Five Percent: Conserve Energy

Climate Change Is Important: Energy Conservation is the First Step

February 26, 2009

Heat and Hot Water Energy Usage for My House

Category: Energy Audit,Household,Save Electricity,Save Fuel,Take Actions – Tom Harrison – 5:50 pm

money-from-chimneyI have spent a lot of time and thought on how to save electricity, but not as much on how to save natural gas. I got a lot of information right from the bills, but I used a cool measuring device to get to some important details.

I have a gas furnace and water heater, and also a gas stove. Sure, I can see how much gas I use from the bill. But what do I do with that information (other than pay the bill?)

I wonder how our gas and electrical usage compare? They are both in dollars, but how does that translate to energy? To get that I need to read the bills and convert to a single unit of energy. Following the excellent model of WattzOn … sort of — they measure power, in Watts — how much power you are using now, and at every moment (watts measures power, which has the time factor, or rate built in).

But here I am looking at the energy that I use over some period of time, like a day, or a month or an hour. So I have decided to measure energy. And so we can compare, I can convert to a standard measure: kilowatt-hours (think: 10 old-fashioned 100W light-bulbs, all on for one hour). When you’re talking about things that use energy like water heaters, furnaces, lights, refrigerators, and so on it’s more important to think of how much you use them in a given day (or week, or moth, or year). I’ll pick “day”.

Read Your Gas and Oil Bills

According to our utility bills from the most recent billing cycle:

  • Electricity: 616 kWh in the 33 day billing cycle, or 18.6 kWh/day
  • Gas: 180 therms in the 25 day billing cycle, or 7.2 therms/day, and 1 therm = 29.3 kWh, so 5274 kWh, or 211 kWh/day

Wow! I used more than 11 times more energy in gas than in electricity. (Maybe I should spend more time focusing on that, especially in the winter!). Ok, how about relative price?

  • Electricity: 616 kWh cost a total of $137.69, or $0.22/kWh
  • Gas: 5,274 kWh cost a total of $291.03, or $0.05/kWh

Yikes! Electricity is 4 times more expensive than gas. (Maybe I should convert my dryer to gas!)

But that’s as much detail as there is to be had from my gas bill. I have put a lot of effort into using less hot water, and we have also put effort into making our house tight and energy efficient. But where I have a Kill-a-Watt and PowerCost Monitor that we can use for real-time electrical use readings, I have nothing like that for gas. I don’t know how much my furnace, water heater and stove each use.

There’s No Problem That Cannot Be Solved With Legos

Legos? I’ll bet you thought I was going to say “coat hanger”, or perhaps “swiss army knife”. Nope! Read on.

I got creative and built a little device to record when my gas burner and water heater were on and off. And now I have a little information to look at that is far more instructive than a monthly bill. Here’s a summary of my findings:

During the gas furnace 24 hour test period, the burner was on for three periods, ranging from 1 hour to 2 hours in duration, for a total of 250 minutes (4 hours 10 minutes). It was an average of 30°F during the time I measured. The gas burner is rated to consume 130,000 BTU/hour, and thus consumed about 542,000 BTUs of gas, or 158.7 kWh (cost: about $7.93).

During the water heater 24 hour test period, the heater was on for seven periods, ranging from 10 minutes to 20 minutes in duration, for a total of 75 minutes, or 1 hour 15 minutes. The water heater is rated to consume 32,000 BTU/hour, and thus consumed 40,000 BTUs of gas, or 11.7 kWh (cost: about $0.59).

Kaazam! I used almost 14 times more gas for heat than for hot water. Said another way, more than 90% of my gas is used to heat the house … in winter.

But there’s more. Because I took readings in 5 minute intervals, I can see that my water heater was on 7 times during the day, for 5 to 10 minutes … except for two times during the day when it was on for 20 minutes. It took me a while to figure this out … then I realized that the test was on the day that we have our house cleaned, and they run a few loads of laundry — it seems like we need to post a note telling the cleaners to use the cold water wash cycle :-). So on a typical day, we probably use even less hot water than measured this day.

The furnace came on three times: once in the afternoon for about an hour (first pre-set of programmable thermostat), once again in the evening for about an hour (temp fell below set-point), and once again in the morning for about 2 hours (second pre-set of the thermostat). I see opportunity in this data — it’s pretty clear that in winter, our biggest opportunity for saving likely comes from heating. I think it’s time for a household energy audit.

The power of measurement and awareness never ceases to amaze me.

Details of My Lego Device

lego-nxtA while back, we bought a Lego Mindstorm robotics kit … for my son and daughter, of course. It has a bunch of sensors (distance, touch, light, sound, etc.), some motors, a little computer called the NXT, and a simple visual programming tool. You can make some really cool robots that move on their own and do various neat things. But you can also build a simple data recorder — all I needed was the light sensor and a little program to write a data file. Every 5 minutes, the program records the light level and saves it to a file on the NXT, and runs for 24 hours. For each test (water heater, furnace) I positioned the light sensor in place where it could “see” the flame of furnace or water heater. After 24 hours, I downloaded the file and read it into Excel and made some pretty graphs.

