Which is better maintaining a cool temp or cooling down a hot house

June 30, 2013 10:12 AM

Which would take less energy - running the AC throughout the day to maintain say 75F (outside 95-100) or letting the house get to 90F+ throughout the day and then cooling it down at night back to 75F?
posted by zeoslap to Science & Nature (27 answers total)

This is one of those questions where I'm sure there's a technically correct answer that absolutely misses the point.

The real real answer is to run the AC all day in a house that's properly insulated. Everything else is a secondary consideration.

And while you're at it, throw up some external shades to block the sun from hitting the west-facing windows. You can get some nice fast-growing trees for twenty bucks at Home Depot and not have this problem next year.
posted by Cool Papa Bell at 10:31 AM on June 30, 2013

I'm pretty sure it's the latter, or else why would programmable thermostats have been invented that do exactly that?

There are a number of other things that help use less energy too:
1. Use blinds and shades appropriately to keep the light from entering and warming the room.
2. Use ceiling or tabletop fans to keep air circulating.
3. Keep your air conditioner well-maintained and your air ducts clean.
4. Make sure your house is properly insulated.
5. Turn the AC down or off at night as well as during the day while you're gone.
posted by peanut_mcgillicuty at 10:35 AM on June 30, 2013

I compromise and turn the temp up to 82 when I'm out and back to 78 when I come home.

I think I'm saving energy, but I'll look forward to hearing more. . .
posted by abirdinthehand at 10:51 AM on June 30, 2013

Okay, here's an attempt at a physics answers: your house is going to heat up faster the higher the temperature difference between the inside and the outside. So, keeping the house cool means that you have to work harder.

If you let the house heat during the day, you have to make up all of the lost ground at night. However, there's less total work, since eventually your house reaches the ambient temperature and stops heating up at all (savings!)

Let me know if this explanation is wrong! I don't know much physics!
posted by zscore at 10:54 AM on June 30, 2013

We've had a few good answers on this in the past - here's one previous question about turning down AC when out of the house... And a couple of similar questions about heating.
posted by LobsterMitten at 10:54 AM on June 30, 2013

zscore, I agree (mechanical engineer here, FWIW). But there's a major confounding factor here: A/C units are more efficient when it's colder outside. So when you try to recover 10 or 15 degrees at 5pm, you've got your A/C running less efficiently than if you did that same cooling at 10am.

I don't know how much this offsets. I'm sure it depends highly on the weather and how well insulated the house is. It would be great to use a logging thermostat (does a Nest do this?) or other method to check the duty cycle for one method vs. the other on similar days.
posted by ftm at 11:01 AM on June 30, 2013

The second one. Take this to its logical conclusion: assume your want to use your vacation home for two weeks at the end of August. What takes more energy, leaving the A/C on all year, or cooling down the house when you arrive in August?

There are other factors to consider as well, like the fact that cooling at night is cheaper (if it cools off enough outside you can do it with just a fan instead of A/C), but generally, keeping your house cool,for a shorter time requires less energy than keeping it cool for a longer time.
posted by tylerkaraszewski at 11:07 AM on June 30, 2013

Depends solely on insulation. If you were to have effectively zero insulation, turning the A/C on only when you were home would use less energy. If you were to have perfect insulation, leaving the A/C on (though, I guess you wouldn't even have to) would use less energy.
posted by wrok at 11:27 AM on June 30, 2013

I remember watching a show about being green in the home and conserving energy. They did an experiment that showed it was more efficient to raise and lower the temperature throughout the day as needed, rather than keeping it constant.
posted by catatethebird at 11:44 AM on June 30, 2013

(This was in a well-insulated home, and would be even more true in a poorly insulated one.)
posted by catatethebird at 11:46 AM on June 30, 2013

I vote second, with a mix. So, here's what I do.

I get hot at night so we run the A/C all night - keeping it around 73 degrees on the thermostat so it cools to about 70-68 in the apartment. It of course seems cooler than that since cold air is blowing on us. We close our home office door so that we save energy, otherwise that room is a total A/C sink.

Then, most of the time during most of the day (I work from home) it's still very cool in the apartment. Keep in mind or apartment seems quite well insulated. Then we don't usually have to turn the A/C on until the evening or later in the day. At that time, we turn it on until it's cool enough then turn it off.

Also, your energy company may charge you based on peak hours. Therefore A/C during the night may be cheaper. It's also cooler outside during the night so it's easier to keep it cool. If it's a bit on the warm side but not too hot then we do fans. Sometimes if it's a little cooler in the evening we open a window with a fan.

This method has been working just fine and it's been between 90 and 100 here. (Looks like high of 102 today!)

