Modeling Light Shelves for Reduction in Cooling/Heating Load?

7 posts / 0 new
Last post

E.G. Would it be accurate to model an external light shelve as a
horizontal building shade with a high reflectance value on the skyward
facing surface? I understand the day lighting benefit in Equest would be
inaccurate, but we are using separate software to obtain accurate day
lighting results. Thanks!

Nilay Desai

Nilay R. Desai's picture
Offline
Joined: 2011-09-30
Reputation: 0

Perhaps these are the answers you're seeking:

1. Building shades and their respective reflectance properties do
affect the modeled behavior of space daylighting systems in eQuest
models. Be sure to pay close attention to all three values (and their
associated help files) on the right of the following image - they may
all play a part depending on what exactly you're trying to model.

2. Based on your query, a caution: eQuest's results do not have
any inherent degree of accuracy/in-accuracy with regard to the modeled
energy consumption resulting from daylighting controls - that depends
pretty much entirely on how closely you define the daylighting systems'
behavior to reality. If you have a program which you perceive to be
producing more accurate hourly percentage consumptions for your spaces,
then by all means - there's nothing stopping you from defining a
schedule in eQuest to match the other program's results.

I expect you mean to say you're utilizing another program to accurately
determine the behavior of daylight in response to adding/modifying light
shelves, but if you are indeed using another program to determine annual
hourly behavior with a weather file, I'd be very interested to hear more
about it!

Best of luck,

NICK CATON, E.I.T.

Nick-Caton's picture
Offline
Joined: 2011-09-30
Reputation: 805

Equest, like DOE2.1e before it provides very
generous estimates of the lighting energy saved
using daylighting controls. I believe that
taking equest predictions at face value for LEED
models would be inaccurately providing more
benefit than warranted. The Appendix G power
factor ratings are likely not conservative in
some cases. I wouldn't expect to see any more
benefit using equest's daylighting than I would
using the Appendix G power factor adjustments.

A paper that attempted to explore various
daylighting energy reduction savings modelling
strategies. DOE2.1e and Daysim are the two programs used for the analysis.
http://www.nrc-cnrc.gc.ca/obj/irc/doc/pubs/nrcc46761/nrcc46761.pdf

Chris Jones

Chris Jones's picture
Offline
Joined: 2011-09-30
Reputation: 0

It's not my intent to "take sides" or anything, but a curious reader
(me) might ask "What do you use, Chris?"

As I'm pointing out - a saving grace of eQuest/DOE2.2 is that you can
spoon-feed whatever scheduling you may have generated in an outside
program if you don't like what you're seeing. I understand some
distrust eQuest's results enough to regularly use DAYsim to generate
custom scheduling. In my practice, I've found it more time-efficient to
work within and around the known limitations of eQuest, for the sort of
energy modeling I'm regularly tasked with. I also prefer to use other
software packages to model instantaneous daylighting behavior - anyone
interested, see archives.

I do think it's a bit of a stretch however to claim the hourly modeled
daylighting behavior of eQuest is comparably dumbed-down as a blanket
percentage reduction, as we do with LPD's for occupancy sensors under
90.1 (I think that's what you're referencing) - eQuest's current
capabilities do allow an end-user to effectively compare multiple custom
daylighting control strategies and their resultant effects on the rest
of a building's systems (i.e. heating/cooling) in a time-efficient
fashion.

I'm not disagreeing the DOE2.1e algorithms aren't the best for figuring
how many fc are incident in a given space for a given hour, but
statements that boil down to "eQuest sucks for daylighting" (for LEED
modeling or otherwise) can certainly be misleading. If you can take
what's known to be inaccurate into account, you can realize eQuest as a
pretty useful tool for designing with daylighting.

NICK CATON, E.I.T.

Nick-Caton's picture
Offline
Joined: 2011-09-30
Reputation: 805

The straight answer is "yes" you may model light shelves as external
building shades for your proposed model and not get into trouble with
regard to 90.1. You would probably not be taking credit for the full
amount of daylight incident in your spaces on an hourly basis however -
you may or may not be able to sleep at night =).

eQuest does not accurately model interreflected daylight, so you are
correct to assert such a model will not in and of itself accurately
account for the effects of light shelves upon your daylighting systems.
Note the values entered for surface/ground reflectance may still alter
the amount of light incident on a given window through their effects.
The effects of those shelves with regard to shading is something you can
be more confident of, however.

If you'd like to take better credit for light incident on your daylit
spaces, I understand some have pursued using a combination of
daysim/SPOT/excel to generate custom scheduling for use within eQuest.
I haven't traveled this road yet, but I've copied what would probably be
a worthwhile discussion below from last Janurary if you'd like to pursue
;).

NICK CATON, E.I.T.

Nick-Caton's picture
Offline
Joined: 2011-09-30
Reputation: 805

In that case I just may have to explore the daysim/SPOT/excel path you
mentioned. Thank you so much for your help!

Nilay Desai

Nilay R. Desai's picture
Offline
Joined: 2011-09-30
Reputation: 0

I agree, the tool must suit the application. My
intended message is that you can use equest for
daylighting, but you need to be aware that equest
is will produce the results based on the user's
input - garbage in, garbage out. One needs to
take into account occupant behaviour. I feel it
is better to use a equest/DOE2.1 for daylighting
savings as it captures the actual lighting demand
peak but I tend to adjust my doe2 models to
achieve a lighting power reduction similar to
what you would see if you used the PAF factors listed in Appendix G.

Chris Jones's picture
Offline
Joined: 2011-09-30
Reputation: 0