|
Vented Attics & Vented Crawl Spaces
(From: Building Science Corporation - Joe's Top Ten
:List of Dumb Things to Do in the South)
"My definition of an unvented attic is an attic where
there are no vents and where the attic insulation follows
the slope of the roof sheathing thereby including the attic
space within the conditioned building enclosure.
The
rationale for venting attics in the South is to "flush"
heat. The dominant heat transfer mechanism in an attic is
radiation. Venting attics will not "flush" radiation.
The air change in a perfectly built and vented attic (code
1:300 ratio) results in an average air change rate of 3
to 6 ach. At this attic air change rate there is approximately
a 2 to 3 percent reduction in heat transfer to the conditioned
space through the vented attic as compared to an unvented
attic insulated to the same level. This assumes an airtight
ceiling and no ductwork in the attic and certainly not leaky
ductwork in the attic. The moment ductwork (assumed airtight
in this instance and insulated at R-6) is installed in a
vented attic, the balance changes. There is approximately
a 5 to 7 percent increase in heat transfer to the conditioned
space as compared to my version of an unvented attic. This
is due to conductive heat gains through the surface of the
ductwork and air handler now located in a "hostile"
location (a hot, vented attic), rather than inside a 75°F
conditioned space (the "house"). The moment leaky
ductwork is installed in a vented attic there is approximately
a 25 percent increase in heat transfer to the conditioned
space. Of course this does not happen if you have airtight
ducts and an airtight ceiling (then the penalty for venting
the attic is only 5 to 7 percent as previously noted).
Now,
if you locate the ducts within the conditioned space and
also build an airtight ceiling, this is approximately 2
to 3 percent more efficient than my version of an unvented
attic. I never said that this wasn't the most energy efficient
way to do it. Of course when is the last time you saw ductwork
below an attic ceiling coupled with an airtight attic ceiling?
Builders put things in attics because they don't leave any
room in the house for the ductwork and air handler. If they
continue to do this, then venting attics is a dumb idea.
So
much for the energy concerns. Now lets talk moisture. What?
Are you all crazy? The air outside is hot, humid and disgusting.
And you want to bring this into an attic where it can diffuse
through the vapor barrier-less attic insulation and get
to the cold, air conditioned ceiling? What were we thinking!
Before it gets there it will see those cold R-6 insulated
ducts, fittings, etc. and drip all over. Give me a break.
Venting attics in the South was dreamed up by some disgruntled
Yankee pissed about the Civil War and wanting to get even.
Be sure when insulating at the roofline in humid climates
to follow moisture control principles as you would with
any insulated wall so that the roof assembly is self-drying
(www.buildingscience.com/resources/walls/exterior_sheathing_systems.pdf).
Lets
now talk about durability of shingles and shingle temperature.
Venting or non-venting a roof has about a 5 percent impact
on shingle temperature and roof sheathing temperature and
even less on shingle durability. The color of the shingle
is more important than venting or non-venting. And temperature
is less important than the shingle getting a sunburn. The
biggest impact on shingle durability is ultra-violet light.
UV is more critical than temperature. The best roof for
hot, humid climates for all applications (including unvented
attics) is a concrete or clay tile roof. Period.
Crawl
spaces are real simple to understand and deal with. When
you vent crawl spaces you bring in hot, humid air and cause
moisture and mold problems. The ground surface is typically
colder than the dew point temperature of the exterior air.
The underside of crawl space floor insulation is radiation
coupled to the ground surface and is very close to the same
temperature of the ground. Moisture droplets can be seen
all over the top surface of typical polyethylene ground
covers as well as hanging from the bottom surface of the
crawl space floor insulation. Gee, I wonder how all the
water got through the poly ground cover? It must have leaked
through the walls. Give me another break. Now, when the
moisture is in the insulation where do you think it wants
to go? Where is the air conditioning? Moisture moves to
the cold surface. Venting crawl spaces made sense only when
you had no air conditioning and no insulation and no crawl
space walls.
Interested
in unvented design strategies? It will take a while to change
common practice as builders and contractors learn to adopt
new approaches. Be sure to consult local experts and code
officials before attempting unvented attics or crawl spaces
on your own. Many design details that cannot be covered
here are important to achieve best performance. See Houses
That Work (www.housesthatwork.com)
for more information."
Back
To Top
|
House Design Recommendations by Climate Region
In
2001, the Building America Building Science Consortium developed
the Web-based Houses That Work (HTW) as a climate-specific
technical resource for designing and building homes that
are 30% or more energy efficient than the 1993 Model Energy
Code (MEC). The resource reflected the experience gained
from five years of Building America teamwork, including
insights gained during the construction of more than 8,000
production homes from across the country. Houses That Work
was and is a freely-accessible learning resource and reference
for builders, building product manufacturers, building researchers,
and the general public.
