re: seek passive solar design faq/guide
8 may 1996
william r stewart wrote:
>george p swanton wrote:
>> can anyone offer a pointer to a set of formulae for glazing/
>> mass/temperature/etc calculations involved in passive solar
>> design? i am interested primarily in direct gain residential
>> space heating applications.
>these calculations are somewhat complex, if you are considering passive
>solar as your main heat source. the passive solar industries council
>has a complete book and software program for this engineering problem.
> passive solar industries council
> 1511 k st., #600
> washington, dc 20005
> (202) 628-7400
yeah, the brick people :-) give them a call if you want to fill up your
sunspace with thermal mass and cripple the performance, while raising
the price dramatically :-) or if you want your "main heat source" to
provide less than half the heat for your house.
two views on sunspace design:
it is hard to think of any other system that supplies so much heat
(to an existing house) at such low cost...
one could shorten the warm-up time of the enclosure and increase
the amount of heat delivered to the rooms by making the enclosure
virtually massless--by greatly reducing its dynamic thermal capacity.
this can be done by spreading a 2-inch-thick layer of lightweight
insulation on the floor and north wall of the enclosure and then
installing a thin black sheet over the insulation. then, practically
no heat is delivered to the massive components of floor or wall;
practically all of the heat is promptly transferred to the air.
and since the thermal capacity of the 100 or 200 lb. of air in
the room is equal to that of one fourth as great a mass of water
(about 25 to 50 lb. of water), the air will heat up very rapidly.
i estimate that its temperature will rise about 40 f. degrees in about
two minutes, after the sun comes out from behind a heavy cloud cover.
at the end of the day, little heat will be "left on base" in the
collector floor or north wall and, accordingly, the enclosure will
cool off very rapidly.
new inventions in low cost solar heating--
100 daring schemes tried and untried
by william a. shurcliff, phd, physics, harvard
brick house publishing, 1979, 293 pages, $12
a sunspace has extensive south-facing glass, so sufficient thermal mass
is important. without it, the sunspace is liable to be uncomfortably hot
during the day, and too cold for plants or people at night.
however, the temperature in the sunspace can vary more than in the
house itself, so about three square feet of four inch thick thermal
mass for each square foot of sunspace glazing should be adequate...
the sunspace floor is a good location for thermal mass. the mass floors
should be dark in color. no more than 15-25% of the floor slab should be
covered with rugs or plants... another good location for thermal mass
is the common wall (the wall separating the sunspace from the rest of
the house)... water in various types of containers is another form of
energy storage often used in sunspaces.
passive solar design guidelines--
guidelines for homebuilders
for philadelphia, pennsylvania
passive solar industries council
national renewable energy laboratory
charles eley associates
current (1995) edition, 88 pages, $50
so, which is the most energy-efficient sunspace in a partly cloudy climate
shurcliff's plastic film sunspace, wearing the green uniform in this
contest, might cost about $1/ft^2, and on an average december day at 36 f,
it would receive about 1000 btu/ft^2 of sun, like the psic sunspace. let's
assume that both sunspaces have a perfectly insulated wall between them and
the house, to avoid the thermal disaster of a poorly insulated trombe wall
in a partly cloudy climate, and let's assume there is no air infiltration
from the outside in either case. the sunspace air would be circulated
through the house with some dampers or fans, keeping the sunspace at 80 f,
say, while the house remains at 70 f. with single glazing, about 900
btu/ft^2 of sun might enter the sunspace during the day, and the amount of
heat lost through a square foot of shurcliff's sunspace over a typical day
would be about 6 hr (80-36)/r1 = 264 btu/ft^2/day, for a net gain of 636
btu/ft^2, ie his $1/ft^2 sunspace would be about 64% efficient, as a solar
collector. a 16' x 32' sunspace like this costing $500, along with a solar
closet containing 20 55 gallon drums full of water, could provide all of
the heat and hot water needed for an attached 32 x 32' two-story house with
an average r20 envelope. as an auxiliary living space, it could be heated
up instantly on some starry night for a party, by moving some warm air from
the house into the sunspace.
