Monday, August 4, 2014

Interaction of Appliances

One of my readers (thanks Andreas) commented about how controlling the appliances could help me.  For example, not letting the compressors of the refrigerator and freezer go on at the same time could lower my peak demand.  I answered that I was a bit leery about that because it could lead to spoiled food.  But, I decided an exercise in looking at it wouldn't take too much time and maybe it would tell me if such a thing would save me money.

For those of you just stumbling upon this site, I have demand billing from my wonderful power company.  There's a long discussion of this here <link>, but suffice it to say that from noon to 7PM I carefully control the usage of power around the house to reduce the power bill.  A single mistake that lasts for 15 minutes during this period could cost me a LOT of money.  Enough to buy several little computers to control things around the house.

So, I looked at adding the usages and creating a new graph to show the total use from my two freezers and refrigerator over a period of time, but had a brainstorm along the way.  I could simply use HighCharts ability to stack graphs.  It worked, and now I have a better understanding of my usage that I want to illustrate for people out there that are considering (or doing) the same kind of thing I do.

First though, a recap of the appliances and what their power usage profile is.  Here's my refrigerator graph as an area chart:

This appliance runs for short bursts to cool down to around 38-40 degrees.  It gets the most activity, but is pretty efficient.  Now, my freezer:

This thing runs a lot.  It has a good compressor that doesn't use much power, but it runs a lot more than I expected.  Still, it doesn't actually chew up much power.  And last, the garage freezer:

This thing runs about half the time during the summer because the garage is hot.  I don't cool the garage and it's on the south side.  

The little spikes in the charts are an artifact of the way I graph the data.  Yes, I could have hunted down the problem and fixed it, but I wanted to study the data, not spend a couple of hours chasing down a bug and fixing it.  So, just ignore the spikes, they only mean I'm lazy.

And finally, the stacked composite graph:

Spikes aside, this shows that my peak usage for all three devices is less than 500 watts.  Sure, it would be more energy efficient to stop one to allow a different one to run, but it wouldn't change my peak usage much.  The way they (the power company) company calculate this is to take a moving average of 15 minutes over the entire peak period for a month and then they bill me for the highest period.  So, any 500 watt period will cost me for the entire month.  I once messed up and used a couple of kW, so I had the freedom to really eat the power for the rest of the month.  Interesting result of peak demand billing, I wonder if they realize it.

At any rate, it doesn't look like I have to worry much about the combination of appliances getting my usage out of control, but it could be useful to keep this in mind if I add an appliance, or need to modify my lifestyle in the future.  Most people don't want to put up with their A/C units being shut off during the hottest part of the day, or their pool motor not running in the afternoon or early evening when they want to use it.  Fortunately, when relatives visit me, they understand my OCD about keeping appliances shut off, and just giggle about me when they get home, so I don't have to worry about them.

See folk, I actually take your suggestions seriously and even act on them from time to time.  It may take months, but I eventually get there.  Now I still have to think about getting a temperature sensor inside the freezer (thanks badhairday).


  1. It's been awhile since I've read any updates on how you watch/manage the air conditioners aside from making sure they don't overlap when running. It would be interesting to see similar data plots of the AC units. We live in similar climates with similar sized houses and equipment, though thankfully for now I don't have demand billing. My interest in the AC units is the age old question... is it cheaper to bring house temperature way down overnight (say to 68) and slowly let the house warm up throughout the day (to 78)... or cheaper to maintain higher average temperature (say 75).

    1. I don't measure the AC units separate from the overall house power. The reason for this is simple, they use so darn much that it's impossible to miss. Another, less important reason is that there isn't an easy to use tool out there to measure power at 220. The stuff all wants to measure 110 and they do a fine job of it, but when we're talking about 6KW, the number of choices gets pretty slim. Sure, I could build one, but I haven't seen the need ... yet.

      But, you can do what you want if you want to start measuring your house power. Just (just ?) build up a power monitor for the house and try your experiments. It should be pretty easy to look at the measurements and tell which way works best.

      For me, I cool the house down and then let it coast through the afternoon without AC at all. Sure, it causes the house temperature to rise up to as much as 85F, but I can handle that better than I can the power bill. At 7PM, they both come on and start cooling. The air feels cool in just a few minutes, but it takes a while to suck the heat out of the tile floors and drywall. Usually by around 10PM, the house has settled down to its new temperature.

      So, my control is off at noon and back on at 7PM. I don't keep them from overlapping, I let them both run. I do control overlapping when I am in recirculate mode to even out the temperature in the house. I don't use this much in the summer, but in the winter it helps to spread the heat created by various appliances to the rest of the house. I use recirculation a lot in the winter.

  2. We have smart meters here and they communicate by Zigbee in a mesh with other houses in the neighborhood. Presumably, one of these homes has a smart meter with a connection to a data sink for reporting back to the utility. I have a receiver in the house, one of these, that lets me know near instantaneous (3 second updates) power usage. I know my AC units pull about 4KW each when running. The same meter reader was used to determine parasitic power usage reported in another post. I'm just too lazy (too many hours at work and other hobbies/interests) to make the pretty plots to see how often they run and I haven't yet compared 24-hour power usage chilling the house down overnight vs maintaining average temperature.

    1. I remember that discussion. If you have a usb port on your device, you can do the experiment pretty easily; well maybe easily. Just capture the data and then stuff it into excel. You should be able to tell which method is the best.

      You'll get the impetus to do this when a huge utility bill comes in the mail. That's what got me started.

  3. How great that you found a good use for my comment without tearing all you freezers and refrigerators apart =)

    The plot quite clearly visualises their individual and combined power usage. In the future when probably most stuff is IoT and possibly reports their power usage, that way of plotting would quickly show what to target first. Even without peak demand billing.

    Keep up the cool work