Lighting densities and Electrical equipment contributing to Space Heating

7 posts / 0 new
Last post

Hi everyone,

I feel that this topic must have been discussed at some point, but an
exhaustive archive search got me nowhere, so I thought I'd throw together a
post of my own, especially since I'm working on an eQuest model of a small
server building which is heated in part by a large number of computers that
run many hours every day.

I'm interested in how eQuest treats electric loads - from lights, computers,
refrigerators, etc - as regards space heating. I had always assumed that,
especially in the case of lights, much of the energy used would be converted
to heat within the building, and contribute to space heating.

However, I set up an experiment to test this and found that the results were
far different from what I'd expected: I upped the lighting density in an
existing natural-gas-heated building model by about 15x and compared the
electric use from before and after the lights were increased. I converted
all values to mmBtus for easy comparison. To my surprise, I found that only
a very small fraction (about 6%) of the mmBtus added to the building through
those lights contributed to space heat. The kWh recorded were increased
hugely, but the heating energy required to keep the building heated was
almost the same in both cases.

I repeated the experiment, upping misc. equipment instead of lighting, and
saw a similar result.

Does anyone have any knowledge of what equations eQuest uses to decide how
much electrical energy use is converted to heat and added to the space in
which it is installed?

Regards,
Taylor Sharpe

Taylor Sharpe Energy Modeler Sharpe Energy Solutions
Taylor Sharpe's picture
Offline
Joined: 2011-09-30
Reputation: 0

Taylor,

A number of things play into the question you are asking.

* Different types of lighting have different efficiencies. Incandescent lights create a lot of heat, LEDs create practically no heat. Fluorescent is somewhere in between but create much less heat then incandescent

* The style of light installed - is it pendant hung in the space? If so, 100% of the created heat is given to the space. Is it recessed? If so, a portion of the heat is given to the plenum, and some of the heat given to the plenum leaves the building as relief air (in fact, it's possible that all the plenum heat leaves the building in the case of a 100% outdoor air system)

* The schedule of the lights. Were you looking at yearly consumption or peak heating load? Peak heating load is typically calculated with the lights (and occupancy and misc loads) at 0%. If you were looking at yearly consumption, but your modeled facility has the lights turned off for a majority of the time, the impact may be minimized.

Point being, you have to consider more than just lighting density when you make this analysis.

John Bixler, EIT, LEED AP BD+C

Garasaki's picture
Offline
Joined: 2011-02-16
Reputation: 1
Taylor, Heat gains from lighting in spaces is shared between the space and its plenum depending on the values for LIGHT-TO-SPACE and LIGHT-TO-RETURN. For other electrical equipment and for occupants, the heat gains all go to the space, but are divided into latent and sensible components. I believe all of the electricity consumed by lighting and equipment is converted to heat gains. The heat gains are converted to cooling loads using weighting factors. The weighting factors can be modified by specifying the convective/radiative splits, through keywords light LIGHT-RAD-FRAC. The radiative portion can be absorbed by walls and furniture and so does not create an immediate (for that hour) cooling load. You can create custom hourly reports to see how this plays out in the building by using the "Building Loads" variable and selecting "Building light heat load", "Building light cool load" and "Building light elec total". Similar variables can be selected for individual spaces. If you haven't already downloaded it, I recommend looking at the DOE-2 Engineers Manual: http://doe2.com/download/DOE-21E/DOE-2EngineersManualVersion2.1A.pdf Section 2.5 - Interior Loads describes the algorithms for loads calculations. Keep in mind that changing your lighting and equipment values will impact cooling energy as well. You may be seeing small changes in heating energy but much larger changes in cooling energy, depending on other factors like your envelope, ventilation loads etc. Regards, Bill
Bishop, Bill2's picture
Offline
Joined: 2011-09-30
Reputation: 0
Classification: UNCLASSIFIED Caveats: NONE All, Are LED lights really magical? We have all used a laptop and felt the AC/DC converter get hot. We know that if a transformers (not the cool kind) are inside a building they put off a lot of heat. What about the transformers for the LED lights? LED lights need DC power, where are all the transformers and were does that heat go? I suspect the LED light is like the electric car. It is cool that the car runs off of electricity, just don't look at the coal that is burnt to make the electricity. Yes LED lights give off almost no heat, but what about the transformer? Anybody have any information on this? What is the heat of rejection from the transformers.... (I have done a little research and one solution was to have a single transformer for the building and wire the lighting of the building with DC power. Talk about changing business as usual.) John Eurek
Jeurek's picture
Offline
Joined: 2010-10-07
Reputation: 0

John,

Hopefully this will help. It is an article that explains heat dissipation
with items that have small hot spots.
http://www.ledjournal.com/eprints/nextreme_sept09.html

Your discussion is far more philosophical and delves into a discussion about
embodied energy. What you are talking about below is addressing multiple
issues which are far more complex than the building energy modeling we are
doing right now. Remember that you are simulating.