this sunspace might have a single layer of mylar glazing made by bayer or
dow chemical, distributed by replex or armin plastics, stretched over some
curved galvanized pipes, with their curved ends tucked under the south eave
of a two story house. the film might be attached with aluminum extrusion
clamps around the perimeter of the sunspace, with a landscaping timber
foundation, staked to the ground with 4' of #4 rebar. the sunspace might
have a layer of green colored greenhouse shadecloth hanging inside to help
absorb the sun. opening some vents and hanging the shadecloth over the
outside in summertime would keep the sunspace and house cooler, and prolong
the life of the glazing. the sunspace might have a crushed stone floor over
black polyethylene film, with a shallow reflecting pond in front, made from
a single layer of epdm rubber draped over a low perimeter earth berm. a
transparent motorized damper in a first floor window would allow house air
to flow into the sunspace, if the sunspace were warmer than 80 f and the
house were cooler than 70 f, and a second floor window fan with a one-way
plastic film damper would move air from the sunspace into the house when
the house needed heat, with the first floor damper open. the fan would also
operate on windy days and nights, perhaps with the downstairs damper
closed, to create a slight vacuum inside the sunspace, to avoid plastic
the psic sunspace, wearing the brown uniform, would perform better with
double glazing. it might cost $10/ft^2, with a 4" concrete thermal mass
with an official psic heat capacity of 8.8 btu//f-ft^2. say the concrete
absorbs 100% of the sun that falls on it, vs the official psic solar
absorptance of 0.65 (table k, page 57.) then about 800 btu/ft^2/day of sun
will enter the double glazing and be absorbed by the concrete, and the
concrete surface will warm up the sunspace air, and that warm air can be
used to heat the house when the sunspace temperature is more than 80 f.
suppose the concrete loses no heat at all to the soil below (i'm giving
quite a few handicaps to the psic sunspace in this efficiency race.) the
concrete might start the day at temperature t, and charges up in the sun to
a max temperature of t + dt, and return to temperature t at dawn. how can
we calculate t and dt? the equivalent electrical circuit looks something
ts sunspace temperature
r2 | d
36 f ---------wwww-----------------|-------------------- 70 f
outdoors glazing | open damper to heat house
w r0.5 concrete - sunspace air resistance
800 btu/ft^2 w
per day |
| --- |
| | ----|-->|------|--tc concrete temperature
| --- |
sun current w
source w r0.4 concrete bulk thermal resistance
------- 26.4 btu/f thermal mass of concrete
let's simplify this by assuming the thermal mass of the concrete is
infinite, vs 8.8 btu/f-ft^2. lots of concrete, or a water wall, or
something with so much thermal mass that the temperature inside the
sunspace never changes at all from day to night over a long string of
average december days, with some sun. this is an optimal sunspace with more
than "adequate" or "sufficient" thermal mass by official psic standards.
let's also assume that the two small resistors have a value of zero, ie
let's ignore the r0.4 bulk thermal resistance of the concrete, that makes
the surface heat up more than the inside, while the sun is warming it up,
and makes it harder to get heat out of the inside of the concrete and into
the sunspace air, and the r0.5 concrete-sunspace air resistance, by
assuming both are r0 conductors. what will tc be in that simplified case?
the sun shines into the sunspace during the day and adds 800 btu to our
concrete capacitor, and over 24 hours, 24(tc-36)1ft^2/r2 = 12 tc - 432 btu
flow out of the capacitor. if ein = eout (providing no heat for the
attached house), then tc = (800+432)/12 = 103 f. pretty nice, but this
sunspace is not providing any heat for the house, just keeping itself warm
on an average day, and losing lots of heat on a cloudy day. suppose we
allow some heat to flow from the sunspace into the house, ie close the
switch, ie turn on the fan or open the damper between the sunspace and the
house often enough to limit the maximum sunspace temp to 80 f instead of
103 f. then the heat loss to the outside world over the course of a day is
24(80-36)1 ft^2/r2 = 528 btu, and the rest of the heat that enters the
double glazing, ie 800 - 528 = 272 btu/ft^2/day goes into heating the
house, so the solar collection efficiency of this $10/ft^2 sunspace in
auxiliary living space, the temperature of this sunspace is largely out of
our control. it takes a long time and a lot of house heat to warm it up on
an evening or cloudy day, and after we leave the space, it stays warm for a
long time, giving up precious house heat to the outside world.
