GBA Logo horizontal Facebook LinkedIn Email Pinterest Twitter X Instagram YouTube Icon Navigation Search Icon Main Search Icon Video Play Icon Plus Icon Minus Icon Picture icon Hamburger Icon Close Icon Sorted
Musings of an Energy Nerd

Choosing Triple-Glazed Windows

Balancing U-factor and solar heat gain

Extra glass keeps out the cold. Triple-glazed Canadian windows like this casement window from Thermotech have U-factors as low as 0.17.
Image Credit: Thermotech

Since 1977, when Sweden introduced its stringent energy code, almost all new homes in Sweden have been equipped with triple-glazed windows. Here in the U.S., where energy codes are more lax, triple-glazed windows are still rare.

For a minority of U.S. builders, however — especially cold-climate builders of superinsulated homes — triple-glazed windows are considered essential. Since few U.S. manufacturers offer high-solar-gain triple-glazed windows, most Americans get these windows from Canadian manufacturers.

Look for a low U-factor and a high SHGC

In any climate, a window with a low U-factor performs better than one with a high U-factor. The lower the U-factor, the better. (For more information on low-U-factor windows, see “Passivhaus Windows.”)

Most cold-climate builders want windows with a high solar heat gain coefficient (SHGC) and a high visible light transmittance (VT). After all, solar gain helps heat a house during the winter. During the summer, when solar heat gain is less desirable, a properly sized roof overhang will shade south-facing windows during the hottest hours of the day. (For more information on the desirability of windows with a high SHGC, see “High-Solar-Gain Glazing” and “Windows That Perform Better Than Walls.”)

Most low-U-factor windows have a SHGC that is unacceptably low — at least for cold climates. Designers of cold-climate houses have to balance conflicting needs — the need for windows with a very low U-factor and as high a SHGC as possible. Most builders end up choosing triple-glazed windows with a U-factor ranging from 0.19 to 0.26 and a SHGC ranging from 0.39 to 0.47.

The higher the visible light transmittance (VT), the better. Windows with a low VT look gray and depressing. To get an idea of what range you’re looking for, consider the advice of Robert Clarke, a technical specialist at Serious Energy and the former president of Alpen Windows. According to…

GBA Prime

This article is only available to GBA Prime Members

Sign up for a free trial and get instant access to this article as well as GBA’s complete library of premium articles and construction details.

Start Free Trial


  1. Anonymous | | #1

    Still Can't get to PH requirements for Northern Climates
    I am building a PH in VT. These are close, but I was able to find imported windows which met PHPP model requirements.

    IV88 Wood Window Frame (88mm) … U-value = 0.21
    Triple Pane Glass CLIMAPLUS ULTRA N …U-value = 0.088 (SHGC 0.5 or 0.6 available)

  2. GBA Editor
    Martin Holladay | | #2

    About Climaplus
    Climaplus is a type of glazing from SGG (Saint-Gobain Glass). The SHGC number you quote appears to be a glazing-only number, not a whole-window number, so it is not directly comparable to the SHGC numbers in my table.

    What brand of windows are you considering purchasing?

  3. Dan | | #3

    SHGC & summer
    I'm an architect & I've always been wondering about how to balance SHGC, I've done projects in MA, NH & VT, but my question is this: Most people that are considering triple pane windows in these climates won't have AC in the summer. Won't a SHGC increase from .24 to .47 produce a large increase in the inside temperature on the hottest days of the summer, regardless of whether the windows are properly shaded, or not? After all, even properly shaded windows will get direct sun some of the day.

  4. GBA Editor
    Martin Holladay | | #4

    Summer overheating
    The answer to your question can be found with good modeling software -- many designers are gravitating to the Passive House Planning Package, or PHPP -- mixed with experience and good judgment.

    There is no single answer to this question. Most designers of superinsulated cold-climate homes specify orientation-specific glazing. While it's clearly advantageous to specify high-solar-gain windows on the south elevation, most designers feel safer specifying low-solar-gain glazing for the west elevation.

    Actual indoor temperatures will depend on the orientation of the windows, the glazing ratio, and whether the windows are exposed to full sun or are partially shaded by foliage. I have lived in a house in northern Vermont without air conditioning for 30 years. Until 2 years ago, all of my windows were glazed with clear double glazing with a very high SHGC. The house has never been uncomfortably hot.

    Here's the problem with specifying low-solar-gain glazing: you're leaving money on the table and condemning the homeowners to years of utility bills that will be higher than necessary. Low-solar-gain glazing results in higher utility bills; that's a fact.

    Here's another fact to consider: if you choose triple-glazed windows, your SHGC numbers won't be that high even if you strive for high solar gain. We're talking 0.39 or 0.44 here -- not 0.60. Just be careful and don't oversize your windows, especially on the west, and think about your roof overhangs.

  5. Kyle | | #5

    Remember the difference of Europe and North Am. U factor calc.

    It has been mentioned here before that the way that a whole window U factor is calculated in North America leads to approximately a 10% higher calculated value than if it were calculated using the European standard method. So keep that in mind when you consider spending considerably more money to purchase European windows.

    A few of the windows that Martin suggested have been used in PH certified homes in North America before and have very good reputations.

