Window R value in calcs?
If the info in “Windows as good as walls”, etc, is true (and I have no reason to doubt it), what does one use for an R-value when making rough heat loss calcs for a house? If an entire window has an R of 5, say, then it’s effective R value is somewhere between +5 and -5, is it not? I am just using the standard normal Q = UA (delta T), as I don’t know how to calc solar heat gain, and it seems to oversimplify what a good window does to treat it like an R-5 wall section. Is there some grossly approximate rule of thumb to use for said window’s R value in a sub-arctic environment? The above referenced article mentioned some windows gaining about as many btu’s as they lose, but I did not see what percent of those gained btu’s are usable, ie garnered when they are needed, and not just let back out the open window because it is warm outside. Am I looking for another holy grail? The house harboring these windows would be in a fairly high, fairly sunny location; nothing deep in tangled black spruce, and nothing on the top of Shaktah Ridge. Thanks. klingel
GBA Detail Library
A collection of one thousand construction details organized by climate and house part
Replies
Thank the Creator that nothing can have a negative R-value. I'm not sure what you mean by +5 to -5 effective R-value (or even what you mean by "effective" R-value).
But for heat loss calculations, all you need is whole unit R-value and area and HDD and design minimum temperature.
Solar gain calculations are done separately and require a bunch of other variables, including average daily solar availability, window orientation, window SHGC, shading coefficient for the site, glazing to floor area ratio, glass to mass ratio, etc.
Robert: OK, and thanks again. That "+ or - 5" was a misguided attempt to get at "either they lose (+R) or gain (-R), ie, they suck heat IN". "Effective R" would mean that "although they are R5 in the lab, in a house that hole only has a net loss of XX btu/yr, the same as a wall that size with an R of YY." I guess I was a bit obscure w/ my terminology.... I tried to edit that whole deal, but the spam filter cop told me I was attacking too much, so it stopped me. I know that solar heat gain is done entirely separately, but I was hoping someone who has done it a lot has realized that "in cold climates, using a window R of 15 to 25 gets you pretty close". In other words, a very crude model could be used instead of all the complex and expensive calcs w/ sun data. I wouldn't care to know a window is going to lose 2,409 btu/yr, but " 2 to 3 K is about right" would suffice. Well, I was just hopin'. klingel
John Klingel,
The original Canadian ER method for rating windows was one attempt at achieving the goal you seek. A window with a -5 rating loses more heat in a year than in gains; a window with a 0 rating is energy-neutral; and a window with a +2 rating gains more heat than it loses.
The ER rating system has since been changed to make all of the values positive. (Marketing gurus hate negative numbers because they are hard to explain to shoppers.)
Read more here:
Windows That Perform Better Than Walls.
Choosing Triple-Glazed Windows.
Robert is quite right, however, when he points out that this discussion has nothing to do with R-value.
Martin: Thanks. I think the old Canadian system may be worth exploring. There must be some way to generalize how a specific window behaves thermally in a specific region. If its R value predicts a loss of 2000 btu/yr, but a region's "general" heat gain during the heating season indicates that it would gain 1500 btu/yr, then it is a simple matter to calculate its "effective R". I am going to have to dig into this solar heat gain stuff and see if there aren't some "gross averages"; every discipline has them, though these may be so gross they are useless, practically speaking. Dunno. Thanks again. klingel
Sorry John but you won't find any generalizations that are worth using (with the possible exception of the Canadian ER ratings).
The heat loss of a window is fixed by the unit U-value and the HDD climate for annual energy use, or minimum design temperature for calculating maximum design heat loss for sizing the heating system.
Passive solar heat gain is a local phenomenon, since it's dependent on the particular site (both exposure and shading), on the window orientation (in both altitude and azimuth angles), and on the house shading elements (such as overhangs and porches and ells). And the usefulness of that solar gain is dependent on glass to floor area and glass to mass ratios.
And don't confuse heat loss/solar gain ratios with "effective R-value". R-value is an inherent property of a material or assembly - the resistance to heat flux. Heavy thermal mass can change the steady state R-value of an assembly into a dynamic "effective" R-value because the mass effects the speed and timing of heat flows.
