re: smart vents
14 mar 1996
david coote wrote:
>one cheap form of ac in a house is provided by opening and closing doors
>and windows depending on the temperature gradient between inside and outside.
sounds good, david. baruch givoni's new _passive and low energy cooling of
buildings_ talks about cooling the thermal mass of a house at night by
ventilation and keeping it cool during the day with no inside-outside air
exchange. to make this work well we want lots of ventilation, say 1 cfm per
square foot of house.
suppose the house is 2,000 ft^2 with 3,000 ft^2 of r20 walls, and the daytime
temp is 100 f, and the average temp in the house is t. then during a 12 hour
day, the house will gain 12(100-t)3,000/r20 = 180k - 1800 t btu. if the house
starts out at dawn at 70 f, and it has a thermal mass equivalent to say, 3000
lb of water, then during the day the house will absorb 180k-1800(70) = 54 k
btu, and since 1 btu raises 1 lb of water 1 degree f, the house temp will rise
by about 54k/3k = 18 f to 88 f. let's try more thermal mass: suppose the house
has masonry walls and ceiling with a thermal mass of 0.16 btu/lb-f x 32 lb =
5 btu/f for each 8" x 16" cement block, with r20 insulation outside. then the
walls and ceiling would have a thermal mass of about 6 x 3,000 = 18 k btu/f,
equivalent to about 18,000 pounds of water, so the temp rise would be 1/6 of
the above, ie 70 f --> 73 f. that's better. the bottom and top rows of blocks
might have holes in them from the inside of the house to allow house air to
flow through them, increasing their heat exchange surface area.
now suppose the temp t of the thermal mass inside the house does not change
at all from night to day. suppose the night temp is 70 f and the fan is 2,000
cfm. then during an 8 hour night, the fan would move roughly (8)(2,000)(t-70)
btu, ie 16kt - 1120k = 180k - 1800t. solving for t,
17,800 t = 1300k, or
t = 1300k/17,800
= 73.03 degrees f. pretty cool.
>if no-one's home all day the house is going to be stifling when
>you get back from work.
this idea is cool the house down well the night before, when the air is cooler
than daytime air, and close up the house during the day and let the thermal
mass of the house keep it cool.
>what might be nice is a micro-processor controlled motorised vent system.
>enter the desired temperature at the system panel and let the system open
>and close vents appropriately depending on internal and external temperatures.
one might also use a differential thermostat to open the vents when it's cooler
outside than in the house, eg a heliotrope general ((800) 552-8838) dtt-94,
which has a trade price of about $75, which would need an output relay to
drive a three terminal damper, or a spring return damper. honeywell probably
makes something like this too. or one might use a passive plastic film damper
that opens itself at night, with a manual door behind it. if the house has
high thermal mass and good insulation, its temperature will change slowly
from day to day. this house with the masonry walls would have a natural "rc
time constant" of r20/3,000 x 18k = 120 hours or 5 days. this is good for
solar heating too.
>the vents could be tucked away under eaves as their primary purpose is
>not to let in light. they'd have to be placed so as not to let in rain.
one might put them in an insulated attic floor, with reflective undersides,
perhaps made of foil faced foamboard, and let them open in winter towards
the sun, under a steep transparent roof, to admit light and heat. they might
also act as return air dampers for solar warmed air blown down from the peak
of the roof to the house in the winter.
>other considerations would include: reliability of the motors;
most damper motors, eg the $50 honeywell 2 watt 6161b1000 actuator, have a
lifetime of at least 100,000 cycles, ie 300 years at once per day. large
low-speed fans are also fairly energy-efficient. the 24 vac version of this
motor has mechanical stops and a 120 degree rotation, max, with a max torque
of 45 in-lb. it has 3 terminals, a common one, one to make it go clockwise
when 24 volts is applied, and one to make it go ccw. honeywell (actually
american warming, now) make a nice d642ls "ultra-low leakage" damper that
works with these, with a trade price of works with these, at a cost of $100
or so, but because of the price and the fact that it is made of metal with
no insulation, i prefer making a little door out of foil faced foamboard.
>how much vent area needed to get useful heat flows;
the air that flows out of an a ft^2 hole in a wall h feet high, with a temp
ti (f) on the inside and a cooler temp to (f) on the outside is roughly
q = 16.6 a square root ((ti = to) h) cfm. q cfm of air flowing through
a temperature difference of d degrees f carries about qd btu/hour.
>anyone know of residential systems of this type?
seems like you can make up a system like this with standard honeywell parts...
>any experience with these systems?
the dampers and motors seem to work nicely.