How Heat Moves Through Homes — Building Science Podcast
Space-age radiant barriers work great in the vacuum of outer space, but here on Earth, heat moves by conduction, convection, and radiation, simultaneously, all the time.
In our last episode, Dr. Joe Lstiburek talked about efflorescence and the serious damage that water and salt can do to masonry. This week, Dr. John Straube explains how the three forms of heat flow work, and debunks the claims of a few common insulating materials.
Comfort is the Primary Purpose of Buildings
So, I actually call this “thermal control, insulation, and thermal bridges,” because thermal bridges are important. Let’s go back again and say, “Why do I want to control the heat flow?” Well, there are a whole bunch of reasons, but we often forget that the No. 1 reason we want to insulate is for thermal comfort. So, we want to make sure that exterior walls, roofs and floor slabs — that the temperature of these components stays above 68° in the wintertime and below 78° in summertime.
If the wall or roof or slab gets too far beyond those ranges, we feel discomfort regardless of what the thermostat says. Actually, a lot of interesting buildings that work well that don’t have a lot of insulation — say, the old five-wythes of masonry built in 1864 or something like that — they work not because they’re that well insulated if you put them into our value test; they work because they maintain the surface temperatures at a relatively high level in the winter and are relatively cool in the summer. Even though the air temperature may not be that well controlled, you feel comfortable in them. And so just maintaining comfort is the No. 1 reason why, when we moved to lightweight assemblies, we had to start adding insulation; it was just too darn cold.
Windows are tough to keep warm
Now, we still have that problem in our windows and our doors. So sliding patio doors — large areas of R-2 — well, they’re not very comfortable either in hot weather or in cold weather to sit next to. They just don’t meet the test of providing sufficient thermal control for comfort. We’re not even talking energy. So, they fail the first test.
We also want to control surface and interstitial condensation. Now that we are quite clear on “condensation occurs on cold surfaces,” we can also now look at the psychometric chart and say, “So, how cold is cold?” Well, it depends on interior relative humidity, right? And if we keep the relative humidity in, say, the 30% range in cold weather, that means the surface temperatures have to be above 35 or 40 degrees. If you’re in the 40% relative humidity range, you now have to keep interior surface temperatures above 45 or so degrees.
And again, where do we get that surface condensation? On our windows, because our windows are the least insulated part of the building enclosure. So, we see that first. But we also can get it in corners and at thermal bridges.
R-5 meets most comfort and condensation concerns
The next thing is to save energy. But really, our struggle first and foremost in terms of stopping the biggest problems is: Make sure we get comfort; make sure we don’t get surface condensation. That’s what we need to have minimum levels of insulation for. And actually, R-5 will solve those two problems by and large in most climates. R-10, and you’re a slam dunk — even though we see codes with R-40s and R-20s and whatever. If we could just get R-5 everywhere, we could solve 95% of comfort and condensation problems — and R-10, we’d solve 99% of them.
But we can’t do that because we can’t even achieve R-5 and R-10. Saving energy, reducing operating costs and pollution is what the building codes are worrying about when they specify R-values — and we’ll find out how useless that is as an approach to saving energy as we work through this presentation.
Energy-efficient buildings need smaller heaters
Then we have to save distribution and heating plant costs. Most people seem to forget that if I actually do a good job of controlling heat flow across my building enclosure, it means I can reduce the size of my air conditioner, the size of my furnace, the size of the ducts, the size of diffusers, the size of the fans, the size of the fan motors, the size of the filter, the size of the space it requires to put all of that into a building.