-Run at night, close any unnecessary rooms or unnecessary vents.
-Run it periodically during the day if fans don't cut it.
-Make sure your place is well insulated.
posted by Crystalinne at 11:53 AM on June 30, 2013

Just a data point:

I have a 2000 sf house built in 2003. I put in a programmable thermostat that has pretty much had the same setting since day one. For AC, we have the schedule set up so it's 72 when we're home, and 84 when we're not (and overnight, but it would never get there). In the winter, 68 when we're home, 62 when we're not (and overnight). It's one of those that is supposed to 'learn' how long it takes to reach the target temp on a given cycle, but on the very hottest days here in MN (95-100), it sometimes is still trying to get there when we get home from work.

In the 10 years I've been here, I have yet to have an electric bill over $130, and the gas bill has never topped $100. Right or wrong, I've always attributed this to having the programmable thermostat, so my vote is for #2.

Other things we do for energy savings:
- Shades on the windows - the majority of the windows face west
- Close the downstairs air vents in the summer - the basement is cool to begin with, and the cooler upstairs air will sink down there anyway.
- Lights on only in the rooms we happen to be in at the time
posted by SquidLips at 12:37 PM on June 30, 2013

This is a relatively easy question to test in one's particular home, as my favorite nerd Michael Bluejay did. You'd have to tweak the procedure a bit to account for leaving it set at a high temp vs. just turning it off.
posted by muddgirl at 12:40 PM on June 30, 2013

I can only think of one thing that cuts against the general consensus.

Your house contains at least one device which uses more power the hotter it gets, and which makes the house itself hotter as it operates, thereby causing a positive feedback loop in which the more power it uses, the more power it needs: your refrigerator.

Therefore, not only will letting the house heat up during the day cause the refrigerator to draw much more power than it otherwise would, the extra heat which that produces in your house increases the amount of power your AC must use when you finally turn it on.
posted by jamjam at 12:40 PM on June 30, 2013

A refrigerator doesn't take that much energy to run, though. I think modern fridges are roughly the equivalent of a single incandescent bulb.
posted by stopgap at 12:46 PM on June 30, 2013

An important consideration only touched on with the programmable thermostat mention, is how long does it take to reach your target temp if it was "off" all day and the place was allowed to get 100f inside from the sun beating down on it? 30 minutes? hours? because it may be(hypothetical examples here obviously) that if it's left on all the the time the ac pulses on and off 6 times a day, and two times in the afternoon are longer. But if it's turned off and only turned on at night the total amount of time it needs to run is longer than if it had been left on.

This entirely depends on how shitty your insulation is. I'm voting for the solution of "set it to 80something during the day, and 70something when you'd be home" rather than totally turning it off. But mostly because i have bitter memories of perpetually coming home to a hot place that hadn't cooled all the way down yet because of living with a bunch of firm believers in the logic of #2.

So yea, look long and hard at how long it takes to cool to a decent temperature from being totally off. it might be wayyy too long depending on the space.
posted by emptythought at 1:06 PM on June 30, 2013

I think modern fridges are roughly the equivalent of a single incandescent bulb.

You must like things pretty bright around your house, stopgap-- this online energy calculator from Consumer Reports suggests an average of 725 watts for a refrigerator.
posted by jamjam at 1:21 PM on June 30, 2013

The only thing you can really do is precooling, if you have a time-of-use energy plan.
posted by RobotVoodooPower at 1:42 PM on June 30, 2013

jamjam, that calculator is just wrong about refrigerators. Mine uses about 50-60 watts averaged over its duty cycle and about 120-140 watts when the compressor is running.

My AC unit uses about 2000 watts while running, so even if a high temperature made the refrigerator compressor run constantly (it doesn't, but it would take 80 watts more than usual if it did), running the AC for just 20 minutes over the course of 8 hours* would cost more than the extra from the refrigerator. Put another way, running the AC uses far more energy than the extra a refrigerator might use due to a higher temperature, even in an extreme hypothetical.

* AC uses 2000/80 = 25 times the energy that a refrigerator running 100% of the time would use above its usual. Thus, the refrigerator over 8 hours could use (in excess of its usual) as much energy as the AC would in 8hr/25 = 19.2 minutes.
posted by whatnotever at 3:35 PM on June 30, 2013

I have a Nest and it does give you data on this sort of thing. My Nest lets the house heat up during the day and then cools it down a little before people start coming home (or vice versa when the heat is running); the money saved is evident in my lower energy bills. When you get your monthly e-mail report from Nest it tells you WHY your energy use was higher or lower this month -- people were gone more often, or the weather was hotter, or you put a higher set point, or whatever.

Another thing people haven't mentioned -- humidity makes a difference (and the latest Nest programming update accounts for humidity). I wouldn't let my house sit at 90+* during the day because it's almost always book-damagingly humid here when it's that hot; a lot of the comfort A/C provides comes from dehumidifying as well as cooling, so the humidity of your climate may make a difference in how often you run the A/C.