The main caveat in the data is the granularity of the measurements — it is possible that 1) each period during which there was a measurement, the measurement could have started 5 minutes earlier and ended 5 minutes later, and 2) if the flame was lit for less than 5 minutes, there may have been no reading. The former case is likely, and should be adjusted for the probable error; a typical reading is likely to be under-counted. The latter case, based on observation, does not appear to happen; that is both burners tend to turn on for 5 minutes or more. So I factored this error into my measurements.

Naturally, what I really want is a full-time, real-time device that records and displays data for these two units, as well as for our larger electricity users like the dryer. All on one pretty graph. On the web. It can’t cost more than $150. Personally, I think getting the data is the hardest part :-)

Here are some pretty (not!) pictures.

The Data Logger attached to my water heater


Detail of light sensor “looking into” the water heater


Five Minute Interval Graph of Water Heater On/Off Cycles


Five Minute Interval Graph of Gas Furnace On/Off Cycles



  1. […] did the math, and I did the mechanical engineering. And I learned a few important things. During February, I […]

    Pingback by Deconstructing Energy Bills | What's on? — March 6, 2009 @ 2:04 pm

  2. I am in the business of solar thermal (hot water) systems so I found this to be a very interesting way to track energy consumption on water heaters and boilers!

    One question with regard to your boiler. It is rated at 130,000 BTU per hour – but are you sure that when it fires up it always comes at the maximum rate of 130,000 BTU per hour? Some boilers modulate and may not always come at “full”.

    Another way separate water heating (summer natural gas consumption) from space heating/water heating (winter gas consumption) is to subtract August from December gas usage. This assumes that other uses remain more or less constant – which is true to varying amounts from house to house. Have you looked at that?

    Keep up the good work

    Comment by todd hoitsma — March 7, 2009 @ 7:52 pm

  3. Todd — I had a similar question about the BTU output of the furnace (and water heater). Based on visual inspection, it looks pretty dumb to me: on and off, but I have to admit I didn’t watch for a full hour it was on. The light sensor was calibrated pretty well, so might have recorded lower values if the burner stepped down its output. I saw no indication of this in the data. But this certainly blows the whole data collection method as a general solution.

    Actually, this makes me a little curious. I had a conversation with a guy this week who mentioned a special recording thermostat that track length of time the thermostat circuit is closed (burner on) — essentially the same as my measuring device, albeit a lot simpler :-). If there are burners that modulate output, these devices would either need to know about that, or make some estimate of average output.

    The real solution here, of course, is that big appliances like furnaces and water heaters should have a built-in data interface that measures output, efficiency, and other stuff. But I’m a dreamer.

    A second solution (for gas, at least) would be to have an inline metering device with a data interface.

    I know that in larger non-residential buildings, heating and ventilation control systems are more sophisticated and some, (like a company I did a little work for last summer) make rather incredibly precise control and feedback systems that optimize the system in real-time.

    But in the lowly house, we’re a long way away, it seems.

    As for the heat-vs-hot-water split, it is certainly true that in summer, it’s all the hot-water. But I want daily, even hourly data, similar to my PowerCost Monitor for electricity. Again, we’re a long way away.

    Thanks for the feedback — it’s all good stuff.

    Comment by Tom Harrison — March 8, 2009 @ 11:17 am

  4. […] now that the heat’s off, spend some effort to save a little hot water. We pay about $30/month to heat water — if you heat with gas or oil, check your bills in the summer — that’s probably […]

    Pingback by Shower Timer: Six Bucks Once Saves That Every Month | Five Percent: Conserve a Little Energy — June 20, 2009 @ 8:29 pm

  5. “the test was on the day that we have our house cleaned, and they run a few loads of laundry — it seems like we need to post a note telling the cleaners to use the cold water wash cycle”

    ummm…. your cleaners probably expend more energy traveling to your house than you could save by changing these small things.

    Instead of spending hours analyzing your energy use maybe you should just do your own cleaning….

    Comment by Obviate_Now — July 2, 2009 @ 9:40 am

  6. They live less than five miles away and clean several other houses on our street on the same day. And while I take your point, this blog is mainly about becoming aware of all changes we can make, small and large (the “Five Percent” title is intended to capture this idea). Through spending hours analyzing our energy use, we have found ways to save a pretty significant amount of energy — energy we don’t use all of the time.

    Through measuring, I learned lots of things. Indeed, I hadn’t thought about the travel to the house factor, and through your comment, I thought about that :-) — on the other hand, we definitely did think about referring the two women who clean our house to our neighbors because it would be more efficient for everyone.

    Ok, so perhaps I am being a little defensive. This is one of a few luxuries we maintain as my wife and I both work and prefer to have as much time with kids and for ourselves.

    Comment by Tom Harrison — July 3, 2009 @ 6:15 am

  7. Thanks for the info. This will help me quite a bit.

    Comment by Chis Smith — May 24, 2011 @ 3:38 pm

RSS feed for comments on this post.

Sorry, the comment form is closed at this time.