# Calculating maximum illuminance levels from daylight

Just a quick one guys....

We are doing some radiance analysis on a new building design. The
building is to be located in Auckland, New Zealand (Southern Hemisphere,
full of orcs and hobbits and elves). The design is a museum building that
specifies a maximum of 350lux on a boat that is on display. We have to
ensure that daylight (sunlight + skylight) does not give levels greater
than this. We have designed the building to completely shade the boat
surface from direct sunlight throughout the year.

So, I have set up a numerical grid for analysis and ran the model under an
overcast sky and then used local meterological outside illuminance data.
This data gives a maximum diffuse illuminance of 50,000lux.

Using this overcast sky model the design works. However, would this be the
'worst case scenario'? I understand that a clear sky gives a greater
luminance distribution around the horizon that an overcast sky.

Following a shading analysis we have found that the most exposure to the
sky and direct sunlight will be at 12pm during the winter solstice (June
21) to the north facade. At this point the sun will be at an altitude of
30 degrees.

In order to calculate absolute illuminace levels on the boat can I just
use the CIE Clear sky model set up for this time period using the default
Or do I have to specify the maximum values?

Ben

Ben,

There are various ways to estimate time-varying daylight
illuminance from a few 'static' simulations. Your application
however seems to require fairly precise values. I think the
only way to achieve them with any certainty is to carry
out a full blown time-varying daylight illuminance calculation
using a full year of 1hr met. data. If you're not convinced,
see Figures 6.36 and 6.37 (Chapter 6) here:

http://www.iesd.dmu.ac.uk/~jm/zxcv-thesis/

for an illustration of the difference between static daylight
factors and time-varying daylight factors.

To generate time-varying daylight illuminances, try Christoph
Reinhart's DAYSIM package:

http://irc.nrc-cnrc.gc.ca/ie/light/daysim.html

-John Mardaljevic

PS. I would have suggested also the DLS (IESD's own Radiance-based
tool to predict time-varying daylight illuminance). But
it is currently broken becuase changes in Java have not yet been
incorporated into the tool.

···

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

http://www.iesd.dmu.ac.uk/~jm

Thanks John,

I understand what you are proposing and have read your thesis previously
during my studies.

Local weather information available contains data for an average or
typical year. The sky may not be sunny during the time when the building
has the highest level of solar exposure (Winter solstice). Since the CIE
Clear Sky represents an extreme sky condition I thought I could pretty
much cover the worst case scenario, not actual measured irradiance data.

So, I am still unsure weather Radiance or Gensky will automatically
calculate the direct and diffuse illuminance given a clear sky, time
location etc. and if it does will these numbers be realistic or
theoretical? I am not concerned about overdesigning to handle an extreme
sky condition.

Surely once light from the sun has penetrated the earths atmosphere it is
at a constant level, varying in intensity by solar altitude (cosine law).

Thanks,
Ben

Ben,

So, I am still unsure weather Radiance or Gensky will automatically
calculate the direct and diffuse illuminance given a clear sky, time
location etc. and if it does will these numbers be realistic or
theoretical?

Yes, gensky can *estimate* absolute values from just location, time
and user specified sky conditions, e.g. CIE clear. But there is no
knowing how good the estimate is until you make comparison with
absolute values from, say, a TRY. My suggestion is: find the
maximum global horizontal irradiance (& associated diffuse horizontal)
in the TRY, and use those to set your -R and -B gensky parameters.
Doesn't matter too much when in the year it occurs, it's more a
case of getting an idea of what likely magnitudes are.
Maybe increase the overall value by a little to account for very extreme
conditions, but be careful not to exceed theoretical limits.

-John

···

-----------------------------------------------
Dr. John Mardaljevic
Senior Research Fellow
Institute of Energy and Sustainable Development
De Montfort University
The Gateway
Leicester
LE1 9BH, UK
+44 (0) 116 257 7972
+44 (0) 116 257 7981 (fax)

http://www.iesd.dmu.ac.uk/~jm