  6. Graham Mink | | #6

    Do you know the approximate cost for any of these windows? Even a relative cost scale would help a ton. Thanks for all of your extremely helpful blogs!

  7. GBA Editor
    Martin Holladay | | #7

    Cost info
    Two years ago I bought Thermotech triple-glazed windows for my own house and paid $56 a square foot.

    I recently got a bid on four Inline triple-glazed windows and was quoted $65 a square foot.

    Another bid on the same sized windows from Paradigm came in at $51 a square foot with krypton gas and $40 a square foot with argon gas.

    Jesse Thompson writes that he budgets $45 to $50 a square foot for Inline windows, but it may be hard to get such low prices. He says he budgets $65 to $70 a square foot for Serious windows.

    A while back, Sam Bargetz of LoadingDock5 Architecture in Brooklyn shared the bids he received on 12 casement windows. Serious Windows came in at $46 a square foot; Fibertec at $49; Thermotech at $66; and Bieber (a European manufacturer) at $196 a square foot. (Bargetz reports that Serious has since raised their prices.)

    Stephan Tanner says that "Three-Wood" Optiwin windows from Germany cost about $110 a square foot in the U.S.

  8. adkjac upstateny | | #8

    measuring a window's area
    For heat loss and window purchasing.... I am assuming that to measure a window at my own home, I should measure an inch past the inside of the finish material that my casing is nailed to.

    And if looking at windows in a catalog, the actual window exterior measurements, or maybe the RO?

    Martin, tell us how to do this right.

  9. GBA Editor
    Martin Holladay | | #9

    Measuring for a replacement window
    It sounds like you are asking how to measure for a replacement window. Check out:

  10. Andrew Henry | | #10

    Energy Rating
    Hey Martin,

    It would be great if that table had the Energy Rating for each window. Though I suppose that not all windows would have an ER number.

    The ER value would help people understand the relationship between SHGC and U-value.


  11. Doug McEvers | | #11

    Energy Rating

    The same window will perform differently in various locations depending on the amount of available solar radiation. Heating degree days also come into play. A quality heat loss program is needed to evaluate window types for a given location.

  12. GBA Editor
    Martin Holladay | | #12

    ER numbers
    Good idea. The ER ratings for many of the windows are available. I'll add them to the table when I get a chance.

  13. Stephen Thwaites | | #13

    Calculating ER
    For those that just can't wait for Martin to add ER numbers, this is how the current ER is calculated;

    ER = (Gains - Losses) +40
    Gains are from the sun
    The gains are an average of 13 or so Canadian cities (can't recall the exact number) and N,S,E &W
    They are also averaged over a 5064 hour heating season
    All this averaging produces an average heating season solar gain of 72.2 W/m^2
    A, in my view debatable, solar reduction factor of 0.8 is applied to the gains
    Losses are from Uwindow and Air Leakage
    For a good hinged window the air leakage losses are low
    for example quadrupling our casement's air leakage decreases the ER by about 0.6
    ( the ER's on Martin's list will range from about 25 to about 40)
    The avg. temperature difference for the same 13 or so cities over the 211 days is 21.9C (39F)
    note, since the temperature difference is in Celcius, then Uwindow needs to have metric units
    (Uwindow,imperial * 5.678 = Uwindow,metric)
    This 'bump-up' or 'fudge factor' was added a few years ago to the formula so that no purveyor would be embarrassed by having a negative ER

    So ignoring the air leakage to do a quick comparison ahead of Martin;

    ER = (72.2 * SHGCwindow * 0.8) - (21.9 * Uwindow,metric) +40

    the units were W/m^2 - an average heat flow over the heating season, until the unitless 40 was added......

    Of course, even better than the ER is an actual simulation. With good software you can 'dial-in' your climate and shading to get a more accurate answer. The Passivhaus's PHPP and HOT2000 (in detaiiled window mode) are particularly good because they will calculate window properties for the sizes of windows in your building -- as opposed to assuming all windows have the same properties as the NFRC sized windows. If a detailed simulation is not in the cards at least use RESFEN.

    For extra points:
    Calculate the BFRC's WER which is calculated in a similar manner to the ER, but based on the cloudier and milder UK climate
    (note the factors applied to gains and losses are different from the ER because the Brits decided to use kWh/m^2/yr as their units)

    Stephen Thwaites P.Eng.
    Thermotech Fiberglass Fenestration

  14. Andrew Henry | | #14

    Fudging to prevent humiliation
    Hi Stephen and Martin,

    Correct me if I'm wrong. Before the fudge constant of 40 was added to the ER equation the boundary at which a window was a net gain towards heating the house was zero?

    So to know if a window has crossed this threshold we'd just have to subtract 40 from the present ER value, if the number was positive then the window is a net gain.

    To make things less complicated. If the stated ER is greater than 40 then the window is a net gain to the warmth of the house? : )

    I know that ideally you'd model the whole house, but the beauty of the ER number for the consumer/homeowner was that it pretty starkly pointed out whether a window was a net benefit to the energy efficiency of their house. Every other value (SHGC, U-Value) makes most peoples eyes glaze over. Pity NRCan added the fudge factor!? A threshold of zero had a nice metric way of focusing one's attention. Though it's a reasonable concession I suppose if it means window manufacturers (in Canada) state their windows' ER number.



