Air Conditioning and power consumption

[Chardok]

So, I've always been told that with home aircon, you should try to keep the temp as consistent as possible because it take more power to reduce the temperature by a large margin than it does to keep it relatively constant - Now, I'm wondering if that is still true with modern AC units. I like my home cool, (74F, 23C) But during the day, I have my thermo programmed to let the temp inside hit 82F (~28C). I'm perfectly okay with setting the daytime temp even higher (it routinely gets to over 100F, 37C)

I mean...not TOO high, 'cause I got two cats. But I can certainly let it slip on up to around 90 while I'm not there. is 90-74 just large a spread to affect any cost savings? What is the typical not-home vs. hom temp I should be shooting for?

[Mr Bean]

Depending on the AC cooling method you have in your house cooling down from 82*F back down to 74*F takes more power than maintaining a constant 74*F.

Cooling after all does not use 74*F air but rather 60*F-70*F (Depending on the air conditioning system) that must be run for longer. Also house dynamics play a big part. Cooling your whole house to 74*F then letting the insulation take care of things is easier if you don't have an active heat source (Like an unventilated but well insulated attic) proving a steady heat source during daylight house.

There are so many individual factors your only real choice is using one method for a month then comparing your electricity bills. My own living situation for example I can cool my entire place down to 66*F with the outside around 90*F and it's cheaper than running my heat at 72*F because my air conditioning is not very flexible and will produce the same amount of heat (98*F) or cooling (62*F Air) and simply runs at that temperature until the place is cooled down to the level I set on the thermos.

[Magis]

Depending on the AC cooling method you have in your house cooling down from 82*F back down to 74*F takes more power than maintaining a constant 74*F.

I really don't see how this makes sense. For starters, power is irrelevant, it's work that's important.

Scenario 1) Over, say, a 24 hour period of time, for an AC to keep a house constantly cool, it has to reject all the heat that leaks into that house over the 24 hours.

Scenario 2) Now, let's say that instead, the house starts cool, then warms up, then the AC has to cool it back down - all again taking place over 24 hours. The AC still only has to reject all the heat that's leaked into the house over a 24 hour period, in order to bring the house down to its original, cool temperature.

The only difference is that the rate of heat transfer increases with temperature gradients. So, if a cool house is in a warm environment, more heat leaks in during a certain amount of time than if the house were warmer. As a result, in Scenario 1, more heat would have leaked into the house, and hence the AC would have to reject more total heat. In Scenario 2, less heat leaked in because the house was more similar to the environmental temperature, hence the AC has to reject less total heat.

If we were to take the "it takes more power to reduce the temp than keeping it constant" theory to an extreme, it would stand to reason that keeping a house at 74F for a thousand years would take less energy than keeping at 82F for 999.999 years and cooling it down to 74F for 9 hours, which is clearly unreasonable.

[Mr Bean]

That only makes sense Magis if outside temperature is constant that entire 999.999 years and ignore the thermocline barrier phenomenon or in other words with a sharp enough temperature difference you can provide your own natural insulation.

[Magis]

That only makes sense Magis if outside temperature is constant that entire 999.999 years

Yes, it was an extreme example. According to the OP, though, the AC would be off during the warmest part of the day, which makes my argument more applicable, not less.

and ignore the thermocline barrier phenomenon or in other words with a sharp enough temperature difference you can provide your own natural insulation.

I'm not sure what you mean by this, but it sounds sketchy.

EDIT: According to the US Department of Energy, turning up the temperature on a home AC for an 8 hour period, will result in approximately a 1% reduction of electricity per degree change (website here).

[sciguy]

That only makes sense Magis if outside temperature is constant...

His underlying reasoning still holds up, though. And heat will leak into your house faster when the temperature difference between inside and outside is larger. I can't see any plausible scenario where it would be more efficient to keep the house constantly cool than to let it warm up and then cooling it back down, unless your AC has some sort of insanely inefficient startup sequence that wastes energy every time it turns on.

...and ignore the thermocline barrier phenomenon or in other words with a sharp enough temperature difference you can provide your own natural insulation.

A quick google provided no links talking about thermocline effects affecting house insulation. Is there actually something to this, or is it just some bullshit that you made up?

[Purple]

Here is a question thou. Is the difference in power consumption and by connection electric bills really that large to justify the bother?

[Akhlut]

His underlying reasoning still holds up, though. And heat will leak into your house faster when the temperature difference between inside and outside is larger. I can't see any plausible scenario where it would be more efficient to keep the house constantly cool than to let it warm up and then cooling it back down, unless your AC has some sort of insanely inefficient startup sequence that wastes energy every time it turns on.

It depends on a lot of factors. A very well-insulated house that is partially sunken into the earth in such a manner as to minimize sun exposure will probably not be as inefficient to keep the house constantly cool. This might hold true for basement apartments as well.

I'd say for most houses, though, it's better to change temperatures between day and night, in most instances.

Here is a question thou. Is the difference in power consumption and by connection electric bills really that large to justify the bother?

If you're just renting a house for a year in a location where you only need A/C for 3 months of the year, probably not. If you live somewhere where A/C is needed for 6+ months of the year and you're going to live there indefinitely, then, yes, it will matter. If you can save $20/month, on average (assuming fairly static energy prices), then that's $1,200 in 10 years.

[sciguy]

It depends on a lot of factors. A very well-insulated house that is partially sunken into the earth in such a manner as to minimize sun exposure will probably not be as inefficient to keep the house constantly cool. This might hold true for basement apartments as well.

Of course it will require less energy to keep a well-insulated house cool than a poorly-insulated house. But the house's insulation doesn't really matter for the purposes of determining whether it's more efficient to leave the AC on all day vs. turning it off and allowing the house to warm for a while before cooling the house back down.

Start with a house at temperature T. Assuming it's warmer outside than in, throughout the day heat will leak into your house, raising the temperature. In order to get back to T, you will have to remove all of that heat that leaked in. So, which house will have more heat leak in throughout the course of the day - the one that was kept at T all day, having the heat contantly pumped out as soon as it came in, or the one that was allowed to warm for a while? The answer is that the house that was allowed to warm will have less heat leak into it, because the rate at which heat enters the house from outside will decrease as the house warms. This is true regardless of how well-insuated the house is.