I have started going down these roads and it turns out they are very
complex. Keep it simple and use averages.

Thanks,

PETER HILLERMANN

Peter Hillermann2's picture
Joined: 2011-09-30
Reputation: 0

LED's use DC power, and do so efficiently. Sounds like you're referring to the AC-to-DC power supply though, not the LED light source itself.

As described in your post below ACDC power converters can be loosely lumped into 2 categories:
- old, inefficient, big, hot, LINEAR converters with their big transformers, and
- newer smaller, cooler, SWITCHING converters ("the cool kind" referred to below)

I think Enerstar has done a lot to phase out the inefficient LINEAR converters. You can still procure the LINEAR power supplies. But you probably won't find one that meets Enerstar requirements. The SWITCHING supplies are much more efficient, and may even cost less now, as they are smaller, lighter, and have less metal in them.

So here's a couple of Enerstar links for some reference material:
http://www.energystar.gov/index.cfm?c=archives.power_supplies
http://www.energystar.gov/ia/partners/product_specs/program_reqs/eps_prog_req.pdf

For more information on SWITCHING ac to dc supplies:
http://en.wikipedia.org/wiki/Power_supply#AC.2FDC_supply

The Wiki article gives some ideas as to the "magic" that is achieved with the new supplies.

If the question was about the LED's themselves and not the transformers/converters that go with them then please excuse my sidetrack.

If anyone has seen central DC converters wired throughout the facility I'd be very interested to hear about it. Certainly many loads are now DC. Perhaps the Tesla Wars (http://en.wikipedia.org/wiki/War_of_Currents) are swinging back towards Edison for efficiency reasons.

Charlie Holleran

Charles Holleran's picture
Joined: 2011-11-12
Reputation: 0

John et al:

The query seems to be branching out pretty quickly ... I have many
thoughts on the broader topic and present state of LED's and their that
wouldn't make everyone happy to read/hear, but I think this addresses
the heart of John's post directly:

The problem of lights as an internal heat load remains pretty simple for
us energy modelers and HVAC designers, and isn't made more difficult by
LED technology.

Consider a light fixture of any source as the boundaries of a
thermodynamic problem: Energy in = Energy out. If we put 100W of
energy into the fixture, then 100W of energy must be produced in some
form. That output is some combination of visible light energy that our
eyeballs can perceive, and invisible other energy which can largely be
perceived as heat/sound after transmitting through the air or other
surfaces to arrive our eardrums/skin. Normally that ratio of output
visible light energy is tiny, but for certain LED sources it's less
tiny**. That fraction is entirely moot however from an building energy
simulation standpoint. All energy output in the form of light will
encounter surfaces (floors/walls/people) and ultimately be absorbed*.
In the process, the energy will briefly excite a few electrons... but
when the lights go off and the excited electron-party is over all that's
left is energy irradiated an invisible wavelengths (excepting
glow-in-the-dark t-shirts). The big point here is, energy input is the
only figure/variable we really care about, because that's exactly the
amount of energy we need to consider from an HVAC heating/cooling
perspective. None of it just goes away.

~Nick

* Acknowledging: A fraction of light hitting exterior windows transmits
out of a building - I'll claim this negligible since that's a fraction
(glass transmittance) of a fraction (window to every-other-surface
ratio) of an already small fraction of the input watts (visible light
vs. other wavelengths).

** If anyone cares to better understand this careful choice of words, or
wishes to better cut through marketing mis-information from the lighting
industry, I would first advise taking some time to review difference in
the terms 'efficacy' & 'efficiency,' then identify the contextual
mis-use in the preceding discussion below. Incorrect conclusions
regarding heat are made as a result. A hint: Both can use units of
"lumens per watt" in the right context, but lumens per watt of light
output is not the same thing as lumen of light output per watt of input
energy ;).

NICK CATON, P.E.

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