If you are on-board with the sentiment that we should strive to reduce the amount of energy we consume as a means to relieve pressure on a world suffering impending energy scarcity, then you probably want to know how one might proceed. In this post, I will describe the single-biggest energy-saving strategy I have employed in my home in the past five years, which slashed my natural gas consumption by almost a factor of five.
Last week, I described how to read gas meters, in the process discovering how onerous pilots lights can be. As a result of initial exploration of my energy footprint in the spring of 2007, I shut off the furnace pilot light for the summer, which I figured accounted for two-thirds of my warm-season natural gas use. When winter came, my wife and I challenged ourselves to hold off on re-igniting the pilot light until it got too cold for us to bear. That day never came. The result was a dramatic reduction in natural gas use.
In this post, I will talk about some of the ups and downs of adjusting to a colder house in the winter. Granted, we live in moderate San Diego, and could not get away with the same tactic in many locales. Even so, I will quantify the gains one might expect elsewhere for similar living conditions.
Below is a plot of natural gas usage for me and my wife since early 2006. It was at this time that we sold our home and moved into a large, 1960′s condominium that used natural gas only for heating and hot water. The data are straight from our utility bill, with a point plotted for each month (months shown as alternating gray/white stripes).
The first month can be disregarded, since we were not occupying the space yet and have no idea why the gas use was so high in March. Also shown (in blue) is the typical San Diego gas usage for a similar vintage/size dwelling that houses one or two people—as provided in the form of an online comparison tool by my utility (SDG&E).
In the year spanning April 2006 through March 2007, we used 307 Therms of gas. To put this into familiar metric units for the plot, I have multiplied by 29.3 kWh/Therm and presented the result as daily energy use—taking into account the variable number of days in each billing cycle’s “month.” In total, we used about 9000 kWh of thermal energy from natural gas over the course of the year. If comparing this to electricity usage, keep in mind that most electricity is generated from a heat engine getting approximately 33% efficiency at converting thermal energy into electricity. So the equivalent amount of delivered electricity requiring a comparable fossil fuel outlay at the power plant is about 3000 kWh. But if this electricity is used for actual heat, we would still require 9000 kWh consumption in the house no matter what the form.
What we see from the plot is that our annual usage was pretty consistent with that of the typical San Diegan in the first year: SDG&E claims 300 Therms is typical for condos and 383 Therms is typical of detached houses of comparable vintage/size/occupancy.
But then something dramatic happened. I looked at my gas meter, and took stock of our monthly history. At first, I tried to reconcile the summertime use of 15 Therms each month, or 0.5 Therms/day—equating to about 15 kWh/day. As explained in the pilot light post, an average of one shower a day plus other uses led to an estimated 60 l of hot water use per day, heated 20°C over (summer) ambient temperature, requiring only 1.4 kWh of thermal energy. How can I be a factor of ten off? So looking at utility bills with just a bit of physics is enough to expose problems—in this case the pilot lights.
Notice what happened after the dawning of my energy awareness. In the last twelve months, we used 66 Therms, compared to over 300 in the year before becoming energy-aware. Admittedly, April and May of 2011 were lower than normal because of a sabbatical quarter spent in Seattle, leaving the house unoccupied (but the hot water pilot left on). We more than made up for this anomaly, however, by an experiment I ran in December with an eye toward this post. More on that later.
Also worth noting is that we moved from the condo to a detached house in late 2009, but the gas use stayed low. Moreover, we see a dramatic change when a house-sitter stayed in our place in the fall of 2010. I don’t think the house-sitter used gas heat, but in any case used far more gas than we normally would in the house (hot water, presumably). We also turned up the thermostat for a visitor who stayed with us for a week in the condo in early 2009.
The lessons we derive from this are that:
Okay, clever trick. Turn off the heat. Brilliant. But then it gets cold, genius.