  15. dankolbert | | #15

    Being from Maine I'd love to use Paradigms but I'm not convinced their frames are that great.

  16. GBA Editor
    Martin Holladay | | #16

    Vinyl frames
    You're right that vinyl frames are less desirable than pultruded fiberglass frames. Vinyl expands and contracts a lot with changes in temperature, and isn't as strong as fiberglass.

    There's another problem with Paradigm's windows: their triple glazing is only 7/8 inch thick. Their glazing gaps are sub-optimal -- chosen because of the limitations of their window frames rather than for thermal performance. If you want good thermal performance, especially with argon gas, you want triple glazing that is 1 3/8 inch thick. That's available from the Canadian manufacturers of fiberglass windows, but not from Paradigm.

    Filling Paradigm's skinny glazing gaps with krypton helps somewhat, but the price of krypton has gone through the roof.

  17. GBA Editor
    Martin Holladay | | #17

    ER numbers
    I have posted a new table (see above) with the ER numbers, using the formula provided by Stephen Thwaites.

  18. Stephen Thwaites | | #18

    ER numbers

    As Andrew suggests, an ER over 40 means that for an average orientation the window is a 'net gainer'

    Keep in mind Martin's list is for casements, fixed windows are about 5 'points' higher due to their smaller frames and larger stnd size for caculating thermal properites (more glass is better). So it's easier for a fixed window to be a 'net gainer'

    Interestingly for people who like numbers, there is something called an ERS. The 'S" is for specific.
    'Specific' means for a given location and orientation
    It is better for picking windows by orientation than the ER

    For a window in Ottawa for a window w/ an ER of 40, the ERS's would be as follows:
    south................ 58
    east& west ..... 37
    north................. 26

    ERS was developed before simple tools like RESFEN became available so ERS is almost never used these days. A nice feature of RESFEN is that it lets you customize shading. This is important because if a south facing window has no shading you'd pick-up another 10 ER 'points'. This is because, as mentioned in an earlier post, the ER assumes 20% shading.

    Hope that helps
    Stephen Thwaites P.Eng.

  19. Jim W | | #19

    ER Numbers, where did they come from?
    Trying to make sense of the ER numbers and this is what I found. Accurate Dorwin's top casement window actually reached a 40 not a 37 as posted above. But the window with a SHGC of 52(highest)had an ER of 32 not 37 as above. Fibertec's top number was 35 not 33.8 as mentioned above. Same case? Either I don't understand or the numbers are incorrect or the website is wrong.

    Can someone clarify and I think I will stick to the U

  20. adkjac upstateny | | #20

    No... measurement used for heat loss calculation
    Anybody... Not asking about window size for window size purchase. I am asking about window size to figure window area... to be used in heat load software.

    Since whole window u includes frame, I am assuming a window's RO is what most people use that do heat load calcs with software?

  21. GBA Editor
    Martin Holladay | | #21

    Where the numbers came from
    You asked where the numbers came from. As I noted in my earlier post, these ER numbers came from Stephen Thwaites' formula:
    ER = (72.2 * SHGCwindow * 0.8) - (21.9 * Uwindow,metric) +40

    Now, if Thwaites' formula is only an approximation, or is incorrect, I will be happy to correct the numbers resulting from that formula. Any technical experts out there want to either confirm or contradict the accuracy of Thwaites' formula?

    The Energy Star Canada Web site is useful, but only up to a point. The Web site doesn't accurately specify the glazing used to calculate its numbers -- it just uses vague terms like "hard coat argon." In my table I have specified the actual glazing product used by the window manufacturers.

    Many window manufacturers have had their windows rated with 30 or 40 glazing types -- but they really only promote or sell two or three of these. I have tried to list the triple glazing options that the manufacturers actually promote or offer.

  22. Stephen Thwaites | | #22

    Where the numbers came from
    What is correct Martin's numbers (based on the formula i supplied) or NRCan's?

    The answer is, of course;
    .....both are correct

    Take Jim W's Dorwin example;
    The NRCan website shows Dorwin can get a +40 for a triple glazed casement
    - the details for this window are shown here:
    - this is not the same window as is on Martin's list -- it has 1 more low-e coating then the window on Martin's list
    - according to the NRCan figures;
    --- the +40 window has a Uwindow,metric of 1.14 (or 0.20 in imperial units)
    --- the +40 window has a SHGCwindow of 0.44

    Put these values into 'my equation' (the one that ignores air leakage) and presto: get an ER of .......+40 !

    So both are correct

    I did leave one incorrect impression about oldER and newER and to correct myself;
    - no shading
    - zero meant zero
    - shading factor of 0.8 (20% shading)
    - fudge factor of +40 added

  23. GBA Editor
    Martin Holladay | | #23

    Thanks for another helpful post. I appreciate all the information you share.

  24. Steve | | #24

    Estimating window costs
    As a window distributor I'd strongly suggest using caution when estimating window costs by the square foot. Most fiberglass & vinyl and now some wood manufacturers use united inch (width plus height) formulas. In addition to the variety of options that can skew a SF price up or down (fixed versus venting etc) different sized windows may fall into the same united inch bracket and therefore could significantly vary in SF cost.

