A deep energy retrofit turns a summer cottage into a snug full-time residence
In 2008, contractor Dwight Holmes went to work with homeowner Mary Sargent on a deep energy retrofit of her 1917 “Tea House.” Nestled into a hillside apple orchard in Vermont, the building was originally constructed as a private, summer-only, communal dining and entertainment hall with living quarters for servant help. The house included many doors and large single-pane French windows to allow summer breezes to pass through the building. While this is ideal for the summertime, maintaining a livable temperature indoors during Vermont winters was a challenge.
Mary (and her dog and two cats for mouse patrol) love wood heat — but also hoped to buy, haul, and stack less wood each season. Thermal comfort was also an issue. Many times during the winter, Mary would return home at the end of the work day to a house that was 45°F inside, with the wood stove fire long gone. The existing building had mostly bare 2x4 stud walls; the only insulation was fiberglass in the walls of the core rooms of the building on the first floor. On more than one occasion, ice dams formed large enough to break off and smash basement windows. The project's primary goal was to expand the thermal envelope of the building to include all of the occupied floor space — and to do so in a super-efficient way.
Energy modeling and budget
Although Mary had a healthy budget for her deep energy retrofit, Dwight knew that the cost of all the energy upgrades would quickly add up. To optimize the design for energy efficiency, energy consultant Doug Snyder used Energy 10 software, a modeling tool well suited to passive-solar projects measuring under 10,000 square feet.
Snyder ran about 20 simulations to compare the energy performance of different building modifications and to explore diminishing returns. The energy modeling results established the optimum combination of dollars spent for energy saved. The plan called for aggressive air sealing, R-20 insulation on the foundation walls, R-40 insulation in the upper story walls, R-60 insulation in the roof, and triple-pane low-e windows.
Energy 10 modeling projected heating savings of 81% over an estimated baseline. (Determining the baseline for the project was difficult, however, since only about half of the home was insulated and the wintertime interior temperature was kept unusually low and would fluctuate greatly.)
Balancing passive and active solar
The site enjoys good solar orientation and access, so solar power was an early consideration. The existing structure had a large sunroom with two stories of south-facing glazing. The sunroom seemed to offer a great opportunity to capitalize on wintertime solar gain.
Since the green building consultant, general contractor, and solar thermal and heating system designer were all concerned about temperature swings in this area, they decided to insulate the two-story interior partition wall separating the sunroom from the remainder of the house. Operable French windows were reused and installed in the upper portion of the partition wall. As the sun room heats up in the wintertime, the French windows can be opened to share space heat generated by passive solar gain with the rest of the house. While the SHGC of the south-facing glazing was not optimized for solar gain, the sun room still heats up to 80°F on sunny winter days. A thin floor of stained and detailed concrete was poured in the sunroom to increase the room's thermal mass.
Building assemblies designed and specified for durability
I think Dwight and I did a decent job of looking at vapor barriers and condensation and drying potentials as we worked out the wall design. I even had some email dialogue with John Straube at BSC regarding drying potential of wood components if spray foam were used as the cavity fill coupled with polyiso exterior foam board. I think eliminating the need for an interior vapor barrier by using exterior sheathing insulation is elegant.
High performance heating with wood stoves?
The original heating system included two wood stoves: one large stove in the basement and a second stove in the main living room on the first floor. The new heating system consists of a wood boiler in the basement and two flat-plate solar hot water collectors.
Both heat sources feed a large built-in-place 600-gallon water storage tank that is insulated to R-60. The domestic hot water tank (with a backup electric element) pulls water from the larger 600-gallon tank.
When the thermostat calls for heat, the wood boiler heats the home directly; otherwise it dumps heat into the large 600-gallon tank. Because of cost constraints, a Marathon brand wood boiler (with an efficiency of around 70%) was chosen. A more efficient (80%+) but more expensive Tarm boiler was a budget-buster.
The multiple-zone heat distribution system consists of radiant tubing under the first floor and baseboard radiators on the second floor. The radiant floor seemed ideal for the first floor because most of the space has a high cathedral ceiling.
There is also a small sealed-combustion wood stove in the living room for backup heat in case there is a power failure and the boiler and associated pumps cannot run. Since the tight home has less natural air infiltration than it used to, some effort was made to find a sealed-combustion wood stove so that backdrafting would not be an issue.
Careful air sealing and superinsulation helped the heating system contractor design a system that was smaller that what would have been needed had conventional insulation levels been used.
As is typical during the first year of operation, the boiler and solar hot water systems had some glitches. The solar system needed to be tweaked; a faulty check valve in the piping circulated warm water to the cold panels at night for several cold months during the heating season. This meant that the heat gained during the daytime by the solar panels, and perhaps even a some of the heat contributed by the boiler to the large hot water storage tank, was lost at night. The wood-fired boiler also required a modification to the control for the return water temperature to the boiler. The first winter’s fuel consumption exceeds what would have been used if the system had not had these initial problems.
During the first winter, the basement was (surprisingly) just as warm as the upstairs occupied space, even though no intentional heating elements had been installed there. Since this space is not occupied, it would be more efficient to save this heat for the upper occupied floors. Two suspected causes for this excess heat are waste heat coming off of the boiler and radiation downward from the radiant tubing installed under the first floor (in the basement ceiling). Better insulation underneath the tubing could help this situation, and this will be considered for future work.
The ERV began making some noise as construction was wrapping up. Metal components on the interior fan cages were frosting up and the increased friction was causing the noise. This was a surprise since, ERVs are less likely to freeze than HRVs. After calling the manufacturer and doing some troubleshooting, we figured that the freeze-up probably occurred as a combined result of very cold winter temperatures and moisture-of-construction in the home (recently installed damp-spray insulation and the drywall compound applied to all of the newly finished interior walls). The frosting over of components inside the ERV did not recur throughout the remainder of the winter.
Mary wishes she had known more about renewable energy systems at the very start of the project. “It has taken us the better part of two heating seasons to work out all the mechanical issues,” Mary says with just a hit of a lament. “But today, the system works much better and I understand a lot more about how to get the most out of the wood boiler.”