I won’t deny that an unheated house is colder than a heated house. But before I get into specifics, let me remind everyone that humans are evolved animals who coped with seasons long before HVAC (heating, ventilation, air conditioning) came along. Are we really such colossal wimps now that we can’t tolerate living very far away from “room temperature?” We actually have ways to adapt to cold and heat—within limits.
We found that our house (and condo) tended to get down to about 12°C (54°F) on the coldest nights with no heat. Neither place is super-well insulated, but I would not call them dreadfully drafty either.
Below is a plot of temperature within our house over the course of 8 days in late December, 2011. Except for the “experiment” on Christmas (day 359–360), the house went along its usual, passive undulation. The average inside (blue) temperature—ignoring the heating period—was 16°C (61°F), reaching a low of 12°C (54°F).
Meanwhile, the outside temperature averaged about 11°C, making for a 5°C (9°F) average difference. Why would the average internal temperature of the house be warmer than the average outdoor temperature? It’s not because of the appliances within, although this contributes some small amount. Mainly, it’s solar heating through windows and rooftop. Our house is not designed to be a passive solar collector, but stick a house in the sun and some of that heat is bound to end up inside.
On the morning of December 25, I turned the thermostat in our house up to a toasty 20°C (68°F). Room temperature! I monitored gas consumption needed both to charge up the house to this temperature, as well as to maintain it for the next 24 hours. The indoor temperature sensor, near the ceiling in the hall, displays spikes when the heat is on and blowing hot air into the house. This appears as “hair” on the blue curve in the plot above. Below is a close-up.
Tracking the gas expenditure, it took 1.3 hcf (hundred cubic feet; 1.33 Therms; 38.9 kWh) over the course of two hours to get the house up to the set-point. Most informative is the following overnight period, when the house was fully stabilized at the set-point. The overnight period required an additional 3.4 hcf (3.5 Therm; 102 kWh). If we kept this up for a month of similar weather, we should expect to expend something like 100 Therms of gas energy in a month, compared to our average of 5. Thus our winter-month gas expenditure would be 20 times larger than our realized average rate! Instead of paying $7 per month for gas, we would get hit with $120 charges!
Overall, the one-day experiment used about 5 Therms of energy, commensurate with our usual monthly expenditure: thus the extra spike on the natural gas history plot at the end of 2011. The daily expenditure of 102 kWh to heat our home translates to an average power of 4,250 W. I react with horror. We try to keep our electricity average power below 200 W, for instance.
First of all, I should express my gratitude that my wife is on board with living a cooler winter existence. It takes two to make such cuts a reality. Living in an unheated house was a bit of a difficult adjustment the first year, but now we hardly notice. It’s supposed to be a bit cold in winter. We wear appropriate layers, have down-boot slippers, and have an electric mattress pad on the bed.
Let me point out that when it comes to staying warm in a house, I really could care less how warm the pictures on the wall are, or the kitchen knives, or the books in the bookshelf, etc. I care how warm I am. And that’s a much easier task. So concentrate on warming the person, not the house, and suddenly you’ll find that not much energy is needed to stay warm.
The mattress pad on the bed is a key feature. It has dual controls, five steps each, amounting to about 6 W of additional power per click. Full-blast, it’s only about 60 W. Compared to a 1500 W space heater, this is impressively small, and more power than is needed to stay comfortable all night. We turn it on a half-hour before we get in bed. It knocks off the chill, eliminating the unpleasantness of crawling between cold sheets. I usually cut mine completely off at this point, relying on the down comforter plus my 100 W metabolism to do the rest of the work. Another key feature is a higher concentration of coils at the feet, where more warmth is appreciated.
Blankets on the couch make curling up for a movie or chatting with friends perfectly comfortable. My wife often sets a 40 W heating pad on her lap or under her feet when extra warmth is needed. On occasion, there might be a little space-heater action in the closet while getting dressed, or around the dining table when friends visit. For short times in small spaces, this can be reasonably effective without being too energy-egregious.