  25. Tim Eian | | #25

    Optiwin Alu2Wood
    We are using Optiwin's much more affordable Alu2Wood (vs. the aforementioned 3Wood) window for a Passive House project in Wisconsin. The frames are insulated, and once properly installed, essentially buried in the wall's insulation package, making for a performance boost of this often poorly performing connection. Glas Trösch in Germany is supplying triple-pane glazing with a metric U-value of 0.6, and a g-value of 63.5%. We were able to boost the heating performance of the entire 2,000SF building by over 20% when using this combination over a previously specified Uw 0.53 and g 52% glazing in the same frame.

    The windows priced out at around $75/SF—including shipping and customs, as well as the current $-EUR exchange rate.

    We are using an exterior motorized venetian-blind shading system on a sun and wind sensor to manage heat-loads throughout the year.

    GBA is following the project at

    We expect the windows to arrive in early May and will document installation on our website.

    On a related note, I would find it very helpful if Optiwin product were featured in the comparison list in this post, as they are well represented throughout the U.S. and readily available in the market place. In our research, we found them to be the best value for money, and with their excellent support, and plethora of glazing options, we were able to custom-tailor the performance of the windows to our application. In addition, our local supplier (Peak Building Products) provided a matching shading solution and helped facilitate installation details—something that can make or break any window selection in the built product.

    Related links:
    TE Studio:
    Peak Building Product:
    Optiwin USA:
    Passive House in the Woods project:

  26. doug | | #26

    Cost delta for triple glazed
    I have had very recent pricing on two projects with triple glazing. One was a Kolbe and Kolbe spec, aluminum clad wood casements, and on a $30,000 window order, the upcharge from double glass was about $4100. The other project, specifying Pella was about 15% higher with triple glass, double low e. Weathershield is another domestic manufacturer that has offered triple glazing for years.

  27. Harry Applin | | #27

    R-5 windows
    Cost? Check this out

    Highly-Insulating (R-5) Windows and Low-e Storm Windows Volume Purchase Program
    The U.S. Department of Energy's Building Technologies Program (BTP) is coordinating a volume purchase of R-5 windows, and low-e storm windows, to expand the market of these high efficiency products.
    Price is the principal barrier to more widespread market commercialization. The aim of this volume purchase initiative is to work with industry and potential buyers to make highly insulated windows more affordable.
    A considerable effort has been made by the Volume Purchase Team to educate a variety of groups on the benefits of highly-efficient R-5 windows; groups within academia, local government, non-profits, weatherization and others. Whether through meetings, webinars or conference calls, various potential buyers, marketers and end-users have learned more about this program, and are excited about the opportunity within it. Below are examples of different audiences the program has been working with, as well as letters from groups such as the Consortium for Energy Efficiency (CEE) (PDF 163 KB) expressing interest in participating in this R-5 Windows Volume Purchase Program, if proper volumes and price points are met.
    Both home owners and buyers can take advantage of the energy savings from windows available in the volume purchase program. R-5 windows can reduce the average heat loss through windows by 40% over common ENERGY STAR® windows with an R-3 value, and low-e storm windows have the potential to double the energy savings over storm windows without a low-e coating.
    Windows in the U.S. account for 30% loss of building heating and cooling energy, representing an annual impact of 4.1 quadrillion Btu (quads) of primary energy. Windows have an even larger impact on peak energy demand and on occupant comfort.

  28. GBA Editor
    Martin Holladay | | #28

    Triple glazing from U.S. manufacturers
    While it is possible to get high-solar-gain triple glazing from some U.S. manufacturers, it's usually quite difficult. If you're tempted to buy triple glazing from a U.S. manufacturer, ask these questions:

    1. Is the triple glazing full thickness (1 3/8 inch) or less than optimal thickness (only 7/8 inch or 1 inch)? Thin triple glazing has poor thermal performance, unless expensive krypton gas is specified.

    2. What is the whole-window SHGC? Most U.S. manufacturers — if they offer triple glazing at all — offer only low-SHGC glazing that is more appropriate for hot climates than cold climates. Most cold-climate builders want high-solar-gain glazing.

    3. Compare the whole-window U-factors and whole-window SHGC to the numbers in the chart. It's hard to beat the Canadian windows with fiberglass frames.

  29. kim_shanahan | | #29

    Serious Windows
    Martin, you mentioned the Serious Windows does not offer triple glazing, but I am certain they do. Including heat mirror films in both cavities. They say they have a line of windows with up to R-ll values that I believe are made with pultruded fiberglass. Is all this true? Or manufacturing hype?

  30. GBA Editor
    Martin Holladay | | #30

    Is Heat Mirror triple glazing?
    You're right (as you can see from the table included in my article) that Serious Energy offers windows with a very low U-factor (U-0.13). You're also right that the frames are made from pultruded fiberglass.

    Technically, though, the Heat Mirror glazing used by Serious is not really triple glazing. As I explained in the article, Heat Mirror uses just two panes of glass. Suspended between these panes of glass is one or more plastic films; these plastic films create distinct air spaces that mimic the performance of triple or quadruple glazing with much less weight.