how curious that by carefully following the current official guidelines of
the passive solar industries council, we can reduce the performance of
shurcliff's low-thermal-mass sunspace from 64% to 27%, while increasing the
price from $1/ft^2 to $10/ft^2, unimproving the cost-effectiveness of the
sunspace by a factor of 12, even with all these psic-slanted assumptions...
here's a quote from the acknowlegements section of the psic guidelines:
_passive solar design strategies: guidelines for home builders_
represents over three years of effort by a unique group of organizations
and individuals. the challenge of creating an effective design tool that
could be customized for the specific needs of builders in cities and t
towns all over the u. s. called for the talents and experience of
specialists in many different areas of expertise.
_passive solar design strategies_ is based on research sponsored by the
united states department of energy (doe) solar buildings program, and
carried out by the los alamos national laboratory, the national
renewable energy laboratory (nrel)... and the florida solar energy
the national association of home builders (nahb) standing committee on
energy has provided invaluable advice and assistance during the
development of the guidelines.
valuable information was drawn from the 14 country international energy
agency (iea) solar heating and cooling program, task vii on passive and
hybrid solar low energy buildings...
although all the members of psic, especially the technical committee,
contributed to the financial and technical support of the guidelines,
several contributed far beyond the call of duty. stephen szoke, director
of national accounts, national concrete masonry association, chairman of
psic's board of directors during the development of the guildlines; and
james tann, brick institute of america, region 4, chairman of psic's
technical committee during the development of these guidelines...
gave unstintingly of their time, their expertise, and their enthusiasm.
>some of the variables involved in such a design include;
>what is the heat loss rate of your structure?
yes, that's a good thing to know... "ohm's law for heatflow"... note glass
is a very poor insulator... a 30 x 30' x 2 story house with r20 walls and
ceiling might have a thermal conductance of 2000 ft^2/r20 = 100 btu/hr-f.
make 10% of the wall area windows by adding 200 ft^2 of r2 glass and this
doubles to 200 btu/hr per degree f--unless the glass is in a thermally
isolated sunspace, in which case the thermal conductance and heat loss
of the house go down, not up...
>what is the solar insolation in your area and when does it occur?
also good to know, eg the amount of sun that falls on a south wall on a
december day, as well as the average temperature in december. if your
house stores heat for several days, these averages are good enough for
design. you don't need to know much more detailed weather data. (altho
it is very nice to be able to simulate the performance of a house design
easily, hour by hour, over the last 30 years, if you want to do that,
with the nrel/noaa cd-rom data.)
joe mccabe, pe, lives there :-)
>what are your backup systems (eg, masonry fireplace, ground-source
ideally none. this is how some people define a "solar house," ie one with
no other form of heat... simple, no? such a house can be easily designed
with some high school physics and algebra, as licensed professional engineer
norman saunders has been doing in cold, cloudy new england since 1944.
or... you can follow the orthodox passive solar industries guidelines above,
which say that no matter how hard you try, you can't provide more than 41%
of the heat needed for a house in the (warmer, sunnier) phildelphia area
using the sun, if you fill up your sunspaces with bricks, that is :-)
* * *
here's the scoop. the way to do this is simple. start by finding 3 numbers:
1. find the heat loss for your house, eg 200 btu/hr per degree f.
2. find the average temperature in december where you live, eg 36 f.
3. find the average amount of sun that falls on a south wall in december
where you live, eg 1000 btu/ft^2/day, using nrel's numbers for
philadelphia, assuming a little more ground reflection.
size a low-thermal-mass sunspace to provide 100% of the heat for the house
on an average december day, with some sun. there are several steps here:
4. find how much heat your house needs on an average december day. if it
needs, say, 200 btu/hr/degree f, using "ohm's law for heatflow," on
an average 36 f day, it will need 24 hours (70-36) 200 = 163k btu/day
to stay at 70 f inside.