In the past couple years I made two changes for reasons unrelated to insulation/energy savings that had surprisingly huge effects on my energy bill -- I had my chimney permanently capped (with a sheet of stainless steel) to stop small animals falling down it, and it resulted in a HUGE unanticipated energy savings. I also had my doors replaced (they were just old damn doors, and very gappy), and that's produced a smaller but significant savings. I'd say the door replacement was on par with installing the Nest and doing the heat-up-when-nobody's-home thing in terms of energy savings, but the chimney capping eclipsed them both by a LOT.

Energy-saving window treatments (insulating curtains and blinds) on the west side of the house have also made a small but significant difference in energy savings (and a big difference in terms of comfort in the afternoon and early evening). I'd be curious to paint my roof white and see how much effect that had.
posted by Eyebrows McGee at 5:15 PM on June 30, 2013

If you let the house heat during the day, you have to make up all of the lost ground at night. However, there's less total work, since eventually your house reaches the ambient temperature and stops heating up at all (savings!)

Yes, turning it off during the day is cheaper. Perhaps only marginally, but it is.

However, the cost is not just in dollars. What's the point of turning the air conditioning on at all if you are going to sit in a hot house for two hours before it cools off? Especially if your house has a lot of thermal mass, it could take forever for it to cool off. And you will have spent all that money pumping heat out of the house just to sit there for three hours sweatting.

So what you have to do is figure out how much of a temperature drop the air conditioner can do in, say, an hour. Will it drop 10 degrees? Then you set the unoccupied temperature to about 15 degrees warmer than your set point, and have it turn on half an hour before you get home. You'll come home to a house that's not hot, not cold, but cooling down nicely. Just experiment with a good tradeoff.

And yes, closing blinds and curtains makes a huge difference. I also have installed functionally clear thermal blocking window clings that make quite a difference. Maybe not in dollars, but in comfort.
posted by gjc at 5:19 PM on June 30, 2013

stopgap: A refrigerator doesn't take that much energy to run, though. I think modern fridges are roughly the equivalent of a single incandescent bulb.

I thought for sure you were wrong, but you're not: a modern 18 cu-ft fridge consumes ~56W.
posted by IAmBroom at 8:01 PM on June 30, 2013

jamjam: You must like things pretty bright around your house, stopgap-- this online energy calculator from Consumer Reports suggests an average of 725 watts for a refrigerator.

The 725W number appears to originate with the Dept of Energy, and it's a one-size-fits-all, outdated number that is much-quoted on the interwebs.
posted by IAmBroom at 8:06 PM on June 30, 2013

Leaving aside all the perfectly sound thermodynamics explanations whereby cooling when you get home is most definitely more efficient, your house breathes. A lot. Like, for a newer house .3 Air Changes per Hour (ACH), for an older house potentially much more (I think mine registered at 1 ACH per hour). So you're not just keeping the air in the house cool, you're constantly cooling air that has just been sucked in from outside.

DOE has an entire program devoted to Buildings and Building Technologies that sits in its Energy Efficiency org (DOE EERE), they sponsor some really interesting research that is worth googling if you're curious to know more.
posted by genug at 8:21 PM on June 30, 2013

My A/C guy, Tommy, says to keep the A/C ticking over all day at a reasonable temp, and then maybe run it up a bit when you get home if you want -- but definitely don't turn it off in the morning and then try to cool down your house at the end of the afternoon.

The equipment is designed to run for long periods of time. But when the house isn't cooled, the attic becomes blazing hot -- and all your cooled air will get heated up as it flows through the ductwork to the vents. (The foam jacket around the ducts is like R4 or, more recently, R8 -- definitely susceptible to getting heated up.)

My wife hates to hear this, because she believes in her heart that leaving the chillers & air-handlers off all day must be cheaper, but Tommy says nuh-uh and to leave them going at a moderate temperature.
posted by wenestvedt at 7:18 AM on July 1, 2013

In most houses with AC, the ducting areas aren't cooled in the first place (that's why the ducts are insulated), so even if your AC is running all day, any uninsulated space like the attic is going to become blazing hot (I know mine does! I've been up there at 5pm on a Texas summer day). Having your house at 90 degrees isn't going to heat up an uninsulated space any more than having your house at 75 degrees, unless you have AC registers blowing into that space, which would be sort of pointless.

and all your cooled air will get heated up as it flows through the ductwork to the vents.

I really do think this difference is going to be negligible. Even with R4 insulation the metal will cool down quickly enough that it won't matter for your total efficiency.

Again, I urge anyone who cares about their energy bill to try this out in their own home. It's not that hard to test if you can find two more-or-less equal-temperature days.
posted by muddgirl at 7:51 AM on July 1, 2013

Gotanda: Also, do you have anything in your house that will suffer from rising to 95+ on a daily basis?

... such as a computer left on. I've had a computer slowly degrade from overheating.
posted by IAmBroom at 10:01 AM on July 1, 2013