    There are a few downsides to Heat Mirror glazing, however. In general the plastic films provide lower SHGC and VT numbers than comparable triple glazing. Moreover, some builders are still skeptical of the longevity of the plastic films used to make Heat Mirror glazing.

  31. Anonymous | | #31

    I did that mistake for my house...
    I live in Canada, Montreal area. I installed triple glazed windows all over. Yes, I noticed a significant decrease in my electricity bills. I also noticed a significant decrease in the light quality inside. I won't make this mistake anymore.

    Take a look at this great research paper from the CNRC on the topic:

    Basically, its more economical to favor heat gains in the winter and pay the extra of cooling in the summer, at least, where I live.

  32. GBA Editor
    Martin Holladay | | #32

    About your mistake
    If I understand you correctly, purchasing triple-glazed windows was not a mistake. Your mistake was purchasing the wrong kind of triple-glazed windows — that is, windows with a low SHGC and a low VT. As long as you aim for a high SHGC and a high VT, you should be fine.

    Do you know what type of glazing you purchased?

  33. Tim Eian | | #33

    Regarding light transmission, and other window musings
    In regards to the light transmission discussion, I can add that the aforementioned Glas Trösch glazing from Germany offers 74.8%. In addition, and simply converted (I know the Germans test a bit different) the 0.6 W/(m²K) U-value translates to an imperial 0.11 Btu/(h ft² F), which helps us to achieve 0.14 to 0.15 Btu/(h ft² F) for the installed windows in our project, depending on size and orientation.

    We are using the Passive House Planning Package 2007 to model our projects. It accepts U-values for glazing and frames, as well as g-values, and thermal-bridge factors for the installation detail.

    While important, I'd like to point out that the simple specs on a window are only part of the equation, as installation into the wall can be crucial for overall performance. The PHPP can model different installation conditions and the associated thermal bridging, which can have a significant impact on the overall performance of a building—at least for something as highly efficient as a Passive House.

    The PHPP also looks at shading and orientation very carefully. Both play a major role. I think it's important to look at an installed window as a subsystem of the house, which itself is a system. Therefore, individual specs, while important, only make sense when looked at in regards to the entire system.

    Window technology is also in evolution. A few years ago, when my colleague Stephan Tanner built the BioHaus in Bemidji, MN (the first certified PH in the US), metric U-values were limited to about 0.7 or 0.6, with g-values around 50%. Within the last 5 years, those numbers have improved to Ug 0.5 and g 64%. At this point, we do not really have to chose between good insulating value and high-solar heat gain, as well as good light transmission. In addition, many of these glazing units do not have to be Krypton-filled, which arguably is not a realistic option for all glazings produced due to the scarceness of this resource.

    I'd like to add that I am not trying to say that North American windows in general are worse than the German windows. For our specific application however, they offered the most technically appropriate solution at an astonishingly affordable price.

    I personally feel that we are lacking decent frames in NA; the glazing can most likely be made available very quickly, as most of the big players in glazing are global anyways. I had a conversation with a marketing director from Cardinal Glass about this in 2008. They appeared to be happy to supply the market with European triple-pane glazing types, but most U.S. American frames can simply not hold them, nor are they thermally adequate in design. Insulated fiberglass frames from Canada may perform better in this department. However, a nailing flange installation in the weather plane with a significant thermal-bridge factor around the edge is still an issue here.

    Thanks for kicking of this very interesting post. I appreciate all your great work at GBA.

  34. GBA Editor
    Martin Holladay | | #34

    On installation and other matters
    Thanks for all of the information you provided. Readers interested in learning more about Optiwin windows and the differences between North American and European protocols for window testing and rating can see my previous blog, "Passivhaus Windows."

    I certainly agree with you, Tim, that we need a "whole building" approach to consideration of these issues -- one that includes sensitivity to orientation, sizing, and window area. The PHPP software appears to do an excellent job of helping designers sort these issues out.

    I'm thinking of an analogy, though. I once read a financial planner talking about retirement savings plans. He pointed out that a person can get obsessed between whether or not he was getting a 3% or a 4% return on retirement investments -- and then could go through a divorce which wipes out 50% of the investments. (Well, it doesn't wipe it out, but it redistributes it.) Similarly, Passivhaus builders are sometimes obsessed with installation details — "overinsulating" the exterior of the window frames, for example — that affect relatively tiny energy flows.

    House designs vary widely. I'm still seeing some very large houses out there -- including some large houses that are aiming at Passivhaus certification. A designer can control the size of the house, the area and orientation of the windows -- and these factors ultimately matter more than whether or not the exterior foam sheathing covers the frames of some of the windows.

    So here's my bottom line: design a small, very tight house. Include very thick insulation. Orient your windows intelligently, shade them in the summer if possible. Don't overglaze or underglaze your facades. If you do these things, you'll be very happy with triple-glazed Canadian windows. And you will save a little money compared to the builders who import windows from Austria or Germany.

  35. Tim Eian | | #35

    RE: On installation and other matters
    Thanks for the insight, Martin.