So, we see a lot of life-cycle cost analysis of insulation and air tightness and such, and almost never do they include the escalating cost of energy — and they never include the rather significant savings that can be achieved in terms of downsizing your air conditioning plant, or your furnace or your ductwork. When you start reducing the size of ductwork or even eliminating it almost entirely, well, that opens up architectural opportunities: “Oh, you mean I don’t have to put in a suspended ceiling here? You mean I don’t actually have to take the duct to the outside wall?” Well, of course you don’t. If you insulate your walls sufficiently and provide good windows, you don’t have to provide a duct to the outside wall; you can provide the heat at the inside wall and simply blow the air over there. If you have a piece of junk wall and a piece of junk window, yeah, not only do you have to put the vent right below the window, you’re still going to be uncomfortable. But the duct savings are significant; it’s just that they’re not usually reaped.
Meeting codes doesn't guarantee good performance
Now, decreasing load diversity is something that matters to commercial buildings; less so to residential. Load diversity is the term used to describe the fact that at one part of the building you may have a significant cooling or heating loadRate at which heat must be added to a space to maintain a desired temperature. See cooling load., while at the other part of the building you have the opposite. It’s very common in commercial buildings, for example, to have the east side require cooling at 10 o’clock in the morning. The sun is shining through the east windows, everyone has moved into the office, it’s all heated up; whereas on the west side, they’re still running the boiler to keep it hot.
The only way that situation could actually occur is if you have a very bad building enclosure. And yet it’s pretty common; it’s a pretty common scenario. And that’s not just energy wasting; that’s annoying for the controls people have to work out. It’s annoying that you have to have systems that can do cooling and heating at the same time — what a pain in the butt. And yet by building bad building enclosures, we have managed to create that as almost a standard in office buildings and schools, etc., in our part of the world.
And of course there’s the “I don’t wanna go to jail reason” to meet codes and specifications — which I always put at the bottom of the list. But actually, most designers that we work with, they put that as No. 1: “I don’t care if it matters; we’re just going to meet the codes.” And building code officials who say, “Look, just put R-20 in there. I don’t care if it actually does anything; jut put it in there to meet the code.”
Heat and how it moves
The Basics of Heat Flow
As construction methods and materials change, and energy gets more expensive, how and why we insulate our homes become more important.
- For thermal comfort
- To save energy
- To stop condensation and the potential for mold and rot
- To reduce the size, cost, and complexity of our HVAC(Heating, ventilation, and air conditioning). Collectively, the mechanical systems that heat, ventilate, and cool a building. systems
In order to slow the flow of heat through our foundations, walls, windows, and roofs, it helps to understand what heat is and how it moves.
What is heat?
- Heat is energy in the form of vibrating particles
- The faster the particles move, the farther they move apart — slow particles make solids, faster ones turn to liquids, and even faster ones become gases
How can it move?
- ConductionMovement of heat through a material as kinetic energy is transferred from molecule to molecule; the handle of an iron skillet on the stove gets hot due to heat conduction. R-value is a measure of resistance to conductive heat flow.: solid things touching other solid things — drywall touching wall studs touching plywood
- Convection: fluids, like water or air, moving around in an open space, like a pipe, wall cavity, or room
- Radiation: heat in the form of electromagnetic energy moves through open space — the less stuff in the way, the better
So, to understand a little of this, we’re going to go through the three modes of heat transfer: conduction, convection and radiation. Those three modes of energy transport are acting all the time and in parallel; they don’t really interact that much. We have to look at each of the three to understand the total picture.
First, we should understand the nature of heat. Heat is vibratory energy in molecules. At any temperature that we’re experiencing, the molecules are vibrating. If you put more heat into them, they vibrate faster. In fact, if I take a piece of stainless steel and I put so much heat into it that I get to about 2,000 degrees, it’ll actually vibrate so much that the whole thing falls apart into a liquid. And if I keep heating that up, it’ll boil and the molecules will fly off into a vapor.
So, whatever the heat is, it’s vibratory molecules, and when we talk about conduction, we have hot and cold as one solid object, and the fast-vibrating hot molecules physically collide with the slower-moving cold molecules and transfer momentum. By transferring some of the speed from the hot to the cold, the hot gets colder, and the cold gets warmer; that’s how heat is transferred by conduction.
