Basement Insulation and the Resulting Energy Savings

I had heard a lot of talk about insulating basements to save heating costs, but I had never seen any particularly informative data. I looked at my heating costs and decided to insulate the basement and document how basement insulation affected my home energy use. Working with the data also provides a practical example of heat flow and the heat flow equation.

What drove me to the decision to insulate was that I had uninsulated exposed ducts in an unfinished basement. Thus, heat lost through the ducts heated the otherwise unheated basement. The basement has eight foot tall concrete walls and R-19 insulated rim joists. About eighteen inches of the concrete foundation sticks up above ground level. There is about 10 ft2 of double glazed windows in the basement. In the winter, when my furnace was running a few hours a day to keep the upper living area habitable (~67oF), the basement temperature would drop to an average of 55oF. During the summer, the basement would warm to 65o F. Obviously, lots of heat was flowing through the concrete walls during both seasons. That's not a big surprise because the R-value of concrete is reportedly as low as 0.08/inch.

The house, built in 2000, is in Missoula, Montana. The walls are 2" x 6" framed on 16" centers with blown in insulation, probably R-20 or so with sheetrock and sheathing. The cathedral ceiling is R-38, the windows are low-E, double glazed, double hungs. The footprint is about 1,200 ft2. The house has about 1800 ft2 of heated living space in a floor and a half. Thus it is fairly modern, meets modern energy codes, has a 92% efficient natural gas forced air system with a programmable thermostat, and we are cheap with heat (an hour in the morning, six hours in the evening).

I insulated the basement by doubling the insulation at the rim joist to R-38. I then glued 4' x 8' panels of 1.5", R-6 expanded polystyrene (beadboard) to the concrete walls and built a 2" x 4" wall inside that which is further insulated with R-11 fiberglass bats, and then sheet rocked. The cost for materials came to about $0.90 per ft2.

The first figure to the right shows what our weather looks like in terms of heating degree days (HDD). The data are from the NOAA weather site and heating degree days are plotted on a monthly basis for nearly four years.

The second figure shows my monthly natural gas use divided by the appropriate heating degree day value. I only plot the data for the heating season, September to May. From each month's gas use I removed the average of August's gas use so that normal water heating was not included in the results. The vertical axis is in dekatherms of natural gas per HDD.

To date, it looks like insulating the basement walls, in my unheated basement, is saving me about 30%, of natural gas use (not just dollars) on my heating season bills. You can still see, even after insulating the basement, what appears to be an increase in gas use per HDD early in the heating season. That is probably due to the ground surrounding the house cooling down. If the insulation was perfect, you would not expect that slope. However, the insulation is not perfect and the slab is not insulated.

Given the initial results, I decided to thermally isolate the basement by insulating the heating ducts and the basement ceiling with R-13 fiberglass insulation. As soon as I did so, the average basement temperatures dropped about 3oF and showed much less variance during the day. The basement is now hovering around 57oF and I heat it independently, with a small natural gas, stand alone stove when I want it warmer. Next winter's data should show a slight additional savings from insulating the ceiling.

Graph displays the average temperatures of several years
Graph displays that the average gas use is lower after insulation.

An Example

If you care to think about what the heat losses are and how they flow through your walls, here's a start:

BTU/hour = Area (ft2) * (T1 - T2)(oF) / R (hour*oF*ft2/BTU)

If your natural gas supplier bills in therms or dekatherms, there is 100,000 BTU/therm.

As an example, consider a square foot of concrete 6" thick, with an R value of 0.08/inch. If you hold the inside of the concrete at 60oF and the outside at 30oF, the heat loss is:

62.5 BTU/ft2/hour, or 6.25 *10-4 therms/hour.

How many square feet of uninsulated foundation wall do you have? Suppose you have 150 ft2of such wall above ground and your power company gets $0.892/therm:

150 ft2 * 24 Hours * 6.25 *10-4 therms/(ft2*hour) * $0.892/therm = $2.01/day.

Now, glue on some R6 beadboard and the energy loss goes to:

4.6 BTU/hour, thus the R6 in this case yields a 93% reduction and the cost drops to about $0.13/day.