We call this Houses That Work II. HTWII is more comprehensive
and more detailed and represents the latest Building Science
Consortium experience and results under the Building America
program. Specifically HTWII includes:
-
Updated North American hygro-thermal regions map
that is aligned with the Department of Energy Model
Energy Code proposed climate zones and map.
-
Climate-specific Best Practices with performance
criteria for high performance home design and construction.
-
Three Building Profiles per climate.
-
A Building Materials
Property Table1 (populated with technical
and performance specifications from product manufacturers
and building research members of the Building Science
Consortium).
HTW
II has five sections - an overall introduction and a section
for each of five hygro-thermal regions: Hot-Humid, Mixed-Humid,
Cold, Very Cold and Hot-Dry/Mixed-Dry. The introduction
contains the hygro-thermal map and each climate section
contains Best Practices and Building Profiles. See the table
below for a quick reference for the components in each hygro-thermal
region's building profiles.
Climate/
Building Profile |
#
stories |
Wall
Type |
Foundation |
Wall
Cladding |
Roof |
Roof
Cladding |
|
Very Cold |
|
Aspen |
2 |
Stick-
framed |
Conditioned
crawlspace |
Fiber
cement battened panels |
Unvented
cathedral |
Standing
seam metal |
|
Concord |
1 |
Stick-
framed |
Full
basement |
Wood
siding |
Vented
unconditioned attic |
Asphalt
shingle |
|
Cold |
|
Beacon Hill |
1.5 |
Stick-
framed |
Full
basement |
Brick
veneer |
Unvented
cathedral |
Slate
roof |
|
Boston |
2 |
Stick-
framed |
Full
basement |
Fibercement
panels and battens |
Conditioned
attic |
Asphalt
shingle |
|
Chicago |
2 |
Stick-
framed |
Full
basement/
cast concrete |
Vinyl
siding |
Vented
unconditioned attic |
Asphalt
shingle |
|
Denver |
2 |
Stick-
framed |
Full
basement w/ sub-crawl |
Brick
veneer/wood siding |
Vented
unconditioned attic |
Asphalt
shingle |
|
Minneapolis |
1.5 |
Stick-
framed |
Slab-on-grade |
Stucco |
Vented
cathedral |
Asphalt
shingle |
|
Vineyard |
2 |
Stick-
framed |
Full
basement |
Cedar
shingle siding |
Unvented
cathedral |
Cedar
shingle |
|
Mixed-Humid |
|
Atlanta |
2 |
Stick-
framed |
Slab-on-grade |
Fibercement
siding |
Vented
unconditioned attic |
Asphalt
shingle |
|
Charlotte |
2 |
Stick-
framed |
Conditioned
crawl/ block wall |
Brick
veneer |
Vented
unconditioned attic |
Asphalt
shingle |
|
Louisville |
1 |
Stick-
framed |
Full
basement/
cast concrete |
Vinyl
siding |
Vented
cathedral |
Asphalt
shingle |
|
Hot-Dry/Mixed-Dry |
|
Albuquerque |
1 |
Stick-
framed |
Slab-on-grade |
Stucco |
Vented
unconditioned attic |
Asphalt
shingle |
|
Sacramento |
2 |
Stick-
framed |
Post-tensioned slab-on-grade |
Fibercement
siding |
Vented
unconditioned attic |
Tile |
|
Tucson |
1 |
Stick-
framed |
Slab-on-grade |
Stucco |
None |
Low-slope
membrane |
|
Hot-Humid |
|
Houston |
2 |
Stick-
framed |
Slab-on-grade |
Brick/
Fibercement |
Conditioned
attic |
Asphalt
shingle |
|
Maitland |
2 |
Block/
stick frame |
Block
stem wall & slab |
Stucco |
Conditioned
attic |
Tile |
|
Orlando |
2 |
SIPS |
Block
stem wall with slab |
Cement
board siding / brick veneer |
Conditioned
attic |
Metal
roof |
|
Montgomery |
1 |
Stick-
framed |
Conditioned
crawl/ cast concrete wall |
Vinyl
or aluminum lap siding |
Vented
unconditioned attic |
Standing
seam metal |
Back
To Top |
Unvented
Attic Discussion
"Sometimes you just have to color outside the lines.
Unvented attics? You've got to be kidding. Well, actually,
no. Unvented attics actually make a lot of sense. In humid
climates, venting attics brings a great deal of moisture into
the structure. Not venting makes this problem go away. In
cold climates, venting attics brings in a great deal of snow.