Conduction requires contact and low R-values
To have this mechanism work, you need solid material so that the molecules are directly in contact. Even having a gap of an eighth of an inch, you stop conduction. You still have conduction through the air, but the two solids have stopped. Our most common experience with this is metallic materials. You have a frying pan — cast iron — you put your hand on it. If it’s been sitting in the fire too long, the heat will conduct up to the handle and it’ll be hot. What we do to stop that, of course, is we put a piece of wood around the handle of the frying pan because the wood is less conductive.
Now, the way we measure the performance of conduction is something we call conductivity. And conductivity is a material property, like vapor permeability is a material property. The symbol is typically K, but in Japan and Germany, they use the Greek symbol lambda. We can figure out what the resistance is. By figuring for a specific thickness, you get a conductance, and from the conductance, you do an inverse and you get R-value. That measures resistance only.
Unfortunately, if this is the science behind it, the R-values that we see installed on our Styrofoam SM or on fiberglass insulation, they actually also include convection and radiation (we haven’t talked about that yet). So, the R-values we use for insulation products aren’t really scientifically based on conduction.
If air or water move freely you get convection
Convection is the movement of the hot molecule here to over here by physically grabbing the whole bundle of molecules and dragging it across. So, it’s a mass flow movement of grabbing the hot fluids, which could be water from the boiler, and transporting it to the second-floor bedroom; or it could be a hot furnace, and transporting it to the kitchen — but just physically moving the fluid. That’s convection.
Convection, actually, is more important in our building assemblies than most people realize — and it’s because our building assemblies have gaps and openings in them and we use materials that allow air to flow through them. And so when we look at something like a fiberglass batt that’s been installed from the inside, as I push this friction-fit batt, these fibers get pulled backward, right? They have to because of friction. Then that causes the fibers to stick out here, so the installer tucks the batt in like that. That’s your standard thing. As the installer tucks the corner in, it compresses the batt insulationInsulation, usually of fiberglass or mineral wool and often faced with paper, typically installed between studs in walls and between joists in ceiling cavities. Correct installation is crucial to performance. and often causes a wrinkle at the middle.
How do we know this? Well, we get people who install fiberglass batts with Plexiglas sheathingMaterial, usually plywood or oriented strand board (OSB), but sometimes wooden boards, installed on the exterior of wall studs, rafters, or roof trusses; siding or roofing installed on the sheathing—sometimes over strapping to create a rainscreen. and we look at the far side and can see all these little gaps and openings form. So, this is what it really looks like.
Batt insulation rarely works well in the real world
The question is: What happens at these gaps? Well, this will be filled with hot air because it’s on the warm side; this will be filled with cold air because it’s on the cold side. Hot air rises and cold air falls, and so we have these two micro-ducts on either side of the fiberglass insulation. As hot air rises, it goes through the insulation and comes down and goes around like that. And that transports energy.
How much energy? Well, it depends on how big those gaps are. Twenty, 30, 40 percent of the total heat flow across the wall can be by this mechanism. Which means your R-20 batt will drop to R-12 or R-10.
What’s neat about this mechanism is that as the temperature difference across the wall increases, there’s a greater and greater proportion of heat flow transported by this mechanism. As you need the R-value more and more, it actually drops. Which is why when they test R-value in labs, they never test it like this because then they wouldn’t get R-19. So, what they do is they turn it on its side and they make darn sure there are absolutely no gaps — no studs, no gaps. They also make sure that the temperature difference across the batt is never more than about 30 degrees. That way, all this shit can’t happen and they can just get good R-value numbers.
Now, if you take an R-19 batt and have the audacity to install it in a vertical application, your R-value goes down. If you then have little gaps — like batts are actually installed — well, then your R-value goes down. The bigger the temperature difference, the more the hit is; it’s a nonlinear relationship.
