This is the third in a series of posts by Craig Anderson describing the off-the-grid house he built with his wife France-Pascale Ménard near Low, Québec. Craig writes about the “Seven Hills Project” in a blog called Sunshine Saved. For a list of Craig’s previous posts, see the list of “Blogs by Craig Anderson” in the sidebar below. This post originally appeared in November 2015.
Being a newcomer to living off the grid, I did not make the best initial choice for our heating system. In order to help others avoid the same mistakes, I’ll lay out what we did and how it was problematic, as well as how we went about fixing the problems.
What we installed initially
We installed a hydronic heating system powered by a propane-fired boiler (a Trinity LX150). The system heats a concrete floor in the lower level with four zones (one for each bedroom and one for the bathroom), and includes two hydronic baseboard registers upstairs. This system provides amazingly warm and comfy floors in the bedrooms.
Some people with heated floors don’t get to have the warm-foot experience very often with high-efficiency homes, but this has not been the case for us. As the house is used primarily as a weekend retreat in the winter, we turn up the heat considerably when we first arrive. We have to wear warm sweaters when we first get in the door, but then have a full day where the floor is toasty warm.
The secondary heating system, and the one that I enjoy using much more, is a free-standing wood stove, a Jøtul F3CB to be precise. It is a relatively small (42,000 BTU/h) high-efficiency stove made in Norway, but it is more than sufficient for our well-insulated home. The stove is located in the open-concept upstairs, and in just a few hours it can take the 1,000-square-foot high-ceiling space from sweater temperature to shorts weather.
We have only one full winter of use to measure our consumption of propane and wood, so it is difficult to draw any big conclusions. Still, so far it’s been about as we had expected. In the 12 months up to November 2015, we burned 400 gallons of propane. Primarily, that was for heating the house, but it also went to domestic hot water, the backup generator, and a propane range in the kitchen. My best guess is that 75% of that, perhaps 300 gallons, went to space heating. For the wood stove, we burned just a bit less than a cord of wood, and had a fire in the stove during at least part of the day most days we were there.
One problem was the system’s complexity
The big problem with our heating system was that it was relatively complex and brittle. We aren’t there all the time to run the wood stove, meaning the boiler really needed to carry the load. The problem is the system requires a constant, and quite significant, supply of electricity at the time of year when it is most difficult to generate power from our photovoltaic array. (For more information on this issue, see “How to Live Comfortably Off the Grid.”)
Running full power, the heating system requires approximately 400 watts of electricity for the boiler and circulation pumps, meaning that if it runs for 10 hours per day (which can happen on the coldest days of winter), the heating system alone needs 4 kWh, almost the full amount of our daily target of electricity.
The other problem is the hydronic system is sensitive to freezing, which occurred last winter (our first full winter). We lost power and had a few frozen water pipes, as well as a break in one of the hydronic heating lines. There was glycol in the mix as an antifreeze, but apparently the installer did not put enough in. It did take a very serious set of combined circumstances to bring down the house — it was the coldest week of the year while we were away visiting family at Christmas; several days of snow covered the solar panels and prevented any electricity from being generated; and finally — the last straw — the generator broke down.
Making the house more resilient
We had no desire to repeat the emergency situation we found ourselves in for a good chunk of last winter, so we’ve taken quite a number of steps to make our home more resilient in the face of future mechanical problems.
First, we installed a monitoring system so we could get daily status reports via email. They include conditions of the PV, including power generated and power used, generator run time, and battery temperature. These daily reminders tell me how the system is functioning, and I know that if I fail to receive a report, there’s a problem with either the power or internet systems.
Second, we increased the number of photovoltaic (PV) panels. During the first winter, the generator was needed relatively frequently over a six-month period from the fall through the spring — far too often for my taste. Therefore, we more than doubled the capacity of the solar system, from 2,820 kW to 5,820 kW. (More on our PV system next time).
Finally, we added a new backup heat source that would not be dependent on either our being there every day or electricity. We did this by using an older, simpler technology, a direct-vent propane wall heater. These have been used for many years in garages, workshops, cabins, and quite a number of off-grid homes. If I had done more research about off-grid heating, or if I had been given better advice, I might have decided to handle all of our heating needs with a couple of these heaters from the beginning.
Their biggest advantage is they require no electricity at all to function. They have a pilot light and a millivolt thermopile thermostat, which uses a temperature gradient to produce the small amount of current needed for the thermostat, and they rely on convection to circulate air past the heating elements. We have installed one to provide for some of the base load of heating, as well as ensure the house would never freeze again, regardless of any issues with the electrical system.
Sizing the new heater
I relied on the boiler company to help make this decision for the original heating system, but this new installation was a much more hands-on endeavor for me. We’re lucky to have a good energy model of our house, needed for the LEED certification (to be discussed in a future post). This shows an estimate of 33,200,000 BTU of heat needed per year for the whole house, which is about a 70% reduction over a similarly sized house built to code.
Heaters are generally rated for the BTU they can produce in an hour (BTU/h). To get an initial approximation of our heating needs, we take the heating load for the year, divide by 100 days to account for the heaviest part of the heating season, and divide by 24 hours in a day to account for a heater running full time. Once we did the math, we got 13,800 BTU/h.
This calculation would be assuming we heat the entire house to 70°F (20°C) with just the wall unit for the entire winter. In actuality, we will keep a lower set-point, and this unit will instead keep just a portion of the house at about 65° to 70°. That allows the rest of the house, heated by the hydronic system, to be cooler when we aren’t there. This calculation is just an estimate of the average heating load, so this amount of heat wouldn’t be able to keep up with heating the whole house on the coldest days of winter.
We selected the Empire DV215 heater, rated at 15,000 BTU/h, which sits in the central bedroom, radiating heat out to the rest of the lower level.
Wall heater was a success
By March 2016, with winter on the run, I knew the wall heater was an amazing success in terms of reliability and reducing the use of electricity with our heating. The heater was able to carry essentially all of the heating load for the house during the weekdays when we were often in the city. The heater was placed in the kids’ bedroom, and set to around 68°F. Upon our arrival after a few days away, the adjacent rooms were always 60°F or warmer, and the upstairs was always warmer than 50°F.
From a starting point like this it was quite easy to turn up the boiler, start a fire in the wood stove, and be down to shirtsleeves in no more than two hours.
One other new (more like forgotten and found again) fact is that our energy evaluation also included a calculation of peak heating loads for the home and boiler system, which actually matching quite closely to the calculation I did above. The HERS calculation of peak heating load for our house is 23,400 BTU/h, which including the caveats that I mention above. This is the amount of heat that would be needed to keep up on the coldest days of the year, not the typical winter day that I tried to estimate for.
This peak load calculation also showed our actual boiler specification, with a rated heat output of 136,000 BTU/h. This is nearly six times our maximum heating load and is majorly overkill, but I have heard time and again that heating and cooling contractors usually overbuild these systems, and our house does require a much smaller heating load than the standard home.
Get building science and energy efficiency advice, plus special offers, in your inbox.