    Maybe one comment regarding Passive House. We find that a lot of these details, while more or less negligible in traditional construction, can have significant impact once one designs a high-performance building like a Passive House. We've seen the entire annual heating energy performance go 10-25% one way or another based on frame conditions, and/or glazing selection. This is of course in climate zone 6, which means that small short comings are amplified by the number of the heating degree days and severity of peak low temperatures.

    One comparison that I sometimes make is this: If you drive a gas-guzzler that gets 4 MPG, and your tires are not properly inflated, you will likely never know that you are missing out on a fraction of an MPG in performance. However, if you build an extremely efficient car that gets 180 MPG, and your tires are not inflated properly, you may experience a very noticeable drop of 20 or more MPGs.

    I completely agree with all the other points you make. Small, compact, properly oriented, air-tight and well insulated will get anybody to a much higher level of performance. It seems that there may be some absolutes in life after all.

  36. GBA Editor
    Martin Holladay | | #36

    The map is not the territory
    When you're fiddling with house specifications at the design phase — entering in different window frame U-factors in hopes of achieving the magic 15 kWh per square meter on the PHPP spreadsheet — there's no doubt that small tweaks can make the difference when it comes to hitting your target.

    But if you take a step away from your drafting table or computer keyboard, and consider the number of BTUs involved, the numbers are small. Many people are quite happy to live in houses that require 18 or 19 kWh per square meter. That looks like a lot more energy than 15 kWh per square meter -- and it may be, as a percentage of your Passivhaus goal — but we're not talking about big numbers.

    What do we want? Well built, well designed, beautiful houses with low energy budgets. Most people want the house, not the Passivhaus certificate. The map is not the territory.

    Here's something else to consider: a smaller house using more kWh per square meter may have lower energy bills than a bigger house with fewer kWh per square meter.

  37. Anonymous | | #37

    high elevation passive house
    Does anyone have ideas/comments for good windows for service in alt. 5500ft? I heard thick tiple panes are only good to 2500 ft in elevation. I'm also in need of narower units with thick glazing. Thick glazings for North America (-40 exterior) seem to be hard to get in narrow units. Also most of the Rockies are in the ES North zone.

  38. Steve McCarthy | | #38

    High solar heat gain.
    One option for getting a high solar heat gain coefficient from a typical window manufacturer in the USA is to order a "replacement sash" from a vinyl replacement window manufacturer and install it backwards. I know this only works for non operable windows and yes plastic sucks but after the sash edge gets partially buried by interior and exterior stop and the low e film is in the correct position , the window performs and the glazing cost is cheap!

  39. lonestar | | #39

    First Folding Glass Wall System with Triple Glazing
    NanaWall Launches First Folding Glass Wall System with Triple Glazing

    Glass walls that provide exceptional thermal performance and weather resistance
    NanaWall Systems, the industry leader in opening glass walls, announced its newest product for architects and builders in December 2009: NanaWall® SL70 with Triple Glazing, providing exceptional thermal performance for the harshest weather conditions.

    The NanaWall triple glazed SL70 system provides flexible solutions to architectural openings from three to thirty six feet wide using one to twelve panels.

    The SL70 system was chosen for the glass pavilion at the Bavarian Zugspitzbahn Bergbahn, a ski resort in the Alps over 8,000 feet above sea level. The pavilion includes forty six panels of the SL70 including three folding glass door systems of six panels each, and seven folding glass window units with four panels each.

    The architect chose the SL70 System because it met the client’s challenging specifications and testing requirements to provide:
    • Comfort despite temperature variations from -40° F (-40° C) to 60° F (+15° C)
    • Large opening walls that can withstand wind speeds of up to 125 miles per hour with the associated pressure and suction.

    The NanaWall SL70 with Triple Glazing system has an NFRC certified U-value of .30 and a Solar Heat Gain Coefficient of .24. It meets Energy Star requirements for all areas of the US and areas A and B of Canada. It also provides high weather resistance and structural performance. For in swing and out swing models with a raised sill, independent testing indicates that there was no water leakage at 12 psf or 68 miles per hour.

    The NanaWall SL70 with Triple Glazing is available in more than 200 finishes as a floor-mounted system with more than 50 stacking configurations (for pics and details click

    “NanaWall Systems is the first folding door company to offer triple glazing,” said Ebrahim Nana, president of NanaWall Systems. “Along with our partner Solarlux, Germany, NanaWall Systems is committed to developing opening glass wall systems for large openings for our customers in the harshest climate conditions – whether a hurricane zone or the top of a glacier.”

  40. Dirk | | #40

    Gas Filled
    I have been told that the gases (Argon/Krypton) in windows will leak out over time and therefore lose the value purchased in the first place? Is this so, and if so, wouldn't that have to be factored into all the calculations?

  41. GBA Editor
    Martin Holladay | | #41

    An answer to Dirk's question
    According by "Understanding Energy-Efficient Windows," a Fine Homebuilding article by Paul Fisette, "Argon and krypton are safe, inert gases, and they will leak from the window over time. Studies suggest a 10% loss over the course of 20 years, but that will reduce the U-factor of the unit by only a few percent."

