The Devil is in the Details – Energy Model for Home with Custom Field Built Windows

The Devil is in the Details – Energy Model for Home with Custom Field Built Windows

Posted on 31. May, 2011 by in Case Studies, Windows and Glass

Remember last week, when we were talking about glass houses? Well, here’s another Title 24 case study on a 4,500 SF house, also from Swatt|Miers Architects. This house had almost 60% glazing to floor area, much of it custom built on site: 564 square feet of single paned butt glazed corner windows, 540 square feet of frameless glazing, a steel framed window, a 30 foot tall translucent window in a stair tower, 300 square feet of skylights, and a custom built wood screen interspersed with glass panels. That’s almost 2,700 square feet of glass.

And, to make the challenge that much more… piquant… it was in California climate zone 2 (Sonoma – HOT)… AND, they needed to beat California’s Title 24 energy standard by 15% because of local ordinances. It was the combination of all that single glazed area with the climate zone that concerned us the most. But, we had a reputation to maintain, and our motto to designers was, “We’ll never tell you that you have to shrink your windows.”

(Above image courtesy Swatt|Miers Architects.)

Well, we did end up telling them that they would have to find a way to make the butt glazing work with double panes (which we described in our article on window performance). So I suppose now we have broken our cardinal rule of never impacting the visible design.

Complex Shapes

Although the design was entirely based on rectangular planes, the volumes didn’t always line up one stacked directly over the other. This meant that there were some extra floor and roof areas to account for, and there were some subtle variations in building height, too. Most of the time, this can be generalized, but in this case we wanted to be as exact with every surface area as we could, so that we could claim the maximum thermal mass credit. I knew the planners might be reviewing the report against the drawings with a fine-toothed comb, and we needed to be prepared to respond to any comments with a solid grasp of facts.

Wood framed overhanging floor areas were modeled separately from the slab flooring on the main level.

Start at the Beginning, Grasshopper

As we mentioned in the previous case study, we try to start at a basic level with whatever systems information we have from the designer, and then work up from there. The main heating system was radiant, with A/C. On the plus side, the design called for slab flooring, with gypcrete on the upper level – this thermal mass gave us a ray of hope. Even so, the first trial was dismal. 73% below the standard. Heating was missing by 75% and cooling was missing by a whopping -136%. Only pride kept me from throwing in the towel – pride and curiosity.

“Patience, Grasshopper,” I said to myself. “Just do what you usually do, and don’t say anything until you have some good news to report.” So, here’s what we did, and what worked the best.

Cool Roof/Radiant Barrier. Although this seems like a minor place to start, we’d have to try a cool roof at some point, and it might actually help in Sonoma. The cool roof did make a difference (down to -67%), although the cooling improvements were offset by a small detriment on the heating side. What we really needed was a selectively cool roof, that changes color based on outside temperature – maybe someday soon there will be such a thing. The radiant barrier helped less, and they would have had to change the roof construction to include it. The gains from the barrier didn’t justify including it – unless we absolutely had to.

Wall insulation. Next, we upped the wall insulation from the requisite R13 to R21. That pushed us from -67% to -55%. Pretty good, but still way behind. The designer had thoughtfully provided us with the wall assemblies, so we knew that the cavities were 2 x 6 – large enough to fit R21.

The designer provided us with complete details, which helped us to ascertain how much insulation we could specify in the energy model. Image courtesy Swatt|Miers Architects.

Ducts. There would be both heating and cooling ducts in this project. Although the main heating system was radiant, there would be a forced air backup. We couldn’t model both systems, but we had to keep the heating ducts in the energy model, which cost us. We did verify with the designer that the ducts would be located within conditioned space, which gave a credit. And, we added the HERS test for duct leakage, which brought us from -55% to -40%. (Actually, we tried eliminating the heating ducts and it didn’t help as much as it had on other projects.)

