Saving Money With Pre-Cooling

[ano]

Water has one of the highest specific heats of any substance out there. Most of the stuff in your house is going to have a specific heat that is a small fraction of what water's is.
 
Fiberglass quilt for example has a C of 840J/kgK.  Water is 4,200.  Also consider that insulation weighs almost nothing (12kg/cubic meter vs 1000kg for water) and you realize that it contributes very little to the thermal mass of a house.

Its just an estimate, the the figure I give is JUST for the CONTENTS of a house. In other words the stuff a mover moves.  I haven't seen too many movers move insulation or a slab.  As for the weight of the actual house structure, that is closer to 50 or 60 pounds per square foot from what I read.  You of course would need to subtract insulation, and everything beyond the insulation, such as your roof, a slab, etc.  Add all that and your more likely would have approx.  130,000 pounds.  Adjust back for your specific heat adjustment, and it STILL comes out to considerably longer than 8 hours to cool a house 12 degrees.  And I an not counting the loss that occurs through the insulation at the same time. 
 
Again, its an estimate and I don't plan on weighing everything in my house and calculating its specific heat to get more accurate. Its very easy to test for. If I cool my home for longer, and it stays cool longer, then I know I'm not cooling it enough. Its that easy. If it doesn't stay cool any longer, then I have cooled it long enough and reached the peak. Its easier to work backwards.  I'm just using the weight calculations for the neighsayers who don't think the mass of your house takes that long to cool. In reality, it does.  You heat your oven to 350 degrees (air temp) then you add your hamburger or turkey, and it takes considerable time for it to reach 170 inside.
 
Any of this really isn't new, and if you search the Internet, there are a bunch of studies out there about pre-cooling, only problem is most seem to be for office buildings not homes.  One looked at a 4 hour, 8 hour and 12 hour pre-cooling, and concluded 8 was best, but it definitely depends on construction and exactly what is in your house.  A house may have much more thermal mass than an office building, and more stuff. 

 

[apostolakisl]

I think the thought trail has gotten a little off from what I am really trying to say.
 
You are hitting 70 degree air temp at a steady rate and then flat-lining there.  You really should not be flat-lining.  The temp should continue to drop, just at a slower rate, provided the AC is pushing 45 degree air (I think that is what you said it was).  But from the looks of your graph and the perfect flat 70 degrees for hours on end, it looks like you could run the thing all day and not lose another degree.
 
Assuming 45 degree air is blowing into your house, the only way that your house would fail to cool further is if heat is entering the air.  The rate of heat entering the air from the objects in the house will decrease as the temp of those items decrease.  But it is not.  After a few hours of 70 degree air, most of your stuff in the house should be 70, and there is no reason that 45 degree air shouldn't then push the temp lower.
 
If you were to get one of those IR target thermometers it would be enlightening and very easy to use.  They don't cost that much and are quite accurate at measuring temp on most objects (they don't work well on reflective surfaces like polished metal).  
 
A granite countertop would be a good thing to measure.  Granite has high thermal mass because of its high density.  The temp of the granite will probably be one of the last interior things in your house to drop.  Once it hits 70, odds are most everything in the house is 70.  Interior wall sheetrock also would be an indicator.  Outside walls and slab are a different animal since they are in contact with the outside world.  
 
But I wouldn't stop at the granite countertop, just walk around the measuring the temps of all kinds of things.  It will help you to find your hot spot.  Something is stopping the air from dropping colder than 70.

 

[garbled]

I live in arizona too, and I'm on the same rate plan you are.  I've been pre-cooling my house for awhile now, and it definately saves money.  I can also now confirm it, because I installed a brultech GEM to monitor electricity.  (I precool by about 2-3 degrees, and because of that, the AC doesn't even turn on for 1.5-2 hours into prime time)
 
However, I think the really big money, is in keeping the multiple AC units in my house from running at the same time.  That peak demand charge is really big, and if you can keep it down, it's something like $15 per kWh.  On that end, my plan is to buy radiothermostat CT80's, and program them to run inside timeslices.  Basically, I'll shut the AC off for 10 minutes, and turn the other one on for 10 minutes and let it do it's thing.  Normally, the AC runs in 15 or so minute slices anyhow, the problem is they are not in sync.  I'm hoping if I can control the big loads like the AC's, and the water heater in prime time, I can cut my demand charge down.