  42. Tim Eian, TE Studio | | #42

    Resource Management & 2000-Watt Society

    Many people are quite happy to live in houses that require 18 or 19 kWh per square meter. That looks like a lot more energy than 15 kWh per square meter

    Your points are well taken and we need to look at the overall energy consumption of a building. I think that in the end, we need to figure out how many resources all of us can use, sharing a single Earth with finite resources. Others, who are much smarter than me, have already done the math on this, and the answer is 2000 Watts of continuous energy per person:

    IMHO and having said that, there should be absolute goals for buildings, as well as other areas of life and production to achieve this level of sustainability. Whether that is Passive House or something else is less important than the fact, that we may need to look for that absolute. Considering this paradigm, the jump from 15 to 19 kWh per square meter and year may make all the difference in the world.

  43. GBA Editor
    Martin Holladay | | #43

    The 2000 Watt Society
    Thanks for mentioning the 2000 Watt Society. I wrote about the 2000-watt goal in a previous blog, "What Does ‘Sustainable’ Mean?"

    I met several people involved with the 2000 Watt Society when I traveled to Switzerland in 2008. It's a good group and I admire their goals — even though a cogent argument can be developed showing that the planet can't afford such a high per-capita energy budget. We probably need to go below 2000 watts per capita.

  44. Anonymous | | #44

    maximum size per pane of triple glazed glass
    wondering maximum size for triple glazed glass? building a home in whistler and want large picture windows with no metal frames in between!

  45. GBA Editor
    Martin Holladay | | #45

    Contact your supplier
    Contact your supplier.

    Here are maximum dimensions for triple-glazed Thermotech fixed windows:
    Maximum width: 120 inches
    Maximum height: 120 inches.

    So, is 10 feet by 10 feet big enough?
    More info here:

  46. Chris Hall | | #46

    Triple Glazed Wood Windows
    Anyone in North America produce a triple glaze wood window? I see them in Europe but nothing on this side of the pond.

  47. GBA Editor
    Martin Holladay | | #47

    Response to Chris
    Yes, several U.S. manufacturers, including Marvin, offer triple glazing as an option for wood windows. However, these windows have thin (7/8 inch or 1 inch) thick glazing instead of full thickness (1 3/8 inch) glazing, so the performance is low unless expensive krypton is specified.

  48. Doug | | #48

    High Elevation windows
    Five months late, but to answer Anonymous who wants to use sealed glazing at 5000 feet elevation, there are companies who make units that will work.
    Serious in particular includes an expansion bulb on the units we get on the east coast, which allows the air pressure to equalize between the sealed unit and the local atmosphere. Once the unit is in place, the installer pinches off the metal tube and removes the bulb.
    I think they do this because their glazing is made in the West and has to cross the Rockies to get to us here on the right coast, and Serious wants their gas fill to stay in the unit even as atmospheric pressure radically changes during the mountain crossing. Whatever the reason, their system will allow their glass to be used at higher elevations than sealed units without the bulb.
    Other companies whose units cross the Rockies often include an open tube from inside the unit to the outside air, and they won't sell you argon or krypton fill for these units. Windows that don't experience major elevation changes (from shipping or final location) don't need either a tube or an expansion bulb.

  49. JAGA Enterprises | | #49

    Northeast Windows
    I have been using windows manufactured in Syracuse, NY by Syron Industries. They have the standard new/replacements that compete pricewise with all windows (other than those provided by the "big boxes"). However, their new Eco-Vue window line is the best I've seen for those looking for energy efficientcy. Uncomparable prices due to their ability to not having to use Krypton to meet u-factors of .17. I buy them thru a retail outlet in Syracuse, since Syron is only the manufacturer.

  50. Bernard Rokicki | | #50

    Fair Window Comparison

    I have researched your comparison chart above with the NFRC.ORG website. I have compared the Thermotech Window with Inline Fiberglass with triple LOF, foamed filled and Super Spacer. Both windows have the same .19U, .42 SHGC and .43VT. Please see link below.

    Therefore would you not agree that we have two identical performing windows?

  51. GBA Editor
    Martin Holladay | | #51

    Response to Bernard
    Assuming their performance specifications are identical, then you should choose the window you prefer based on hardware, appearance, or the reputation of your dealer for service.

  52. | | #52

    window cost savings
    A local solar building designer (Don Roscoe) makes his own fixed windows out of cellular PVC (to make the frame) and IGU's from a local glazing manufacturer (usually PPG).
    The cost for a 20sf IGU with triple glazing, 2 lowE, argon fill, & superspacer is around $300ea in quantities of 5 or more.

  53. | | #53

    1 15/16" thickness better than 1 3/8"
    I modeled a few window configurations in LBNL Window 5.2.
    I used 3mm clear glass (ID# 9801) for the outer lite, and LOF-EA2 (ID#9921), and 90% argon fill. A good manufacturer who uses a Sparklite tester will have an average of 90 to 95% argon fill.
    With an overall thickness of 1 3/8", that leaves room for 1/2" between the glazing layers. The result in Window 5.2 is a COG U-value of .159 Btu/h-ft^2-F.
    Increasing the inner gap to .563" improves the U-value to .157.
    With the inner and outer gap at .563", the U-value is still .157, as it is with .563" and .625".
    With both gaps at .625", the U-value is .158.
    I think the best configuration is .688 and .787", which gives a COG U-value of .161.
    With CEN environmental conditions, which are more in line with the typcial heating season of Toronto than NFRC-2001 conditions, the U-value is .144. With CEN and 2 gaps of 1/2", the COG U-value is .163.
    When looking at whole window U-values instead of COG, there is an even bigger improvement.
    Even with a superspacer, the U-value through the spacer is much worse than COG. So a total spacer thickness of 1.5" will have 33% less heat loss than 1" (2 x 1/2" spacers).