And in the end, the motivation as to why we go without heat is itself sustaining: we simultaneously have a smaller impact on the world—living more within our collective means, and we are conditioning ourselves to be tougher so that we will more easily adapt to potentially harsher situations in the future. We also remind ourselves that humans have not always been pampered with luxuriant heat during the winter, and we got by as a species just fine. Less princess; more villager.
Despite the accomplishment of trimming our gas usage by a factor of five, I’m sure many readers are wholly unimpressed. “Any fool can turn off the heat in a San Diego winter—if you can even call it winter. Where I live, oh-ho! This hair-brained strategy has no chance.”
Several quick points, then some details.
Let’s spend some time on this last claim. Roughly speaking, the energy expended to heat a house is directly proportional to the temperature difference between inside and outside. More accurately, it is proportional to the difference between the set-point temperature and the temperature that would be achieved passively in the house. Windows admitting sunlight (plus roof absorption) generally mean the average passive temperature inside is higher than the average outside temperature, as is seen in my home.
So in a colder climate, let’s say it’s cold enough that the heat might come on at any time of day (e.g., it does not warm up enough outside to obviate the need). Let’s say the average daily temperature is freezing: 0°C (32°F). Let’s further assume that the passive temperature in the house would average about 3°C (5°F) above the ambient outside temperature. Then to keep the temperature set-point at 12°C (54°F) requires an average offset of 9°C, whereas maintaining the standard room temperature of 21°C (70°F) requires twice the offset: 18°C. This is the basis for my statement that a factor of two savings may be gained even in colder environments if you’re willing to have your house as cold as we let ours get in San Diego. Granted, it would stay this temperature all the time during cold periods. But I’ll wager it can seem normal, and will still be a comfort compared to the frigid world outside.
To get more specific, the National Oceanographic and Atmospheric Administration (NOAA) keeps data on the amount of cooling and heating required based on the weather at a given location. The metric is called heating degree-days, and is effectively a sum of average departures from 65°F (18°C). A temperature slightly cooler than normal “room temperature” is picked to allow for the passive offset of a normal home. Normal values for cities throughout the U.S. on a monthly basis can be found here.
As an example, heating degree-day values for select cities are shown below, in both °C (°F) formats. Both the annual sum and the value for January—typically the coldest month—are shown.
|City, State||January||Annual||Jan. at 12°C|
|Miami, FL||32 (58)||86 (155)||0|
|San Diego, CA||126 (227)||591 (1063)||0|
|Atlanta, GA||384 (692)||1571 (2827)||0.27|
|Washington, D.C.||503 (903)||2222 (3999)||0.45|
|St. Louis, MO||609 (1097)||2643 (4757)||0.54|
|Seattle, WA||415 (747)||2665 (4797)||0.33|
|Boston, MA||613 (1104)||3128 (5630)||0.54|
|Minneapolis, MN||898 (1616)||4379 (7882)||0.69|
|Anchorage, AK||848 (1526)||5817 (10470)||0.67|
Since January has 31 days, we can divide each of the January numbers by 31 to get the average offset from 18°C (65°F). For instance, San Diego’s average January temperature computes to 14°C (58°F)—a bit warmer than the period displayed on the 8-day temperature record above.
In Atlanta, the average January temperature lands at 12°C cooler than the 18°C reference point, putting it at about 6°C (42°F). Maintaining an indoor temperature of 12°C therefore costs one quarter of the energy as maintaining it at 21°C—again relying on the standard 3°C bonus from passive solar heating. To see this more clearly, the passive house would average 9°C when it’s 6°C on average outside. To bring the average to 12°C requires a ?T of only 3°C compared to a ?T of 12°C needed to achieve 21°C inside. Thus the one-quarter expenditure.
Before objecting that clouds nullify any passive gain: not true. Diminished, yes—but not gone, actually. Moreover, clouds keep your house from losing as much heat—especially overnight—by blocking radiation to space. In any case, the 3°C offset implicit in the NOAA data is there for a reason.
