Actually the real question is whether an all-glass pavilion can still comply with the new version of California’s Title 24 energy code. Although Title 24 has been around since the 1970s, it is only now that designers are feeling the pinch. Given the increasing strictness of the energy code, what can an architect do if he (or she) wants to create designs with dramatic glass curtain walls?
The “glass house” shown on the cover image is, of course, Philip Johnson’s famous Modernist masterpiece, also called the Glass House. Even that house could, with the right high-performing window system, comply with Title 24 requirements – I tested it out. But, let’s talk about some more current designs for our case study.
When we got our first Title 24 project from Swatt|Miers Architects, I didn’t really know if we’d be able to finesse it. We had never had a project quite like it before. But then I remember seeing an award-winning Net Zero home in Truro, Massachusetts (by Zero Energy Design) which had acres of glass everywhere. If they could make it work in a cold place like Massachusetts, surely we could do the same in California!
Our first Title 24 project from Swatt|Miers Architects was a 6,000 SF all-glass pavilion with a generous roof, a custom window wall on one side, and heated slab floors. On the plus side, the roof included large shading overhangs, and we had ample time and flexibility to select high-efficiency heating and water heating systems. There were no large skylights or ductwork to worry about, either. The house was located in a moderate climate zone, so they weren’t even going to install any cooling systems at all.
Another big plus was they didn’t have any “beyond compliance” goals such as GreenPoints. So, we didn’t have to beat the standard by 15% or more. As you’ll see in a minute, that would have been possible, but very expensive.
On the challenge side, there was almost 5,000 square feet of glazing – 75% of the conditioned floor area. In addition, the design called for metal framed windows, which are inherently less efficient than windows framed in wood or vinyl. Even thermally broken metal windows can’t always match the performance of wood.
Strategy: Start Low, Increment Up
We usually start by running each project with baseline assumptions, usually the same ones used for the prescriptive Title 24 method. The Title 24 prescriptive method sets forth minimum requirements for things like heating and cooling performance, insulation levels, maximum allowable glazing areas, and window performance – but it’s all or nothing. Either you meet every requirement, or you have to use the performance modeling method to satisfy Title 24. The software model has some built-in generic inputs for things like water heaters that also assumes a very basic level of compliance.
In this case, the first trial missed by -76%, with shortfalls on heating and water heating. Surprisingly, cooling was not a problem – at least, not in the model. Still -76% seems like a big, scary number – one which we attacked incrementally.
Preliminary Trials with Insulation and Water Heating
Increasing insulation levels from R13 to R21 brought it to -67%. There wasn’t that much wall to work with, the walls being mostly glass, which is measured by window performance rather than insulation. The roof already had plenty of insulation, and adding more past a certain point had a negligible impact.
Just as a test, adding an overhang over the South window wall (which didn’t have one) actually made it worse! That’s one thing about trade-offs. Sometimes a measure that helps with cooling will create more load on the heating side, and heating was what we needed to fix. But this design didn’t specify an overhang there, and even when there are overhangs, modeling them is optional.
Next, we tried upping the water heater performance. Originally we’d specified a generic setting, because the mech hadn’t been worked out yet. For a house this size, over 6,000 SF, you probably wouldn’t have a single heater anyway, and you definitely wouldn’t be running all the hot water AND the radiant through one heater. So we had started with a 75 gallon storage tank with a .75 energy factor – legacy numbers.
A tankless heater with a storage tank, with 2 separate systems, and a higher energy factor, got us down to only -49% below Title 24 requirements.
Main Trials: Windows
But all this was simply prep, because the windows were going to be the biggest challenge. The window performance factors that are important for Title 24 are the U value, which measures thermal performance (keeps winter heat in), and the Solar Heat Gain Coefficient (keeps summer sun’s heat out). We had all been thinking that with all that glass, cooling would be the concern, and we’d have to get the lowest SHGC we could find. Not so. It was the U factor on all those windows that would make or break our compliance.
In fact, having a higher SHGC might help because we could afford to lose some cooling margin if the solar heat gain would actually help on the heating side. Passive solar designs in far North latitudes try to leverage the sun’s heat, especially in the wintertime when there’s so much less of it. We don’t normally think that way in California, but occasionally a higher SHGC can actually help, especially on the East wall. The reasoning is that solar energy on the east, when the sun is just rising, can help the house to recover from cooler nighttime temperatures.
The Title 24 prescriptive minimum for window performance is .40 U/.35 SHGC, lower being better for both these numbers. That’s pretty tough, considering that most ordinary metal framed windows run around .50 or .60 U factor, even for double glazed, “Low E” glass. We had started with the generic “metal double glazed low e” input which is around .65/.40. This project specified Fleetwood windows, which has several product lines with a wide array of options for glazing, thermal breaks, gas fills, and even glass types. All of their test results are easy to find, too.
We also modeled each window opening separately (over 77 of them), so that we could use different performance numbers for each window type: there were Casements, Awnings, Fixed, Sliding, and a couple of custom settings. That way if a better performance could be had from the fixed windows but not from the operable ones, we could include it. Sometimes clients have also asked us to use different settings for different cardinal directions, for example, using a more expensive but spectrally selective glass on a western or southern wall where it might matter the most.
U Factor or SHGC – Which Is More Important?
Although we were more concerned with the U factor, it was possible that testing different combinations of U and SHGC factors might yield some interesting information. So, we spent a lot of time trying very low U/low SHGC combinations. We also tried some average U with very high SHGC windows, as well as trials where we kept one number the same and varying the other one up and down to see what would happen.