But it gets worse. Not only do we have the gaps and openings around it, we have the studs in between that, and of course the heat flows more easily through the studs than it does through the batt. So, who cares about the R-19 of the batt when the heat’s actually flowing through the triple studs around the window? And the steel studs?
With wood studs, you can argue that if you were a perfect tradesperson, you would install with no gaps at all. If you were given the three minutes per batt to install — most installers of course use at least three minutes to install a batt — you could probably make this work. But if you put it in a steel stud, the lips on the steel stud guarantee that no matter what level of trade quality is provided, you will have gaps around that steel stud. And those gaps will be the size of that lip — both 3/8 of an inch by an inch and a half wide. So, you’re guaranteed to lose between 30 and 40 percent of your R-value just right there. Isn’t that cool? Or hot, depending on the weather really.
This same loopy stuff can go on in wintertime around insulation installed in attics, because there’s nothing stopping the air from looping around up here, and if there are any voids around the rafters or joists, you get loops going on. In some cases, when the temperature drops to 10 degrees, some types of blown-in fiberglass (which mostly aren’t sold anymore) at 20-degree outside temperatures were getting half the rated R-value, and at 0 degrees, which does happen on the coldest day of the year in northern Minnesota, they were getting something like 30% of the minimum rated R-value.
So, you’d blow in R-40 and you’d get R-15 on the coldest day of the year, which of course is precisely when you need the most R-value. By the way, the way they fixed that these days is you go in and blow 4 to 8 inches of cellulose over top to provide a kind of air-impermeable cap on top of the fiberglass. So, if you do have problems, it’s often quite convenient and inexpensive just to blow a foot of cellulose on top and you’ll get a major improvement in performance.
Radiation likes empty space
Then we have radiation. These vibrating molecules create waves in space-time, which we call electromagnetic radiation. At the temperatures that we’re talking about, they’re infrared radiation.
If you were to make these molecules move fast enough, they would eventually glow red-hot; you’d be able to see them. They’d go from infrared to actual red. And if you keep heating them up, they’d get white hot. And if you kept heating them up, they would actually start giving off ultraviolet radiation — and then a nuclear explosion, you’d get gamma radiation, they’re so hot.
But most of our building applications aren’t worried about that. We’re worried about infrared radiation. That’s what infrared cameras look for: They look for the emission of radiation given off by the temperatures that the molecules have. Now, for radiation to be important, it really likes to transfer through — not solids; it likes to go through voids. It doesn’t even like to have gases in the way. That’s why the sun is able to transfer its energy from 93 million miles away to the planet Earth; because basically there’s nothing between us and the sun other than vacuum, except for the last 30 miles or so. Not even — it’s really only about 5 miles of air between us and the sun. So, radiation is quite effective. About 90% of the radiation given off by the sun hits the planet’s surface. We’re trying to change that, of course.
Now, what you need is a gap. If you have aluminum foil, which does not emit radiation very well, it does not change the heat flow across the building assembly unless there’s an air gap. So, you need a gap.
The Thermos bottles — they’re shiny glass on the inside, and that’s so there’s an air gap, and the shininess deals with the radiation. If you filled that void up with foam, the R-value would go down, not up, because you would eliminate any radiation benefit of the shiny metal. So, you have to have the gap; a slightly bigger gap would be good.
Now, within the pores of insulation like fiberglass or rockwool, there are so many voids that radiation actually does play a role in jumping from fiber to fiber inside that product. And in a fiberglass batt, about 40% of the heat transfer at common temperatures is due to radiation. Foam, it’s about 30%. Now, the reason that matters is that as the temperature changes, the contribution of radiation changes.
We’ve all probably been around a fire on a cold night; watch the campfire burn and you can feel the heat radiating to your face. That’s because it’s hot. It really makes a difference whether that fire is hot or cold whether you feel the radiation on your face or not. As you get down to building-related temperatures — 100 degrees, 50 degrees — radiation gets less and less important.