5. find how much net heat a square foot of low-thermal-mass sunspace can
gather on an average day where you live. suppose the sunspace takes
in 1000 btu/ft^2/day with r1 single glazing. then if we let the sunspace
temperature rise to, say, 80 f during an average 6 hour december day,
so it can provide warm air to heat the 70 f house, the loss will be about
6 hours (80-36)1 ft^2/r1 = 264 btu, for a net gain of 736 btu/ft^2/day.
6. size the sunspace glazing. in this case, we need 163k btu/day divided by
736 btu/ft^2/day, ie 221 ft^2 of low-thermal-mass sunspace. this might be
a lean-to sunspace 16' tall and 16' wide and 8' deep, made with 5, 20' long
curved galvanized pipes on 4' centers, costing $35 each, with some clear
mylar film costing 10 cents/ft^2 stretched over the pipes and a "foundation"
consisting of railroad ties ("landscape timbers") spiked to the ground
with 4' of rebar, with a floor made of black poly film covered with round
pebbles. or it might be an "expensive" sunspace made with $80 worth of
2x6s with some clear single layer polycarbonate plastic costing $1/ft^2,
or one might just use the clear polycarbonate plastic instead of vinyl
siding on the south side of the house, as "solar siding."
7. size the thermal mass in a solar closet or pure attic warmstore to provide
heat for the house for a week or so without sun. for 5 days, say, we need
to store 163k btu/day x 5 days = 815k btu of heat. if we use 55 gallon
drums full of water starting at 130 f, and cooling to 80 f over 5 days,
each storing 25k btu of heat, we need 815k/25k = 32 of them, each 3' tall
by 2' in diameter, tucked away somewhere, receiving the sun thru an inner
layer of glazing on the back wall of the sunspace, with some insulation
behind that glazing to make a passive air heater.
8. other options: add a water heater and 10' of fin tube pipe inside the
solar closet or warmstore to make domestic hot water, using a little
more glazing, or add a sauna, or a place to dry clothes...
* * *
there, that wasn't hard, was it? that's how to design an inexpensive 100%
solar house, with no backup heating system, that also makes hot water, just
like norman saunders has been doing since 1944...
there are a few more little details to check, but this isn't rocket science,
or even college physics. it's somewhere between figuring out a restaurant tip
and doing a simple beam strength calcualtion, as architects know how to do.
>do you plan to use a trombe wall, free-standing thermal mass, floor mass, etc?
ah yes, you might use a trombe wall, invented by felix trombe in 1964 (and
patented by edward morse of salem, ma, in 1881) or a picture window in the
living room, with a masonry floor in front of that... a "direct loss" house,
like the one architect george f. keck called a "solar house" in 1934 :-)
a few years ago, i spent some time explaining to a local architect, a more
technical person than most, who had taken a few engineering courses on the
way to architecting, that a "trombe wall" with some insulation on the outside
and some passive plastic film dampers to the inside of the house, that
opened up during the day, was probably a lot more efficient at collecting and
keeping solar heat in the house than a plain old "traditional" trombe wall,
with masonry right behind the glass, with no insulation. here's what i said:
a "trombe wall" with insulation on the outside, and 1 square foot of south-
facing single-glazed area and an r-value of 20, will receive about 1000
btu/day of heat on an average 32f december day, where i live. if the room
behind it has a constant temp of 70f, and the sun shines 6 hours a day,
on the average, the energy that leaks out of the glass will be about 6 hours
x (70f-32f) x 1 ft^2/r1 = 228 btu during the day, and 18 hours x (70-32) x
1 ft^2/r20 = 34 btu at night, a net gain of 1000 -228 -34 = 738 btu/day.
simple, no? (750 btu, net, with double glazing, which passes less sun.)