In the table, I have also included the fraction of heating energy that would be needed in January if thermostats are pulled down from 21°C (70°F) to 12°C (54°F). In essence, this reduction frees up 279 degree days (31 days times 9°C) for anyone who still needs to run heat to maintain a 12°C house in winter. Even in wickedly cold locations, a substantial savings may be affected. And this is in the worst month. If I do a similar calculation for Minneapolis across its coldest 7 months, I find a 50% reduction in total number of degree-days to maintain 12°C instead of 21°C.
How much heating energy does a monthly 279 degree-day savings translate into? One handle I have is that my house used 3.5 Therms to stay toasty on a night that totaled 6.9 degree-days below the 18°C reference point (possible to assess from overnight dip in close-up plot of day 360 above). So my house uses about 0.5 Therms per (Celsius) degree-day offset. Applying this scaling to 280 degree-days means a potential savings of 140 Therms, or about $170 at the rate I pay for gas service. A smaller, better-insulated house will require less thermal energy per degree-day. Scouting a bit online, I find that my house is not great (no surprise that a San Diego house would not be built to good thermal standards). In colder climates, houses tend to be built with heating efficiencies around 0.2–0.4 Therms per (Celsius) degree day. Divide by 1.8 to get the equivalent Fahrenheit measure. So for a more reasonably insulated house, shaving 280 degree-days by adjusting to 12°C instead of 21°C should save about 70 Therms, or close to $100 each month. A bit of that might get eaten up in blankets, down slippers, and mattress pads. But these things will last years.
For assessing the heat-tightness of a house, we have several units from which to choose. The Therms per degree day measure may be somewhat convenient in the U.S., but a more natural unit is W/°C. One Therm per day equates to a power of 1.22 kW. Here is a useful table for comparison for U.S. houses in the northeast, midwest, and west with a square-footage of 2000 ft² (186 m²). Data are from this useful website.
|Characteristic||Therms/deg-day (°F)||Therms/deg-day (°C)||W/°C|
One caveat worth pointing out: the amount of savings calculated in the foregoing paragraphs is based only on average temperatures. This will work perfectly well for the months when the outdoor temperature stays well below the set-point (by at least 3°C) all day. In milder weather the daily swings may still force the heat to come on in the wee hours, even if the calculation based on averages suggests that heat will not be needed. A well-insulated house will suffer smaller swings and may be able to resist kicking the heat on during the nightly cycle.
We are not utterly victim to our energy demands: we create those demands. Adopting an attitude that houses do not need to be warm as long as the person can remain warm means that we can have a substantial impact on our energy use. By adopting a winter-time set-point temperature that would be thought of as a mild spring day outside, we could slash our national heating demand by a factor of two at least! Further reductions could come in the form of smaller houses (or heated areas within houses), better insulation, and deliberate enhancement of passive solar heat inputs.
So. Are you ready to be a low-heat trooper, donning layers and blankets, or do you insist on shorts and tee-shirts inside in the winter? Perhaps our ancient ancestors would have opted for the same, given a choice. But they survived their winters, and we’re made of the same stern stuff.
In this and other posts that I write about energy reduction schemes, please note that I am not trying to establish my own efforts as superior (my location plays a big role in my success). Lots of folks have been far more effective than I have been in reducing energy usage—and they have my admiration. For me, reduction by factors of 2, 3, 4, etc. are satisfying enough to give me a sense of what can be done. I am now more interested in getting more people to catch up with me than I am in pushing to further extremes. Also, Do the Math readers may wonder why I fuss over home efforts, when many of these tactics do not directly address the impending liquid fuels shortage—identified as our biggest near-term hardship. My answer is that I have adopted a philosophy that all energy is precious and that its consumption (together with consumerism) should be reduced as substantially as we can tolerate. At scale, such an approach relieves pressures on many fronts: environmental; transport; agricultural; resource depletion; etc. Let’s not dink around, and instead change the game across the board. We have the power!