Blower Door Test. With heating and cooling ducts within conditioned space and the duct test taking us to -40%, we added another HERS test – the blower door test, which measures the airtightness of the entire home. Doing these HERS tests on a sizeable house such as this was bound to be challenging, so we stressed to the designer that they, and their builder, should read our article on HERS tests so that they knew what was involved. The builder in particular would need to know that the project was required to pass these tests. The blower door took us from -40% to -37%, not that much. Well, we’d keep it in for now, since it was looking like we’d need every last inch of compliance.

A/C Verifications. Next, we tried adding the HERS tests that apply to air conditioning systems: test for refrigerant charge, airflow, fan watt draw. These took us from -37% to -32%. I had hoped for more. We tried upping the A/C SEER from the standard 13 up to 18 SEER, which, together with the HERS tests, brought us to -29.5%. One thing to note is that often, the HERS tests have more of an impact on compliance than simply upping the SEER. But, all of this was simply postponing the day of reckoning, which was to attack the windows.

Window Performance

As with the other Swatt|Miers case study, we divided up all the window areas by type: Butt glazed corners, frameless wall insets, the stair tower, the custom steel window, various sliding pocket doors, operable casements, the 40 foot long wood screen window on the upper gallery, and the skylights. The design called for various overhangs, including a large canopy extending over the main house and a separate guest house. Most of the window framing was metal, which is not as good an insulator as wood.

Initially, I used the performance specs from Efficient Windows as a starting point for estimating all the custom areas, assuming that all windows with the exception of the corner butt glazing would be double paned, low e glass. There was a lot of back and forth with the designer to establish the composition and construction of the various custom windows. We couldn’t go any better than the standard on most of these (we were lucky to get the standard). The casements and the sliding pocket doors were Fleetwood window products, with numbers that we could look up.

I asked around and searched for information on whether any sort of single glazed window could ever be “high performing”. Alas, there was no magic glass. The experts all informed me that the main factor in window performance is 1) multiple panes with insulating layers of air or gas fill 2) airtightness of the frame itself and 3) insulating properties of the framing material. There was no such thing as a thermally  broken, metal framed window with single glazing, because why? There wouldn’t be any demand for it.

But we were still at -29%. Something had to give. So, I broke our rule and made all those butt glazed corner windows double glazed. That took us to -11%. And then, I modeled the Fleetwood windows using the best numbers they had available for each type. That took us to -6%. At the same time, I put out the word to see if anyone had successfully built a corner glazed window with double paned glass, because I knew that the designer really wanted to keep that transparent appearance, and putting a spacer bar on the corners would be a major disappointment, to say the least.

And it still wasn’t enough. Even adding the dreaded QII test (a HERS test where every bit of insulation is inspected as it’s installed during construction) wasn’t enough, although the QII did bring us from -11% to -0.8%. So close – and yet not close enough, considering that we had another 15% to go.

Interior Mass Surfaces

At this point I dragged Mark English over and made him review the entire drawing set plus all the details. As an experienced architect who’s been designing and building homes for 25 years, I figured he’d see a few things that I had missed, and he would make sure we didn’t take too many liberties with the wall and roof cavities.

Based on consultations with Mark and numerous exchanges with the designer to verify the exact location of every wall and floor finish, we added the thermal mass of a dramatic 2-story stone veneer wall, over 1475 SF of thermal mass. Additionally we included all the tile flooring in the bathrooms, countertops, and the gypcrete from the second floor’s suspended slab floor. Even though this floor was largely covered by wood or carpet, it still yielded some credit. That took us from -0.8% to +12%.

Upon our request, the designer provided us with the location of vertical thermal mass surfaces - stone veneer walls - which we could then include in the energy model. Image courtesy Swatt|Miers Architects.

And it still wasn’t quite enough. I felt like a magician reaching into a hat for another rabbit and coming up with a hamster instead.

What the Designer Said

I figured it was time to fill in the designer with our progress to date, and test the waters about making the butt glazed windows double paned. It might be a good time to insist on an uber-efficient water heater. We’d actually started with a reasonably efficient one, a .80 energy factor, but without further information, I was hesitant to commit to anything extreme. Eventually we would have to include the actual models they were using, and there would likely be more than one with a house that size anyway. I don’t think they had worked out the mechanical systems to that detail, so now was a great time to test and suggest a few things.