Not venting also makes this problem go away. In roof design
with complex geometries, venting roof assemblies can be extremely
difficult. Not venting makes this problem go away. Ever try
to install an air barrier in a complicated roof system? Even
in hot-dry climates, not venting attics can make sense. Of
course in all of these cases you have to know how to do it
right.
What about moisture? What about shingle temperature and sheathing
temperature? What about the energy costs? What about the code?
Yeah, yeah, yeah, and yeah. O.k., everybody take a valium.
We are not about to violate the laws of physics here; we are
actually going to use them to our advantage.
Back To Top
|
Moisture
and Attic Ventilation
We traditionally vent attics to prevent moisture build up
in roof sheathing. Where does this moisture come from? In
cold climates the moisture source is from the inside. In hot-humid
climates the moisture source is from the outside.
In cold climates, building unvented roof assemblies requires
understanding the dynamics of moisture. The key is roof sheathing
temperature. If the underside of the roof sheathing, typically
the first condensing surface is kept above the dew point temperature
of the interior air-vapor mix, condensation and moisture accumulation
will not occur.
Not venting in a hot-humid climate to control moisture build
up in roof sheathing is a no-brainer. Well, maybe a
half-brainer. You have to build your roof assembly right.
In houses in this climate with asphalt shingle roofs, the
morning dew collects on the outside of the roof, wicks up
between the shingles, and then the moisture gets driven through
the roof sheathing into the attic by "solar thermal drive"
(remember, that's moisture moving from hot to cold). The typical
asphalt felt under the shingles is permeable enough that the
moisture heads right through, where it can cause problems
like (at least) adding to your air conditioner's latent load,
or (worse than that) growing mold on your roof sheathing.
So what's the solution? Simple-if you're using asphalt shingles,
you have to use an impermeable underlayment-yup, a vapor barrier.
Since polyethylene is difficult to walk on, we recommend either
self-adhesive membranes or a low permeability roofing "paper"
(actually a type of housewrap) such as Flexia Tri-Flex@ 30.
Note that other roof materials (tile, metal) do not suffer
from this problem. Tile typically has a vent space beneath
and is usually installed over tar flood-coated roofing paper
(yup, a vapor barrier). And metal roof cladding is not a reservoir-it
can't store water. Wood shingles and shakes, on the other
hand, do have a similar problem (See "Unvented Roofs,
Hot-Humid Climates, and Asphalt Roofing Shingles").
In hot-dry climates with asphalt shingles we have a reservoir
but we don't have the moisture. And in hot-dry climates with
tile the question is moot -we don't have the moisture and
the tile is back vented.
Air sealing for moisture control and durability in an unvented
roof assembly is also important, particularly under any conditions
where there is moisture-laden air (Translation: humid climates
where air leakage can introduce moisture into the roof assembly
from the outside and cold climates where air leakage can introduce
moisture from the inside).
Air leakage can be a really powerful mechanism for moisture
transport; its control requires careful air sealing details.
In our experience, that means spray foam, particularly at
junctions (like soffits where the roof and wall planes meet
(Photograph 1) and ridges where the two roof planes meet Photograph
2). Do you have to spray foam to full roof framing cavity
depth-no, this is an air sealing detail, not an insulation
detail. Can I count on the air flow restriction afforded by
a cavity insulation such as fiberglass batt or even dense-pack
cellulose-no, this is an air sealing detail (Even dense-pack
cellulose passes air-cellulose is a convection suppressor,
not an air barrier). Can I get sufficient air sealing without
spray foam-yes, blocking and caulking can work but depending
on the pitch and complexity of the roof system, it can be
a bugger. The real test? After you think you have an air-sealed
unvented attic assembly, do a blower door test twice-once
with the attic hatch open and again with the attic hatch closed.
If you get a significant difference, think (and air seal)
again.
Photograph
1

|
Photograph
2

|
Back To Top
|
Shingle
Temperatures and Shingle Life
What about shingle temperature? Well the obvious answer to
that is don't use asphalt shingles. They're a dumb idea anyway.
They burn. They're an energy heat gain nightmare. They are
sensitive to ultra-violet light and can't be made to last
more than 15 to 20 years despite what the warranty says. Anybody
out there ever collect on a shingle warranty? They also off-gas
horrible stuff. Hail just kills them. But they are cheap.
And in cold climates, they are the roof coverings of choice.
In most hot-dry and some hot-humid climates, builders use
concrete or clay tiles so the issue becomes moot. Ditto for
steel, copper and wood roofing. Constructing unvented roof
assemblies with these types of roof coverings is not a problem.
With asphalt shingles, the operating temperature of the singles
increases slightly, on the order of 2 to 3 percent (not to
mention the moisture issues in hot-humid climates). That means
that a black asphalt shingle roof that is typically at 150
degrees F, now will be at 153 to 155 degrees F .