But at high temperatures, radiation is important and it’s a major transfer mechanism; at low temperatures, it doesn’t play as big a role. So, radiant barriers are very good for high temperatures — say, the roof in a sunny climate. They’re less important for cold conditions — say, the underside of a crawlspace; they don’t play as big a role. But in every case they need an air gap.
NASA's radiant barriers are useless when you pour concrete over them
With radiant floor heat, it is actually kind of misnamed. There is radiation transfer, but actually most of it is by convection. So, convection matters, radiation matters, but more importantly, when I think of radiant floor heat [I think of] snake-oil salesmen who sell these radiant barriers underneath radiant slabs. Radiant-radiant, right? They should go together; they’re both named radiant.
But of course, when you pour concrete on the aluminum foil, there’s no air gap, is there? So, there’s no R-value benefit. The R-value of a piece of aluminum foil underneath a chunk of concrete on top of soil is around 0 — somewhere between 0 and bupkus. However, they don’t test them that way, do they? They test them in horizontal apparatuses with an air gap above and an air gap below with ridiculous temperature differences across them. And then they get, like, R-8. But it’s hard to suspend that slab 4 inches above the radiant foil in most of my radiant slab designs.
So, what they’ve done to address that is they put the little bubble wraps — the radiant foil bubble wraps, and so on — and they have, well, tiny bubbles (Don Ho sang about that until his recent death). These tiny bubbles in between the aluminum foil do help the R-value, and you can get as much as R-1 on some of the bigger bubble products.
Now, the cost of that R-1 bubble wrap per R is about 3 times the cost of buying extruded polystyrene foam, but you can market this stuff as “space age.” Well, NASA uses it. OK, let’s think about this. I’m in outer space. There is no air. So, what are the heat transfer mechanisms? Well, there’s no convection; there’s no air. All I’ve got is conduction and radiation — so, if I’m not touching it, of course there’s only radiation. NASA uses radiant barriers because radiation is the only way they can transfer heat from them to other spatial bodies. It’s the only mechanism that works. '
But as long as you’re building your buildings on Earth, in an air-filled environment, there are other mechanisms that are actually more important. But the NASA technology and the “ceramic balls” — it’s all just bullshit.
But somehow they manage to sell this stuff by playing on people’s ignorance. They’re not sure about how all this works.
Apr 14, 2010 10:07 AM ET
Apr 14, 2010 4:06 PM ET
Apr 14, 2010 4:47 PM ET
Apr 14, 2010 9:29 PM ET
Apr 15, 2010 7:24 AM ET
Apr 16, 2010 12:41 AM ET
Apr 19, 2010 6:28 PM ET
Apr 19, 2010 8:03 PM ET
Apr 19, 2010 8:06 PM ET
Apr 19, 2010 8:52 PM ET
Apr 19, 2010 8:58 PM ET
Apr 19, 2010 9:52 PM ET
Apr 19, 2010 10:03 PM ET
Apr 19, 2010 10:24 PM ET
Apr 19, 2010 11:09 PM ET
Apr 19, 2010 11:48 PM ET
Apr 20, 2010 1:37 AM ET
Apr 20, 2010 3:33 AM ET
Apr 20, 2010 8:26 AM ET
Apr 20, 2010 12:13 PM ET
Apr 20, 2010 1:04 PM ET
Apr 20, 2010 1:05 PM ET
Apr 20, 2010 3:22 PM ET
Apr 20, 2010 8:34 PM ET
Apr 20, 2010 11:44 PM ET
Apr 21, 2010 1:25 AM ET
Apr 22, 2010 11:31 AM ET
Apr 24, 2010 8:58 AM ET
Apr 24, 2010 9:20 AM ET
Apr 28, 2010 10:43 PM ET
May 9, 2010 3:28 PM ET
May 10, 2010 6:12 AM ET
Jul 10, 2010 6:00 AM ET
Jul 10, 2010 6:26 AM ET
May 3, 2012 9:17 AM ET
May 3, 2012 9:25 AM ET