  54. David Jenkins | | #54

    New TripleGlazed Windows
    We recently replaced all of our old large casement windows with triple glazed windows from Loewen and Accurate Dorwin. Our triple glazed Accurate Dorwin windows cost about the same as double glazed Marvin fiberglass windows and the Accurate Dorwin quality is excellent! The Loewen windows were a bit pricier, but that is to be expected due to the douglas fir construction. We are thrilled with these windows.

  55. Jay | | #55

    choosing windows for each side of a VT passive house
    I'm in the final stages of my design. I've gotten some good advice from this forum so far (thanks again, esp to Martin) but am in a little bit of a quandary about how important it is to select for different U and SHGC dependent on which side of the house the windows will go on.

    The house will be approx 30ft x 48ft with the long axis aligned to true East-West. The south side will get full sun all day. The west side is also unshaded. Both south and west walls will be walk out basement with ICF, north and east basement will be buried. The east may be shaded early in the AM but if the sun ever decides it wants to shine while I'm up on my lot then I can probably identify and remove some trees to reduce or eliminate this.

    I'm probably going to go with either Inline 325 (either Dual Low E Hard Coat or Triple Low E Hard Coat x2) or Thermotec (possibly 211 gain #3, 322 gain #3 #5 or 321 gain #5). Most windows will be out swing casements (as I understand these seal better against air infiltration) and a few fixed (like on the top of the west wall gable end) plus one sliding door, 2 double french doors and 2 inswing standard doors type TBD.

    I've seen a few things here saying that even for north facing windows the higher ER rated ones can have a net gain due to passive light. Only one paper was cited and even that paper seemed to be based on a single side by side study. So is there really any strong, conclusive evidence that choosing a higher ER rated window for the north side (as opposed to going with optimizing U values without concern for SHGC)?

    Likewise, on the East and West walls can anyone give me an approximate quantification of the effects of choosing windows that balance U and SHGC to maximize heat gain versus just optimizing the U values alone?

    Is it only a modest gain like 10% or less, or despite the fact that at least on Dec 21st the sun won't even directly strike those walls and for much of the rest of the winter will still be at a low angle to them, can you really achieve a substantial heat gain on the East and West sides of a home?

    BTW the reason I ask is because sometimes discussions on the various energy efficient building forums I've seen focus on what is technically possible and don't always consider the cost vs benefit trade offs. Or at least it isn't always apparent from the discussion whether the tradeoffs were considered.

    Because these kinds of windows are pricey (initial cost estimates appear to be about $60/sf based on casing dimensions) IMO I'm not convinced it makes sense to spend an extra 20% or more (which can be $300 or more per window) to buy windows that maximize heat gain on certain walls when it might take 20yrs for that extra heat gain to pay for the additional cost.


  56. GBA Editor
    Martin Holladay | | #56

    Response to Jay
    1. Don't sweat the decision on the north windows; it doesn't really matter what the SHGC is to the north.

    2. It's fine to use low-SHGC on the west side if you want, even in Vermont.

    3. South, of course, must be high-SHGC.

    4. In Vermont, I believe that high SHGC windows make sense on the east side. Mornings are cool, even during the summer, and a little morning heat is almost always welcome.

    5. When in doubt, model your options with RESFEN or PHPP.

  57. DoctorBeer | | #57

    ER vs u and SHGC in comparing windows for passive solar
    Thanks for the tips.

    For the Canadian windows I notice they use a rating system called ER. From what I've read it's supposed to account for the U and SHGC values and give you an indication of total performance based on calculations done for a standard window size. If I understand this correctly it shows you the relative net gain of a window.

    So, for example, a window with a lower SHGC can actually outperform one with a higher SHGC (for passive solar applications) if it has a higher U value that's enough to reduce heat loss by a large enough amount to more than offset the lower passive solar heat gain, and thus result in a higher net total gain.

    The windows I'm considering have ER values for casement types starting at 30 and going to 41 and for fixed starting at 30 and going to 48.

    Are there any caveats to know when using ER as a means of comparison?


  58. GBA Editor
    Martin Holladay | | #58

    Response to Jay Hersh
    I think you have a good understanding of ER. If you live in a cold climate and you care about solar gain during the winter, the higher the ER, the better.

    I don't know if it's a caveat, but an argument can be made that ER is not very relevant for north-facing windows. I could easily understand a decision to specify a north window based on U-factor without regard to ER.

  59. Domenic | | #59

    best windows for sound
    live on highway in new house sound driving me crazy...some recommendations pls

  60. GBA Editor
    Martin Holladay | | #60

    Response to Domenic
    I suggest you post your question on our Q&A page. That way you are more likely to receive answers from GBA readers. Here's the page:

Log in or become a member to post a comment.



Recent Questions and Replies

  • |
  • |
  • |
  • |