algorithmic modeling for Rhino
First of all: Great Work! I can't even imagine how much you have put into this wonderful project. Thank you!
I am running solar radiation simulation using ladybug and ecotect and I am getting slightly different results. I am calculating average daily solar radiation for entire year for a massing. The ecotect shows flat roofs to be exposed to over 2200 Wh/m2 while ladybug max daily average is about 1700Wh/m2
In ladybug, i just divide the total by 365 days and since output is in kWh/m2, I multiply by 1000 to get Wh/m2 (using the Re-Color mesh to display final results and the new legend)
If anyone knows why, could you please point me in the right direction? My model is in meters.
Replies are closed for this discussion.
Thank you for the nice comment.
Let's make sure we are comparing similar values together. First of all make sure that you are using the same weather file for both Ecotect and Ladybug.
Moreover Ecotect's sky model is different from Ladybug's sky model (http://naturalfrequency.com/wiki/sky-subdivision). Ladybug uses Tregenza/Reinhart sky models and Radiance's gendaymtx which are both validated methods.
For what you are calculating it is easy to compare the values with the weather data. Check global Horizontal Radiation values from the weather file and compare them to your results of the simulation for a roof. They should be pretty close. The more you add to sky density the closer the values will be. Check the attached file for an example.
You are right, I am kinda comparing apples and oranges. The weather file is not the same. The project is in Guangzhou, but Ecotect only has a file for Hong Kong. It isn't that far, but still. So in Ecotect I used the weather for Hong Kong, in Ladybug I used the weather for Guangzhou. I am also not sure how Ecotect and Ladybug take samples of the surfaces and how face normals play into that. Just last week I started testing with solar shading devices and Ladybug's results tend to come out better. They look like something we might expect in real life. I am simplifying to a one bay of the office tower and I will do the tests again and will share the results.
Thanks again for responding and your help!
when you are done this study in both tools would be good if you could share your results.
Results are different at the end. I made sure to have the same weather file this time. The size of each sample I tried to make the same. In ecotect, I divided the surface into 50mm squares. In ladybug, I've setup the sample to be taken every 50mm. (Ecotect generated 17712 panels, Ladybug 17856) I ran both tests from 8am to 6pm, everyday for the entire year. If the sky is pre-calculated, ladybug is much faster in analyzing. About 3 times as fast.
Next I will do the same test, but non-shaded. I guess it would be crucial to see if the difference is the same. I am not so concerned about the actual numbers, but about percent differences.
PS I posting screenshots. If you see errors let me know.
And here are the screen-shoots for ladybug.
Ok, so there is a bit of a difference. On average values, the results are very close (6% difference), on the extremes, not so much. It looks like on the lower extreme, ladybug and ecotect have a difference of about 30%. While on the upper extreme, the difference is only about 8%. It looks to me ecotect is more conservative on average and ladybug is picking up more values on the lower extreme... way more.
Please be aware that I am running the ecotect radiation simulation using the "fast calculation" option. I am going to re-run without that checkmark and share results. But so far it takes very long time(many, many hours). I don't know if I have the patience.
So, I don't know what to do now.