As far as being a beautiful place to live, the dome is a complete success. As an example of superinsulated passive solar design, though, it falls a bit short of optimal. For one thing, it's not as superinsulated as it was supposed to be. For another, there's essentially no passive solar gain! Fortunately, I think I can make up for the lack of passive solar by adding some active solar.

Here's the whole story:

Spray Foam in 2" Layers -- and Get That in Writing

Spraying the foam insulation was one of the few jobs I couldn't do myself because the equipment is so expensive. Do-it-yourself foam kits exist, but at that time they all used propellant harmful to the ozone layer, which I wanted to avoid, and for a large job they cost more than hiring a professional.

In Canada, where thick spray-foam walls are common, foam contractors know that closed-cell polyurethane foam needs to be built up in 2" layers, with no more than two layers applied per day. Unfortunately I didn't know this, and neither did the foam guy I hired. Starting at the top of the dome, he sprayed each triangle with the whole 10" at once. This was bad for two reasons.

First, the foam wasn't able to expand like it should, so he went through his chemicals much faster than expected. Rather than absorb this cost himself, he waited til I was out of town, then finished the job by spraying a lot of cheap fluffy open-cell foam, hidden under an inch or two of the dense closed-cell foam he was supposed to be using throughout.

I didn't discover his deception until after he had cashed my check. Instead of the uniform R-60 I paid for, I got the dome equivalent of an R-60 ceiling and R-40 walls. That still qualifies as superinsulation, but I'll never know how much better the dome would have performed if it had been insulated to my original specifications.

Getting ripped off was bad, but the other drawback of not shooting in layers was even worse. Foam has a fishy chemical smell, and shooting in thin layers lets it escape. The dome reeked for months while I was working on the interior. The foam was applied in late 2006, and in 2011 we were still getting the occasional whiff of foam smell in hot humid weather. We can get rid of it by turning up the energy recovery ventilator, but it's obnoxious.

Better Planning for Solar in the Design Phase

As built, the dome really can't be considered a passive solar house.

I started building before I had fully figured out how I would handle the south windows. Our first summer in Iowa was a scorcher, and it made me very concerned about summer overheating. I made sure the south glass had overhangs, which was smart, but I also undersized it. The rule of thumb for passive solar superinsulated houses is to have the area of south glass equal to 8% of your floor area, or even more if there's a lot of thermal mass. We have under 2%.

Fortunately domes distribute light efficiently, so it doesn't feel dark inside unless the sky is quite overcast. However, on sunny below-zero days, my earlier superinsulated house would reach 70° inside just from sunlight through its triple-glazed plain-glass windows. This one never gets warm from winter sun alone, even on much milder days.

If an insufficient area of south glass was my first mistake, the windows themselves were the second. I found them marked down at an outlet store, and was charmed by how their unusual shapes could fit into dormers built around my available triangles. I overlooked their thermally-conductive aluminum spacers, the fact that they were only double-glazed rather than triple-, and their total lack of documentation. It seems clear now that at least two of them use the type of low-e glass that resists solar heat gain -- the exact opposite of what you want on the south wall of a passive solar house. I felt like an idiot when I figured this out, but an idiot with a lot of company. Apparently any window eligible for federal tax credits comes with a low-e coating specifically designed to sabotage passive solar gain.

It would have been better to special-order triple-glazed, argon- or krypton-filled windows with high-solar-gain glass and low-conductivity spacers.

On the other hand, even the best windows lose heat at night and on cloudy days, and Midwestern winters have a lot of cloudy days. I think the reason my first superinsulated house worked so well was that it had a convenient way to cover the windows with R-5 insulation at night, and I did that religiously.

There was a troubling incident with the first house's window insulation, though. One morning I raised the insulated shades off the glass, and a window's middle pane cracked when the warm inside air hit the freezing cold window. Tempered glass would have prevented that, but ordinary windows don't have tempered glass. Because of this experience, I have been reluctant to make insulated shades for the windows of the dome. I guess some people use patio door replacement glass for fixed windows subject to thermal shock, because that is tempered.

At this point, adding new windows for better passive solar is out of the question. For one thing, I'm pleased to report that we currently don't have much trouble with summer overheating, so that at least worked. Having relatively few windows also means that in winter the house loses heat quite slowly. For example, we turned off the heat and went away for three days at Christmas, with outside temperatures between 19° and 32°F. Friends thought we would come home to burst pipes, but it only dropped to 57.6° inside.

Rather than add windows that both gain and lose heat, I would like to add some kind of active solar heating system for both space heat and domestic hot water. Active systems either add heat or they don't, but they never conduct heat out of your living space. This would have been much easier to incorporate into the dome's design when it was still a CAD file.