 

[ano]

I think the thought trail has gotten a little off from what I am really trying to say.
 
You are hitting 70 degree air temp at a steady rate and then flat-lining there.  You really should not be flat-lining.  The temp should continue to drop, just at a slower rate, provided the AC is pushing 45 degree air (I think that is what you said it was).  But from the looks of your graph and the perfect flat 70 degrees for hours on end, it looks like you could run the thing all day and not lose another degree.

Maybe I didn't communicate what I'm doing clearly.  I do have a device that prevents my air temps from dropping below 70 degrees. That device is called a thermostat.  I do have to live in this house, even when precooling, so 70 is the limit.
 
But in reality its more complex than that. I have a three zone system on one side, so what tends to happen is while the individual thermostats turn zones on and off, if any zones are on, the AC stays on, its just that the dampers open and close.  So temps DO stay at 70, and AC generally stays on, but not always. 
 
I also do use an IR thermometer, a Fieldpiece SIG-1, but remember it measures the SURFACE temperature, and I'm not sure its incredibly accurate.
 

Assuming 45 degree air is blowing into your house, the only way that your house would fail to cool further is if heat is entering the air.  The rate of heat entering the air from the objects in the house will decrease as the temp of those items decrease.  But it is not.  After a few hours of 70 degree air, most of your stuff in the house should be 70, and there is no reason that 45 degree air shouldn't then push the temp lower.

you know it SEEMS like that should be true, but I think reality isn't as simple.  The reality is things don't cool as uniformly as you may think.  The air from the ceiling just doesn't make its way equally to every sq. in of space, and the thermal mass of "stuff" just takes time to change temps.
 

I live in arizona too, and I'm on the same rate plan you are.  I've been pre-cooling my house for awhile now, and it definately saves money.  I can also now confirm it, because I installed a brultech GEM to monitor electricity.  (I precool by about 2-3 degrees, and because of that, the AC doesn't even turn on for 1.5-2 hours into prime time)
 
However, I think the really big money, is in keeping the multiple AC units in my house from running at the same time.  That peak demand charge is really big, and if you can keep it down, it's something like $15 per kWh.  On that end, my plan is to buy radiothermostat CT80's, and program them to run inside timeslices.  Basically, I'll shut the AC off for 10 minutes, and turn the other one on for 10 minutes and let it do it's thing.  Normally, the AC runs in 15 or so minute slices anyhow, the problem is they are not in sync.  I'm hoping if I can control the big loads like the AC's, and the water heater in prime time, I can cut my demand charge down.

I really agree with you on alternating the AC units. Its definitely an art.  Does each cool half your house?
 
I used to have two 10 SEER units that didn't work great. One mainly is for my master, and the other cools most of the rest of the house.  What I used to do was detect when one side was running, then turn the opposite thermostat off. When that side was complete, I would switch, and run that side until complete. Seems logical until it got to 120, and the large one never stopped. 
 
So I just replaced my two units with new ones that seem to work much better.  I still have the master unit (3 tons) and the other one (4 tons) is split into three zones.  So what I do now is this, I detect when ALL three zone of the big unit are compete, then I turn them off for 12 minutes, and run the master bedroom unit for the 12 minutes. When time is up, it gets shut off, the other units are again turned on, and nothing happens until at least one zone starts, then finishes, and it repeats.
 
I did this because I was discovering that my three zones were not synced at all. One zone would run, then finish, and a minute later another would start, and this is not very efficient for the AC. I wanted all to run at once if I could, so by turning off the zones for 12 minutes, and turning them back on at once, they would more likely be in sync.
 