But to suggest that the sun changes the "effective" R-value of a window is like saying that the thermostat setting (internal heat level) changes the "effective" R-value of the window since more heat is lost.
Robert: Yes, all that is true; solar heat gain is complex, and obviously very dependent upon the angle of the sun, duration that it hits a window, etc. And, no, the sun does not significantly change the physical properties of a window; its R value is what it is, set at the factory ad infinitum. But, what the sun does do is change how much NET energy goes out of a window, or in. I am apparently not effectively articulating my point, so let me try again. My point is this: A window with R5 will lose, say, 2000 btu/yr if the sun never shines on it. That is set in stone, at the factory. If that window takes in 1500 usable btu/yr, then it's "effective" or "felt" R value is 20. A good window is not a wall and can't be treated as such when one is simply using the equation I mentioned just to get a rough idea of what a house's actual heat loss will be. I don't have the time to learn all about solar heat gain, but am seeking a gross approximation thereof. I think there is an answer to this, despite it being unimportant in the grand scheme of things and despite it being rough. The windows are going in regardless. Again, thanks for the reply, though. klingel
John,
You are mistaken when you state that a certain window "with R5 will lose, say, 2000 btu/yr if the sun never shines on it. That is set in stone, at the factory."
In fact, the heat loss through this hypothetical window depends on the climate where it is installed (the exterior temperature) and the interior conditions.
Oh, yes -- and one more factor: the amount of reflected sunlight, if any, that it receives. (Even if the window is shaded, reflected light can provide heat gain -- especially in snowy climates.)
John, is this part of a cost/benefit calculation to help you choose between different windows or are you trying to figure out how much glazing area to use?
I read somewhere once to take a 10% bonus on the energy received at the outer surface of the window for this reason... Could it be that high?
Martin: I think you are looking at this too intensely, or too specifically. I am not trying to calculate heat loss for the world with one window. Of course a window in Fla is not subjected to the delta T we have in Fairbanks, and of course there is some incidental reflected light, even here when we are down to 3.75 hrs of sun per day. But that reflected light is trivia. Again, I'm not looking for 1,983 + or - 2 btu/yr (or hour or minute) accuracy, just a ballpark. I just picked the number 2,000 out of the air for illustration. Fact is, my spread sheet says my 184 sf of windows will lose 4,269 btu/hr, using a design temp of -51, or a delta T of 116 F..
Lucas: Thank you for the reply. That is what I am looking for. With that, I would suggest we just close this topic and move along. I don't seem able to get my idea across, so let's just drop it, OK?
The sun reflected off snow-covered ground gives a 39% bonus (4% on bare ground) on a vertical glass surface. Which is why vertical glass is far better for a sunspace in snow country than the 60° tilted glass that used to be popular (which offers only 30% bonus for both angle of incidence and ground reflection).
John,
How do you figure your design minimum temperature is -51? Design minimum temp isn't the lowest temp ever recorded, it's the lowest annual expected temperature ignoring the 2.5% coldest days. And design minimum temp is only useful for sizing your heating system. For figuring out annual heat load, you use average heating season temp, which looks like more like 10°F for Fairbanks.
I know what you're trying to do, but you need to shift your thinking - it's apples and oranges. You need to separate hourly heat loss (which happens 24/7) from solar gain, which is measured in average daily vertical surface insolation per square foot (which seems to be about 1800 btu in March in Fairbanks and way more than I get in VT).
