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

# How far away should I model shading blocks?

I'm working on a project where we are interested in the effect of shading from nearby buildings. I plan to model these as shading blocks and am wondering how far away I need to consider.

Is there a rule of thumb or calculation I can use that takes account of latitude, bearing from the modelled building, relative height of the modelled building and shading block? Or am I overthinking this and should just model everything within say, 20, 50, or 100 m?

Pointers to any studies of this, guidance from ASHRAE/CIBSE etc, or regulatory guidance would all be useful.

Edit

For context, we're looking at hundreds of buildings so are looking for a solution which is both repeatable (doesn't involve a judgement call) and can be automated to some extent.

edit retag close merge delete

Sort by ยป oldest newest most voted

Go as far as needed until it doesn't affect your building anymore.

Couple of pointers as to where I would start:

• Model it in sketchup and enable shading so you can get visual 3D feel for what's happening.

• Use any software such as Climate Consultant to easily produce a Sun Shading Chart. It also has an "obstruction elevation" where you can enter neighboring buildings (distance, elevations, and which orientation) and it'll show you how much of your sun you're blocking.

Below I've added two obstructions, one that has no impact (100 ft high, 15 ft away, but East orientation) and one that has some impact (full south, 50ft high, 100 ft away)

If you feel like going further, it'll be a good occasion to refresh your skills in trigonometry in 3D polar coordinates.

Edit:

I just remembered that there's some Python framework called PV_LIB to model solar PV. Sure enough, there's already a function to calculate the position of the sun given location: latitude, longitude and elevation (and time zone). Check the source code for the pvl_ephemeris function here

more

The first part is a great answer if we were talking about modelling a single building, and I'll certainly use it to get a feel for the question. I'm modelling several hundred buildings though, and so trying to automate the process and make it repeatable. The second part of your answer seems like it could be more promising (you guessed correctly that a Python library would be appreciated!).

( 2015-09-04 09:28:44 -0600 )edit
1

I know you share the Python love :)

( 2015-09-04 09:56:44 -0600 )edit

I was checking for licensing on that code and notice that it's no longer being maintained. The new home is at pvlib-python - and the licence is BSD 3-clause.

( 2015-09-04 10:01:01 -0600 )edit

I would make sure you only model the facades of adjacent buildings necessary for shading vs. modeling the entire building shape. I do think at a certain distance away the return on investment (in modeling time and run time) stops pay off for the impact it has on the model. I wouldn't use a set distance away as what to shade as I would the distance away relative to the height of the building. So you may not model a 1 story building 2 block away but would model a 40 story building 4 blocks away.

( 2015-09-04 10:28:15 -0600 )edit
1

Also if you have a block of similar 2 story buildings that are not exactly touching or exactly the same height, you may just want to model it as a single shading surface. My initial thought about distance away was related to the vertical sun angle, but another factor is that as you get further away the buildings shadow moves more quickly horizontally across your building, minimizing its impact with distance. So of course a tall wide building will be much more important than a tall skinny building. You could also skip anything too far north unless you are going to look at reflections off them.

( 2015-09-04 10:39:28 -0600 )edit