a standard unvented trombe wall (table iv-14b of mazria's book says vented
ones don't work much better) with a very large uninsulated thermal mass
right behind the glass and an r-value of, say 2 (roughly 1' of masonry),
would have an average temperature at the outside wall surface of about
32f + r1 x (70f-32f)/(r2+r1) = 45f, if there were no sun. if you add a
heatflow of 1000 btu/day of sun to that model, falling on the outside
of the wall, the outside wall surface will have an average temperature
of about 45f + 1000/24 x (r=2/3) = 72.4f, which contributes 24 hours x
(72.4f- 70f) x 1 ft^2/r2 = 29 btu/day to the room behind the wall.
so the "improved trombe wall" above, (actually an air heater with the
thermal storage inside the house) is more than ***25 times*** as efficient
(738/29) at collecting and keeping heat in the room behind it, than
the usual trombe wall. this is somewhat oversimplified, of course...
and do you know what the architect said? "i agree with you completely, but
if you do that, you will violate the integrity of the traditional trombe wall,
which has a magical, wonderful way of *flywheeling*, and transporting the
heat through the wall, so it is available at the other side *precisely* when
it is needed, the next morning!" and he went on and on about this conceptual
delight, this conceit, completely ignoring numerical performance... :-)
trombe walls are also thermal disasters during long strings of cloudy days.
when the sun goes in for a week or two, they lose their stored heat in less
than a day, and then leak house heat badly, dramatically raising backup
heat or other solar thermal storage requirements.
i'm amazed that so many people, even in these newgroups, are still so
interested in trombe walls, or their passive solar equivalents, like
direct gain houses or high-thermal mass sunspaces. a lot of people are
apparently still willing to settle for high-cost, low-performance passive
solar house heating techniques, that get them a 30% yearly savings in backup
space heating costs over a 20 year payback period, vs. warmstores, solar
closets, sunspaces and transparent siding, which really can save close to 100%
of the space heating energy needed for a house and provide close to 100% of
the hot water needed for a house, as well.
>you'll find water is far and away the highest-capacity, non-phase change
>storage medium, in the form of freestanding tanks...
>how do you plan to provide for effective daylighting while preventing glare?
how about skylights with reflective south-facing sunscoops over the top?
or a transparent steep south roof with some reflective motorized dampers
in the insulated attic floor, or a bubble-ceiling?
>i'm presently looking for windows that block uv radiation...
you would want to look for windows made out of glass, or polycarbonate
plastic with a uv light transmission of less than 1%...
>but are not low-e (because of the amount of infrared radiation they block).
you seem to be seriously confused, will. for solar heating, blocking ir is
good, as is passing visible light... window glass and polycarbonate plastic
pass 85% of the solar radiation between 380 and 2,000 nanometers (the sun
being like a 10,000 degree f black body) and block more than 98% of the ir
re-radiation between 4,000 and 10,000 nanometers from an 80 f room. this
"greenhouse effect" has been around long before ozone concerns.
>uv will fade furniture
something fades furniture next to south-facing windows... is it the visible
light or the heat or the tiny amoutn of uv that gets through the glass?
at any rate, an intelligently-designed solar home (not yours, apparently)
will not have a lot of glass opening into the living space.
>and increase skin cancer likelihood.
i doubt you would ever get skin cancer with glass between you and the sun.
>i am in the middle of a similar design effort myself, and a utilizing the
>designs of a modular home manufacturer that has participated with doe
>on this subject (see solar today, sept/oct 1995).
kurt smith at avis? :-) he should know how to design better solar houses
by now... doe also sponsored the passive solar industry council guidelines.
perhaps we should abolish them, if they continue to do more harm than good.
>if anybody has any additional info, i'd be interested as well.
i'm trying to help you...
nicholson l. pine system design and consulting
pine associates, ltd. (610) 489-0545
821 collegeville road fax: (610) 489-7057
collegeville, pa 19426 email: firstname.lastname@example.org
computer simulation and modeling. high performance, low cost, residential
solar heating and cogeneration system design. bsee, msee. senior member,
ieee. registered us patent agent. fluent in french.