I like your alternating approach, but I see potential problems.  First, your turning on and off ACs pretty often.  That might cause some added wear and tear, and also, when its hottest, you might be losing some efficiency because it takes a while for them to cool ducts and run at top efficiency. Just when they get there, you are turning them off. I've heard it takes 10 - 12 minutes for them to reach peak efficiency.
 
How does it perform when its really hot?

 

[ano]

One other comment for Lou especially is to realize air is a very poor conductor of heat.  If you pour a gallon of water in a bowl, and then take a 1500w hair dryer a foot away and point it at that bowl of water, even if the air is VERY hot, its going to take a very long time to heat that water up.  Put that water in a pot on a stove and it will heat much faster.  The same is true with AC coming in your room.  The air feels cool, but that doesn't mean the stuff in a room is going to instantly cool. It takes hours and hours.  Then it takes hours and hours to give back its cool "energy." 

 

[apostolakisl]

One other comment for Lou especially is to realize air is a very poor conductor of heat.  If you pour a gallon of water in a bowl, and then take a 1500w hair dryer a foot away and point it at that bowl of water, even if the air is VERY hot, its going to take a very long time to heat that water up.  Put that water in a pot on a stove and it will heat much faster.  The same is true with AC coming in your room.  The air feels cool, but that doesn't mean the stuff in a room is going to instantly cool. It takes hours and hours.  Then it takes hours and hours to give back its cool "energy." 

 
Yes, I am aware of how to use air as an insulator.  This is not relevant to the point I am making.
 
You are blowing 45 degree air into a room which is mixing with 70 degree air, but the air isn't getting any colder.  Something is dumping heat into your air, just as fast as you AC is taking it out.  Thermal mass can't explain that.  The objects aren't generating heat, only giving up what heat they have, and in the process cooling off.  As they cool off, they aren't capable of warming the air, so the air should get colder as you pump more 45 degree air into the room.  But the air isn't getting colder.  Heat is coming from something else besides the thermal mass of the stuff in your house.  Since this plateau is happening early in the day before it is very hot outside, I find it difficult to believe you are getting much heat from outside.  The other possibility is that your AC isn't pumping in 45 degree air.

 

[garbled]

I really agree with you on alternating the AC units. Its definitely an art.  Does each cool half your house?
 

I like your alternating approach, but I see potential problems.  First, your turning on and off ACs pretty often.  That might cause some added wear and tear, and also, when its hottest, you might be losing some efficiency because it takes a while for them to cool ducts and run at top efficiency. Just when they get there, you are turning them off. I've heard it takes 10 - 12 minutes for them to reach peak efficiency.
 
How does it perform when its really hot?

 
So, I have 3 units.  One for a guest house, and two for the main house.  Normally, they tend to run for about 5-10 minutes, every 15-30 minutes, left to thier own devices.  So based on that alone, I don't think I'd really be hurting them at all by telling the thermostat to "shut off" every so often.  If it's already off anyhow, it will be a nop, and if it's on, it's no different than the end of the cycle.
 
Currently I have Honeywell T8000's running them all on a schedule.  It may be the case that they should run longer, less often, but that seems to not be the operation mode of the T8000.
 
As for how it works in the summer, it works great.  It's colder at night, which I prefer for sleeping, and the set temp during the day is warm, like I usually like it, and it keeps up fine.  When I finally get around to wiring up some controllable thermostats, I'll let you know how that turns out.

 

[az1324]

It's demand per hour though so it shouldn't really matter whether they run at the same time or not unless you are spreading it out enough to get around that. Also I've found that it's somewhat unrealistic to try and keep the demand very low because undoubtedly one hour of one day during the month someone is going to run it up for one reason or another. Also other household members don't like it when you scold them for turning things on during the day. So once you take that into consideration and the fact that on peak is 2x off peak rates (ignoring demand), it may be better to do the periodic operation with limited pre-cooling. But if you're really able to keep the demand low then maybe not.

