. I’ve attached my test recipe and the weather file I’m using. The direct normal radiation from the weather file in that period is ~1000 Wh at each hour, and the sun almost directly overhead, so it seems the CB sky is reasonable

Is this discrepancy due to the way genCumSky and gendaylit models the sky with Trazenga and Perez models?

Or am I doing something wrong to generate different results?

Is there another way to simulate solar radiation through glazing by generating a simulation sky by accumulating sky radiation from various parts of the sky?

Thanks for your help!

sunlight skies comparison.gh (515.3 KB)
Par.zip (159.3 KB)