A low U factor was the best choice, regardless of SHGC – but, this might not be achievable, even if they used the more expensive, thermally broken frames. Of course, triple glazing with argon gas fill would help a lot, too – I’m not sure what the quote was for 4,500 square feet of windows with triple glazing/argon/thermal breaks, but it would have killed the project budget. We had to find some way to make it work with double glazing.
A U value of .35 or under would be better overall. A U value of .25 would be awesome. If the U value was low enough, like .30, it didn’t matter as much what the SHGC was. In fact, with a U value of .25 some of the runs came out almost 40% over compliance! Some interesting window results: With a higher U value, a moderate solar heat gain acted to compensate somewhat for heating loss. It was almost as if there existed a second, smaller “peak” of outliers which, if I had a lot more time, I’d try to graph in more detail.
Of course, it’s all very well to say that we need a window with a U value under .30. The U value was the one thing that most limited us, because with the window products and framing materials selected, it simply wasn’t possible or cost-effective to get down that low. And, specifying an artificially high SHGC just to “make the numbers” didn’t make sense. But then again, neither did specifying triple glazed, argon filled windows in a mild California climate zone.
Custom Window Wall
The custom window wall on the South wall was another challenge. Fortunately, this area was wood framed – not aluminum like the rest of the windows. The wood framing would help with thermal performance, and the plans already called for double glazing with Low E glass. However, since there were no formal NFRC test results for it, what numbers could we safely assume for performance? As it turns out, we checked with the local Planning department for the project’s location, and they indicated that using the Title 24 prescriptive standard, .40 U/.35 SHGC, would be acceptable. This standard already assumes a wood framed, double glazed, low E glass window – so it made sense to use it for the custom window areas.
We might not be so lucky in another jurisdiction, though. The plan checkers have some leeway to use their own judgment and, if we’d been in another area, we could have been asked a lot more questions about materials, glass types, etc. – and possibly we might have had to use less forgiving numbers.
In the actual project, we tried a few more things. One was a super-efficient water heater. Most water heaters are in the .65-.80 range for energy efficiency, but there are some out there that go almost to .95, so by creatively assuming that they would use the best systems available, we could at least present that as an option for compliance.
We called out each area of interior exposed thermal mass – exposed stone, tile, and concrete – because usually this helps with Title 24 calculations.
And, as a last resort, we could have called for some HERS verifications for extra compliance credit. Our article on HERS testing describes each of these tests in detail: duct blaster, blower door, QII, and various tests on the A/C system. Of course, not all these tests were available on this project. With radiant heating and no A/C, there were no ducts to test. And, the various A/C tests – refrigerant charge, fan watt draw, airflow, and verified EER – couldn’t be used, either.
Would Cooling Be an Issue in the Real World?
Initially we were surprised that cooling was not a problem, and we were all wondering how realistic that would be once the house was actually built and inhabited. Title 24 intentionally ignores some real-world conditions such as landforms and shade trees, but it does apply location-specific climate data on top of the broader “climate zone” designation. The designer expressed concern that with a high SHGC window, the house would be too warm during the summer, especially along the South side. And really, to us it didn’t make sense to use an artificially obtained number when the most straightforward thing to do would be to use California standard Low-E, low SHGC glazing.
Would Philip Johnson’s Glass House Meet Title 24 Energy Requirements?
Just for fun, I did a quick test of Philip Johnson’s Glass House, using floor plans and dimensions that I was able to find on Google… no guarantees on accuracy, but it seemed like it was worth a try. The trials included three California climate zones: Woodside (Zone 3), Tahoe (Zone 16) and Livermore (Zone 2) and a few compass orientations for 0, 45, and 90. For starters, I assumed double glazing and Low E glass, with the same numbers as a wood frame would be. And, for heating I assumed radiant heated slab flooring… which I think is the actual method.
The compass orientation didn’t matter that much – it mattered a little – but, the location had a bigger effect, mainly in the balance between the shortfalls between heating and cooling. No surprise, Livermore had the biggest cooling problem initially.
Setting the glazing to something more like the actual (metal framing, clear glazing) was disastrous for Title 24 compliance. However, when I changed the windows to the “best available” performance numbers – something around .20 U/.20 SHGC – the house complied in all three zones without any further changes.
The Catch – Custom Built Windows Can’t Use NFRC Test Results
In reality, it may not be possible to even obtain pre-manufactured windows of the size that are used in Johnson’s Glass House. And, in order for performance numbers to be valid and acceptable for Title 24, the window units have to be NFRC rated. We’ve discussed this in a previous article, so I can’t assume that the Glass House design as it is now would ever be able to fully comply with Title 24. And, of course, the Title 24 modeling software only has climate data for California, not Connecticut where the house is actually located. Nonetheless, it could comply a lot more easily than some other projects we’ve worked on.
Another Glass House: Olson Kundig’s Glass Farmhouse
As I was looking around for any accounts of the actual comfort of Johnson’s Glass House, I found another, much more recent take on the all-glass house from Olson Kundig Architects, located in Oregon. Alas, no reports on actual energy use or comfort here, either – but Kundig did employ solar design features and high-performing glass and took some care to adapt the house to the local climate. You can see some nice images and a description at The Contemporist.
About the author
Rebecca Firestone has been working in the Bay Area since 1998 as a technical writer, business content developer, architectural filing lady, marketing director, and sorcerer’s apprentice.