 

[Automate]

I also do use an IR thermometer, a Fieldpiece SIG-1, but remember it measures the SURFACE temperature

 
In general objects with high thermal mass also have high thermal conductivity so there should not be too much difference between the surface and the interior
 
Another thing that would help is to have good air flow across the surface.  Such as a ceiling fan running and blowing air over your tile floor and/or granite..

 

[ano]

It's demand per hour though so it shouldn't really matter whether they run at the same time or not unless you are spreading it out enough to get around that. Also I've found that it's somewhat unrealistic to try and keep the demand very low because undoubtedly one hour of one day during the month someone is going to run it up for one reason or another. Also other household members don't like it when you scold them for turning things on during the day. So once you take that into consideration and the fact that on peak is 2x off peak rates (ignoring demand), it may be better to do the periodic operation with limited pre-cooling. But if you're really able to keep the demand low then maybe not.

You are right. Looks like they have changed the rates YET again. Now peak is only double off-peak, but the one hour peak charge is up to $13.50 KWH.  That is pretty nasty.  Do you know EXACTLY how they calculate that? So its the highest AVERAGE peak during an hour, but is it on-the hour hours, or is it a sliding hour? 
 
I also see they have more rates. They have one with a "super-peak" where from 3 to 6pm in July and August electricity is 10X low-peak, and 5X low-peak noon-3pm and 6pm-7pm.  Then they have indicate to you that its REAL expensive somehow. I'm not sure how that happens though.  Then they have an "experimental" with a super-low rate from like 11pm to 5am. 
 
They used to let you change rates every month, and when they put the smartmeters in they claimed you could change rates once a day, but now it looks like they only let you change rates once per year. 
 
I think there is potential to save money with these rates, but it will take a super computer to figure out how.  :angry2:
 
 

So, I have 3 units.  One for a guest house, and two for the main house.  Normally, they tend to run for about 5-10 minutes, every 15-30 minutes, left to thier own devices.  So based on that alone, I don't think I'd really be hurting them at all by telling the thermostat to "shut off" every so often.  If it's already off anyhow, it will be a nop, and if it's on, it's no different than the end of the cycle.

 
All I can say is that I heard residential AC units are rated for four cycles per hour of operation.  If you exceed that you are risking shorter life.
 
I use HAI OmniStat2's and they have a setting for minimum ON time and minimum OFF time.  I have the minimum ON time set to around 10 minutes and the minimum off time to 12 minutes.  This means I never exceed 2.7 cycles per hour.  Most times it is much less.  Longer cycle times are also more efficient, but temps fluctuate a bit more.

 

[apostolakisl]

Maybe I didn't communicate what I'm doing clearly.  I do have a device that prevents my air temps from dropping below 70 degrees. That device is called a thermostat.  I do have to live in this house, even when precooling, so 70 is the limit.

 
Well that changes things.  I was under the impression that the unit was running continuously.  
 
At this point you really only need to do is wait for the hvac to get your house to 70.  Then start recording the temps of various objects in the house.  See how long it takes for everything to reach a steady state temp.  Indoor walls and objects should hit 70.  Outside walls (and maybe the floor depending on soil temp under the house) will find a higher steady state.  Once it hits steady state, there is no added benefit of keeping things cold longer.  As mentioned by Automate, things with high thermal mass usually have high conductivity, so the surface temp pretty well tells you the core temp.  Basically, high thermal mass is a result of high mass density, and dense stuff rarely is an insulator.  Atoms that are tightly packed together transfer the kinetic energy of their vibrations to each other very efficiently.
 
Air is a great insulator, but only when it is contained in small packets.  Air moves very easily, so the heat doesn't need to pass through it, the hot/cold air itself just moves.  Convection being the most obvious method.  That is why insulation is always made of some material full of air spaces that don't allow easy movement.  Fabrics, down blankets, foam insulation, fiberglass, hair, etc.  It all works the same way.  Look at the insulating ability of solid glass.  It's not very good.  But fiberglass is a great insulator because of the air trapped in between the fibers.