But before we reached into our top hat for that last rabbit (the water heater itself), we tried a few more things just to see what would happen.

Solar water heating credit. The design hadn’t specified solar equipment of any kind. Well so what? Maybe it would let them keep that single glazing, although I doubted that.  Even though there’s no credit for the use of renewable energy for electricity or heat, there is Title 24 credit for solar hot water. It’s based on the percentage of hot water that the home is expected to get from solar, and sure enough, set this percentage high enough and the compliance score improved.  So, by pushing this number to an unrealistically high 50% we were able to inch our compliance from +12% over to 17% over, although I doubted that this would actually work.

And why not? Well, the problem with it is that you still need some kind of indirect storage tank to ensure hot water in the evening, unless you only plan to shower at high noon. In addition, this home would have extra water heating demand because of the radiant heating. But hey… we reached our goal, in theory at least.

Higher Solar Heat Gain on Some Windows

On the last case study, I had, purely out of curiosity, tested a series of window performance combinations just to see what would happen. Although a low U value window was always the best choice, because it provided thermal insulation for both hot and cold temperatures, what could we do if the best U value we could find was average – if that? Since we had so many custom windows that would be built in the field, we couldn’t make aggressive assumptions about them.

So, I inched up just the solar heat gain on the Fleetwood windows, while keeping the U value the same. This would allow at least some glazing areas to keep up their insulating value, while allowing a little more solar heat. Although originally the design had lagged more on cooling, it was now the heating side that had all the shortfall. We were actually ahead on cooling. This latest change brought us from 17% over to 18% over, so it didn’t make a huge difference. Heating was better, cooling lost a bit.


The skylights were another big unknown. There were a lot of them, and the designer indicated they didn’t want wood framed because they were concerned about leaks and such. Then they selected a manufacturer who actually had pretty good numbers, which brought us from +18% to +21% over.

Highest Efficiency Boiler

It was time to pull out our very last rabbit, which was to boost the water heater performance as far as it would go. By specifying a boiler with a 95% efficiency, and keeping a solar hot water credit at 25% (still probably too aggressive), we got the house to exceed Title 24 by 32%. At this point, we dialed back the Fleetwood windows to the actual numbers (Westwood and Norwood product lines, dual glazed low E, thermally broken, but no argon – air fill), and removed the solar hot water credit altogether. We kept the HERS testing for credits: the three A/C tests and the duct test.

The final score? It came out as exceeding Title 24 by 25%. This would give them a little margin if by some chance it didn’t pass every HERS test. Some of those tests were worth GreenPoints, too. So, chances are this house will earn a respectable GreenPoint Rated score as well once construction is complete.

A few final notes follow, on what the designer did to help us, and a note on modeling multiple zones.

Detailed Wall and Floor Assemblies

This was one of the few projects where the designer provided us with detailed wall and floor assemblies, 3 drawing sheets of it. This was great, because we could see exactly where the gypcrete was, which portions of the floor were covered (and with what) and which walls had interior stone veneers.

Although we don't always have to model every layer in the floor, in some cases it helps to know exactly what's in there. Image courtesy Swatt|Miers Architects.

Modeling Multiple Heating and Cooling Zones

This was something that we would have resorted to only at gunpoint, because it’s very time-consuming. Essentially, you model each zone as a separate volume – including floors, walls, and ceilings – as if it were its own little house. Title 24 gives credit for this, but I’ve only ever done it once on a private home, and we had no way of knowing how much it would help unless we tried it.


One Response to “The Devil is in the Details – Energy Model for Home with Custom Field Built Windows”

  1. Fiberglass Services Delhi

    02. Jun, 2014

    I like your blog very much because you shared a nice post which is helpful to everyone to build their custom windows at their home. The performance of the custom windows are better than compared to normal. I’d like to say thanks for sharing this informative post. Keep up blogging.

Leave a Reply

Read previous post:
Can Glass Houses Be Energy-Efficient?

Actually the real question is whether an all-glass pavilion can still comply with the new version of California's Title 24...