If you want some basic data on solar gain through windows by climate and month of year and window orientation, you should check out NRELs solar radiation database -- see http://rredc.nrel.gov/solar/old_data/nsrdb/
Michael: Thanks. I'll have a look-see. Robert: Well, one of the reasons I posted my numbers was because I am self-taught in all this, and therefor not 100% sure. (-51 is the design temp, though. I have lived through -60, North Pole has had -70, and I believe it was Fort Yukon that got to -81; back in the early 1970's.) Way back when, I studied an NAHB Insulation Manual, and in there they demo'd heat loss in a house in Baltimore, Maryland, and used a delta T of 60 to calc the hourly heat loss. After all the house parts' losses were summed, they then divided by the dT (eliminating it) and multiplied by 24 and the HDD. That gave the annual heat loss. I asked someone who was in this line of work if I had interpreted correctly, and he said "yes". So, I proceeded. My calcs are pretty well (*) in line w/ what we actually use, too, which also lead me to believe I had things right. So, that is all I know; right or wrong. If this is incorrect, please send me to a place where I can study this all again and get it right. I don't want this forum to turn into a piece-by-piece course on heat loss. Thanks. klingel
Why not? It will undoubtedly be useful to others.
That may have been the maximum hourly heat loss during that coldest part of the one night of the year that reached the design minimum temperature. And that hourly heat loss is important only for sizing the heating plant to match the maximum design heat loss of the building.
But, for calculating annual heat load and heating cost, you have to use the average winter temperature, and the average winter Delta-T. The average Sept-Apr temperature for Fairbanks is 12°F.
John, it would be interesting to see a sun plot for your location.
The University of Oregon has a useful online program that can help.
I wonder if winter "daylighting" even makes sense for someone so far north?
So far north and with such extreme cold, would the "best performing" house have no windows at all?
Not that I'm advocating that you have no windows... Just wondering.
Robert: I will run this by someone again, w/ paper and pencil in front of us. Something "ain't quite raght cheer", and that all may be sitting in Frbs at the moment. Martin: Regarding "...Oh, yes -- and one more factor: the amount of reflected sunlight, if any, that it receives. (Even if the window is shaded, reflected light can provide heat gain -- especially in snowy climates.)..." Agreed. (I am re-reading a few of my posts; I should not eat and write/read simultaneously.) I believe you mentioned in "Windows as good as walls" that even N facing windows can gain quite a bit of heat from reflected light; that was news, and good news, to me. What I was referring to when I said "no sun" was not even reflected sun; just blank, dark-as-night dark. And, I realize that the reported R5 value MAY change depending on the temp; dunno that one. I have heard that fiberglass batts do have a reduced R value, from the reported R, when it gets cold; great, right when you need them. That I don't know for a fact, either; just read it, and I thought on a reputable site. klingel
Lucas: That is an interesting question. I chose to have very small windows back in '80, but that was not a good move for Wife 2. Mental health needs to be considered, too, so the new house will have bigger, and more, windows. I have not seen a house here w/out windows, but I think a few foam geo-domes down the road have very little. If I can find out any sun data soon, I'll post it. Maybe I can learn a bit about that along the way. OK, Robert: Whilst mudding, I had this thought. If the annual heat loss is Q = (sum of all UA) x 24 x HDD, no design temp is needed (it is buried in the HDD). I think the HDD is what differentiates heat loss in Frb from the same (sum of all UA) in Ohio, or wherever, that has a lower HDD. That is how the book I referred to, my guru buddy, and everywhere I read online have calc'd it. And all these references say "annual house heat loss", not boiler size. I will ask Thorsten next Tues, too. I'm sure he's been down this road 100 times. j
I just found this site; no idea if it is accurate or not, but fwiw.... http://www.susdesign.com/windowheatgain
Yes, for annual heat loss. But you said "Fact is, my spread sheet says my 184 sf of windows will lose 4,269 btu/hr, using a design temp of -51, or a delta T of 116 F."
The only reason for using minimum design temperature is for calculating maximum design heat loss for sizing the heating plant.
To determine annual fuel consumption, you use HDD for the season to determine heat loss, subtract from that internal heat gains from human occupancy, and then subtract solar heat gain (which is usually in average BTU/SF-day for each month and has to be reduced by shading coefficient and SHGC).
The Susdesign window heat gain calculator is only as accurate as your inputs. The biggest variable is average monthly cloud cover. It's easier just to look up annual climate data from the nearest weather station, which often includes monthly solar availability.