Large discrepancy of daylight and energy results between new and old versions

Hi Mostapha, Chris and all,

I tried to run both daylight and energy simulations using a simple geometry. But I found the there is a huge difference between the results of the two different version of HB (0.056 and 0.058).

It looks really wired. Because I used exactly the same geometry, materials, and other settings, but the results are quite different. For instance, the DLA results in the two versions have two times difference, and the energy load (cooling load and heating load) are also different.

Could you please so kind to check the gh files in attachments, and let me know why there exists so big difference and which one is reliable?

Thank you so much in advance!

Best,

Ding

OLD VERSION DLA RESULTS

NEW VERSION DLA RESULTS

OLD VERSION ENERGY RESULTS

NEW VERSION ENERGY RESULTS

Attachments:

CASE_1(newversion)_box.gh, 220 KB
CASE_1(oldversion)_box.gh, 208 KB

Replies are closed for this discussion.

Replies to This Discussion

Permalink Reply by Mostapha Sadeghipour Roudsari on January 14, 2016 at 11:55pm

Hi Ding, We are working on a new release for the 24th. I hope someone can check the files and get back to you sooner but in case that didn't happen I will check the file after the release.

Permalink Reply by Ding on January 15, 2016 at 5:33am

Hi Mostahpa, thank you so much for your reply.

By further checking all the illuminance values over the year at all the analysis points, I found that illuminance values are kind of the same between two versions (i.e. difference is limited in a acceptable range).

But when I further check/verify the calculation of DLA, UDLI, LightControlProfile (Read Annual Result 1) by using gh native component based on the above mentioned illuminance values, I found that the old version (VER 0.0.56) is reasonably right (because good correlation between calculated and verified values exist); but in the new version (VER 0.0.58), there is not any correlations between calculated and verified values.

I assume it might be good for you to have a look at the comparison of the calculated and verified values before your coming new release, in order to avoid the potential mistakes occurred in the Read Annual Result 1 component again.

Ps: Please use the attached files in this reply post. I have added the "let the ladybug/honeybee fly" components in the canvas, which was forgot in my previous post. The simulations will automatically run when you open these files.

Thank you once again.

Best,

Ding

Attachments:

CASE_1(newversion)_box1.gh, 648 KB
CASE_1(oldversion)_box1.gh, 610 KB

Permalink Reply by Mostapha Sadeghipour Roudsari on January 15, 2016 at 12:43pm

Hi Ding,

The only change between the two releases is the profile. Have you tried to check the occupancy profile output from lightControlProfile? I suspect that the issue is there. I will take a look at your file and let you know about the results.

Mostapha

Permalink Reply by Mostapha Sadeghipour Roudsari on January 15, 2016 at 1:02pm

That was it. The issue was with the input type of Occupancy Generator was set to float and not int. Daysim was generating wrong occupancy schedules from the csv file. It should be fixed now. Check the attached file which is updated to the latest version. I didn't check the Energy simulation components.

Attachments: