Thursday, May 10, 2018

Supercooling a House in the Desert

A while back I ranted a while about power politics in the desert <link>, and after some thought and because I've been doing this for years now, I decided to post a few articles on 'supercooling' a house to lower power bills. Another impetus is that the latest rate increase has caused many people in my area to feel the 'heat' of the new power company rate increase.

First, what the heck is 'supercooling'?

This term came about over time because people (me!) were clobbered by power bills based on the 'demand' billing method. I first posted about this quite a while ago and even dedicated a page to it back then <link>. As I pondered and tested, I created thermostats and temperature gauges, and a lot of other stuff. This blog was actually the result of experiments and actual devices that have been copied by people all over the world to mimic some portion of my efforts.

One of the methods I tried was to cool the house way down during the off peak hours and then turn the AC off entirely during the peak period, and let the house coast through the peak period. That made my power bill the envy of all the neighbors. They weren't willing to do the same thing, but they commented on my work many times.

So, supercooling is cooling the house down a bunch before the peak period begins, turning the AC completely off during the peak and then turn it back on after the peak period ends. This is all the result of demand billing explained on the link above <convenience link>. For folk that live in the cold country, this will work the opposite way as well. You can heat it way up and then let it coast. However, I don't have any direct experience with that living here.

My early efforts resulted in great savings under what I would have been paying. So much so that solar power companies can't match me. I had inadvertently beaten the ability of solar power to save money. Here is a chart of my power usage back then:
I know, ancient charting software. But forgive me, this was in 2009 and I was new to this blogging thing.

It was very important to me that I kept the power usage to the absolute minimum during the peak period, and in those days it was from noon to 7 PM. I actually used more power this way, but paid a much smaller bill. The huge spike in power just before the peak period is me supercooling the house before I shut off the AC units entirely from noon to seven.

Basically it worked ! I was a happy puppy and decided to automate the house and take control of my power usage so I couldn't get screwed by an accident again. Thus, the adventure of the the last years began.

So, that's supercooling and my adventure into it.

Now it's time to study it some more because people want to know what to do to accomplish some of my results. Not everyone wants to tear out walls, climb up on the roof or bury wire to the septic tank. They just want to understand how to save some money.

That's quite enough background for now. Let's get to the first experiment I did.

Before the new rate increases came into effect I modified my thermostat code <link> so that the peak period was from noon until 8PM; this is a union of the new peak period 3PM to 8PM and the old, 12 noon to 7PM.  I did this so that whenever they actually implemented the rate change I wouldn't be caught by surprise and get an outrageous bill. It worked, the transition was smooth and I didn't get any bad surprises. Since it was for a very long period, it was perfect to test the heat rise inside the house; I just had to wait for a hot day.

Naturally it came with a vengence, and I was already gathering data, so first I did one set of rooms that are wide open to each other no restrictions on the air flow and are air conditioned (duh). For the test I let them cut off automatically at NOON, because I wanted enough time for good measurements. I kicked them back on at 8PM .

Here's a graph showing the set of rooms and the outside temperature. You can tell which is which. At noon, the room temperature was 76F and at 8PM the room temperature had risen to 81F for an increase of 5 degrees.  I haven't done it for only the period of 3PM to 8PM yet, because I wanted to understand the rise in temperature so I would know the settings to try first. Here's the graph:

I was surprised how little the temperature changed without AC when the outside temperature varied up to 105F. Of course your mileage will vary due to every factor known to man, things like: did you open the doors, curtains open, insulation level, ceiling fans, etc. Most everything inside and outside your house will have an impact. However, for me it means I can lower the temperature by five degrees and survive until 8PM from 3PM when the temperature is in the low hundreds. I'm sure that I'll have to drop another degree or two when it gets in the teens, but I have a place to start.

I did notice that the concrete floor was absorbing heat. It seems the sun shining on the foundation outside was travelling UNDER the insulated walls and radiating into the house. The walls in that area were cool to the touch, but that's subjective. I'll have to go around the house and check the temperature on the exposed areas to see what the floor temperature is; it may just be over-sensitive bare feet. Good excuse to buy an infra-red thermometer.

What? you want to see the power usage during the same period to see how things are going power-wise?

Notice that from noon until 8PM I use almost no power. The little rise towards the end of the period is my TV and the hump around 3:50 is me warming up some stuff in the microwave. I didn't supercool the house for this test, but I did let everything kick on normally at 8PM. That's the huge spike you see on the right.

Yes, I watch the power that closely, at least during peak periods.

One of the fallacies I see mentioned a lot is staggering appliances. This will help during the peak period by keeping the dryer from adding to the AC or something similar. That keeps the demand number lower during ONLY this period. The rest of the time (off peak) it doesn't matter at all. During those periods power is power, so go ahead, run the dryer. During peak, keep that darn thing turned off. You don't need that shirt right now, wear the least dirty one from the laundry instead.

I'll be doing more observations and experiments over the next weeks, months, whatever, and I'll try to keep links between the posts so you can follow along (if you actually care about this kind of stuff) without hearing about hydraulics, septic sensors, old tractors, wifi controls, etc.

This could get interesting and a good place for me to look when I forget why I started this mess.

Oh, this page will attract spam like crazy. Every solar service in the world will be posting something with a hidden link in their name or some word inside. There will be arguments aplenty about the costs of solar vs nothing. Off grid suggestions will probably abound. I know how to deal with this ... I delete it. So, solar sales people, don't bother, you'll just get deleted. Same for people that want to compare their solar savings, gone. Also, at some point, I'll actually have to turn off comments because I get tired of messing with the constant spam. And, I love suggestions for things to look at (if they don't cost me much).

But for now, comment away.

Previous post in this series <link> Next post <link>


  1. It's interesting for me as a European how different countries 'manage' us to limit the number of power stations needed. In the UK a 22kW maximum demand supply is normal and you are limited by your wallet; France, Spain and Germany I think, charge you both a higher unit rate and a higher standing charge; in France a 9kW supply is normal and costs ~$100 pa, with a kWh costing ~$0.13. It rarely hits 95F in the south of France and air conditioning is rarely seen.
    You must have a good amount of insulation; a 5F rise inside with the outside at over 100F is impressive. I'll get a rise of say 15F at 8pm from a starting temperature of 16F at 8AM when the outside hits 95F. I've got >12" of insulation in the roof and 8" in the walls and the slab has ~4". Keeping the shutters down keeps the heat outside, but I'll be checking the slab temperature - good tip. And it was the excuse I needed to buy a infra-red thermometer.
    How does the peak demand work with you. Are you charged a huge premium for any units 3-8pm, or just once you hit a threshold?
    Good post as usual.

    1. Peak demand is measured by the hour, starting on the hour. So there are (currently) five periods they measure 3-4, 4-5, 5-6, 6-7 and 7-8. So, if the oven is on over the hour change, you can split the demand measurement between hours. The difficulty with doing that is they have a 10% allowance on both sided of the hour; 6 minutes before the hour and 6 minutes after. How they can get away with that in the 21st century is beyond me. It's easy to get the time to a second accuracy these days, but they are allowed to be minutes off....

      They keep the highest demand number stored and that's what they charge me for. Actually, I should do a recap of how people are actually charged the demand number. That's another place where rumors abound about what to do about it.


  2. One important thing I haven't seen you mention is thermal mass. Very important with super cooling .

    1. Strangely, thermal mass doesn't seem to count much if you don't let the interior of the house get out of hand and keep the air moving. I think it's because the houses here have low mass. They sit on a concrete slab, but it's covered from the sun; the rest is wood and stucco.

      Now, if our houses were made from brick, adobe, concrete thermal mass would be a huge concern.