Question-and-Answer Resource for the Building Energy Modeling Community
Get started with the Help page
Ask Your Question

OpenStudio seems to be wrong in daylight calculations

asked 2015-02-14 05:19:35 -0500

Dinosaver's avatar

updated 2015-02-15 14:20:06 -0500

If you create a building orient strongly along coordinate system in SketchUp and do daylight analysis, you'll get a correct illuminance map.

image description

2 same windows produce the same picture. If you turn the building a little bit, the map will change dramatically:

image description

To understand why it is, you can edit an illuminance map and turn it manually back to be parallel to SketchUp axes. The map is the same:

image description

You can find the original model here Please clarify this.

Update (moved from answers)

Things getting interesting. Simple model which produces weird results for daylight :)

image description

each spot is duplicated by its previous position (one hour ago). Any idea?

edit retag flag offensive close merge delete


How are your rotating your building? Illuminance maps can only be rotated using the building's rotation and the space's rotation. If you close your OSM, then re-open it, SketchUp will show the illuminance map as it is being simulated in E+.

macumber's avatar macumber  ( 2015-02-14 20:33:44 -0500 )edit

Also, are you using E+ or radiance as the daylighting simulation engine?

macumber's avatar macumber  ( 2015-02-14 20:34:24 -0500 )edit

@Dinosaver, I apologize, I was wrong earlier. OS illuminance maps do have fields that represent rotation, but this rotation is not translated to E+. To see how your illuminance map will look in E+ you will have to set all the illuminance map rotation fields to 0.

macumber's avatar macumber  ( 2015-02-15 13:57:26 -0500 )edit

@Dinosaver - please be mindful of where you are posting. This is not a forum style site, it's a Q&A site. Your posts below are in the answers section but should either be added here as a comment or added above using edit. Thanks.

MatthewSteen's avatar MatthewSteen  ( 2015-02-15 14:16:22 -0500 )edit

I'm sorry, but it is already a little mess... @macumber If you look at the illuminance for the right zone after 1 p.m. appr when there is a direct light, you'll see that the map contains not 2 sharp spots for each window, but also 2 minor spots which are placed exactly where 2 major ones were 1 hour before. I've checked E+ csv - it produces the same.

Dinosaver's avatar Dinosaver  ( 2015-02-15 20:53:14 -0500 )edit

1 Answer

Sort by ยป oldest newest most voted

answered 2015-02-15 04:21:31 -0500

Dinosaver's avatar

updated 2015-02-18 03:18:29 -0500

I cannot do Radiance simulation, it starts, takes a lot of time and then says "annual simulation failed" although I asked for one day simulation only :(

How to run Radiance in OpenStudio see here. I believe, it should be recorded as an issue, because we do scan for tools, but responsible variable in the Ruby file are not substituted.

The second issue is the disk space. SQL from Radiance takes A LOT of space on your C drive in the /users/username/AppData/Local/Temp/Openstudio*/ results folder. Be sure you have apr 5-10 Gb for each map in a real model. The temporal solution could be reassigning the path to the results folder on another drive with enough space. Is it possible?

I have played with rotation all ways. Actually, I mentioned this when I analyzed a script-generated building and it looked correct all the time but illuminance was strange, so I created this primitive model. I know, that the map is oriented along zone's coordinates (btw, why there is a rotation angle parameter in OS?). It is very confusing when we have a curved facade and cannot model daylight just along it. In E+ we can create a map bigger than the zone to cover its shape. The results for points outside the zone are marked by * in the table, so we can filter them and exclude from visualization. Why we see them in Results viewer in OS? How can we edit the sql file to exclude those points?

edit flag offensive delete link more


Annual simulation is the only analysis period supported on the Radiance side, sorry. We will add better warnings for this. We are also adding support for arbitrary illuminance map shapes, which will work much better for your curved facade scenario, when using Radiance.

rpg777's avatar rpg777  ( 2015-02-15 10:52:41 -0500 )edit

The use of the three phase method in OS radiance simulations means annual simulations do not take much longer than single day simulations. However, they do generate much more data. Given that, is it worthwhile to support shorter run periods just to save on disk space?

macumber's avatar macumber  ( 2015-02-15 11:44:45 -0500 )edit

How much disc space do I need to run those models in Radiance? I have 8Gb RAM and as I tracked it, Radiance didn't exceed it. Win7x64

Dinosaver's avatar Dinosaver  ( 2015-02-15 12:05:31 -0500 )edit

Also, if you are interested we will be starting beta-testing for OS 1.7.0 in the next few weeks. There are some significant improvements to the radiance simulation capabilities and stability coming in 1.7.0. If you are interested, sign up for the developer mailing list to get news about beta-testing.

macumber's avatar macumber  ( 2015-02-15 12:13:43 -0500 )edit

Thank you for the invitation, I'll think about that. I really appreciate what you guys 're doing, but my first need now is integrate energy modelling into our entire process. I'm tired from receiving glazed south facades and whatever... :) I used to use commercial software but now I need smthg intuitive, reliable and impressive (and free :)) to show my architects that it is not so hard to think before doing :)

Dinosaver's avatar Dinosaver  ( 2015-02-15 12:32:23 -0500 )edit

Your Answer

Please start posting anonymously - your entry will be published after you log in or create a new account.

Add Answer

Question Tools

1 follower


Asked: 2015-02-14 05:19:35 -0500

Seen: 421 times

Last updated: Feb 18 '15