However, that 3 to 5 degrees F increase is important, since
it translates into an approximate 15 percent reduction in
the useful service life of the shingle. Where does all this
come from? Well, a good rule of thumb in physics and materials
science is that for every 10 degree C (18 degrees F) increase
in the temperature of a material, you double the chemical
potential. Potential for what you ask? Potential for bad things
to happen is what! Like a decrease in useful service
life. Divide 3 degrees by 18 degrees and you get around 15
percent. On a 15 year shingle roof, that means you lose 2
to 3 years in the service life of the shingle.
Why is there only a 3 to 5 degree F increase in shingle temperature?
Shouldn't it be much higher? Actually, no it shouldn't and
it's not. Heat is transferred three ways: convection, conduction
and radiation. Radiation is the dominant factor in roof assemblies.
Venting your roof does not affect the radiation heat transfer
(well, not by much). And the under side of the roof sheathing
is not designed as an efficient plywood-to-air heat exchanger.
When you measure the temperature of the air going into the
roof and the temperature of the air going out and look at
the overall air change (i.e. the mass flow rate), the heat
removal by ventilation is pretty pitiful compared to the heat
into the assembly from solar gain.
Now remember that all this is based on some simple rules of
thumb. But all of this is backed up
by some real consistent field observations. I have about 1
,000 unvented roofs with shingles under my belt Most of them
are in CANADA. Yeah I know, the laws of physics are different
up there.
But a lot of them are in NEW ENGLAND, MICHIGAN and COLORADO.
Over a third of them are now over 10 years old and they are
doing fine. These roofs are constructed from rigid foam sheathing
sandwiched between plywood or OSB. My own house is built this
way (but not my mother-in-law's, we won't go there!).
Back To Top
|
Space
Conditioning Energy Use and Building Durability
Remember there is a trade-off. I choose to give up the 2 to
3 years in the life of my shingles in exchange for better
performance for the entire system. Not all people will want
to make this choice. I already see their logic: Yes, I prefer
mold in my house in Orlando in exchange for 2 to 3 years more
on my shingle life; and yes, I want my shingles to be black
or brown in Austin so that I can install a 1 to 2 ton larger
air conditioner. Of course, there are people who also root
for the Cubs, but we are slowly getting them out of the gene
pool. These are also the same people who vent crawl spaces,
but that will be the subject of another article.
Now, where the real effect of not venting roof assemblies
is felt is the temperature of the
underside of the roof sheathing. Our field measurements and
computer modeling show that the temperature of the under side
of the roof sheathing increases between 10 and 20 degrees
F. Why the huge difference here and not in the shingles? Well,
compare the R-value of a shingle and the R-value of roof plywood.
A temperature gradient can actually exist across the plywood.
And, ventilation air on the underside of the roof plywood
does remove heat. Take away the ventilation air, and you do
increase the temperature of the underside of the plywood.
But does this matter? Yes and no. Depends on the overall system
design.
Not venting roof assemblies in most climates increases the
air conditioning load on a typical home by approximately 3
to 5 percent. If the ducts are inside the conditioned space
and we are not worrying about mold, humidity, ice-damming,
or blowing snow issues, it's better to vent the roof and not
give up the 3 to 5 percent.
But if you are stupid enough to put ducts in attics, if you
are stupid enough to put air handlers in attics, if you are
stupid enough to hire interior designers and architects that
design incredibly complicated roof structures that cannot
be air sealed at the interior drywall because of jigs, jogs,
shelves, coffers, pot lights, valleys, hips, dormers, beams,
skylights etc. and etc. and etc., give up the 3 to 5 percent.
Guess what you gain? You gain between 10 and 30 percent savings
from the airtightness of the roof sheathing and lack of conductive
gain on ductwork in vented attics.
In Las Vegas, we are averaging heating and cooling utility
savings of 20 to 30 percent with our unvented roof designs.
Bye-bye duct leakage. Oh, the ducts still leak, but they don't
leak to the outside. Bye-bye conductive gains on the ducts.
The ducts are inside 80 degree attics. Yeah, but the surface
area is larger and the attic volume is added to the house.
Big deal, those trade-offs are nothing compared to the ease
of constructing an air tight roof. The roof sheathing is now
the air barrier (or the pressure boundary if you prefer that
term).
In the humid south you also gain the humidity control. In
fact, in the humid south you do not have any other intelligent
choice. Unvented is the way to go. We've been fixing moldy
houses in the humid south by turning vented attics into unvented
attics for over 15 years.
Guess who hates all of this? The shingle manufacturers and
the roof vent manufacturers. Duh."
Back To Top
|
|
|