 

[az1324]

You are right. Looks like they have changed the rates YET again. Now peak is only double off-peak, but the one hour peak charge is up to $13.50 KWH.  That is pretty nasty.  Do you know EXACTLY how they calculate that? So its the highest AVERAGE peak during an hour, but is it on-the hour hours, or is it a sliding hour? 
 
I also see they have more rates. They have one with a "super-peak" where from 3 to 6pm in July and August electricity is 10X low-peak, and 5X low-peak noon-3pm and 6pm-7pm.  Then they have indicate to you that its REAL expensive somehow. I'm not sure how that happens though.  Then they have an "experimental" with a super-low rate from like 11pm to 5am. 
 
They used to let you change rates every month, and when they put the smartmeters in they claimed you could change rates once a day, but now it looks like they only let you change rates once per year.

I don't think the rates have changed very much for this plan. You may have been looking at the rates for another similar plan. I believe it is hour on the hour then take the total # of kW you use within that hour (same as avg kWH) and take the maximum such value for the billing period.

 

[IndyMike]

Hello ano,
 
Let me start by saying that I think your pre-cooling approach has value.  However, you should not be viewing the outdoor temperature as the main culprit.
 
Given that you have a well sealed (limited infiltration) home, your main enemy is the Sun's solar radiation.  The majority of the heat flux will be delivered through the walls and windows.  For vertical surfaces (wall and windows) the solar radiation will be at a maximum at sunrise and sunset (sun angle near 90 degrees to the surface).  The outdoor temperature lags your wall/window temperature by many hours since it is principally influenced by horizontal surface heating.
 
I generated the plot below from one of the Phoenix area WeatherUnderground stations that measures both Air Temperature and Solar Radiation (Horizontal).  I generated the Vertical radiation curve based on the observed 1000 W/m^2 maximum and the sun's angle above the horizon. 
 
As shown, you will begin taking serious heating from the sun's Rs immediately after sunrise on your South and East exposures.  This heat will take time to make it through walls due to the thermal resistance of your insulation.  The heat path through windows will be much faster (even Low E windows) - draw the blinds on these exposures.
 
I've taken a lot of liberties in the vertical calculation below.  There are other terms (reflected and diffuse radiation) that will predominate when the sun's angle of incidence is near 0 (i.e. at noon you may still be taking significant Solar Radiation on vertical surfaces).
 
I'm making this distinction because your vertical surfaces are heating far earlier than the outdoor temperature would indicate.  Windows on the exposed walls could be a significant contributor to your heat loading even though the outdoor temperature is cool.
 
If you're interested in the radiant heat load calcs (with examples) : http://nptel.iitm.ac.in/courses/Webcourse-contents/IIT%20Kharagpur/Ref%20and%20Air%20Cond/pdf/R&AC%20Lecture%2032.pdf
 
 

 

[ano]

I don't think the rates have changed very much for this plan. You may have been looking at the rates for another similar plan. I believe it is hour on the hour then take the total # of kW you use within that hour (same as avg kWH) and take the maximum such value for the billing period.

I'm sure you are correct. I think digging more into these rates will be my next task.
 
 

Hello ano,
 
Let me start by saying that I think your pre-cooling approach has value.  However, you should not be viewing the outdoor temperature as the main culprit.
 
Given that you have a well sealed (limited infiltration) home, your main enemy is the Sun's solar radiation.  The majority of the heat flux will be delivered through the walls and windows.  For vertical surfaces (wall and windows) the solar radiation will be at a maximum at sunrise and sunset (sun angle near 90 degrees to the surface).  The outdoor temperature lags your wall/window temperature by many hours since it is principally influenced by horizontal surface heating.

Wow those are some good observations.  So the next question would be how do I use that information? While certainly South/East walls take exposure earlier, its rooms with walls with Western exposure that get the hottest by 7pm.  Maybe its wall sun exposure AND outdoor temps that combine.  South/East rooms maintain their cooling the best.
 
My goal is to save money with a minimum of discomfort, of course.  ^_^  As I said, I think I next need to REALLY dig into these different rate plans.  I think its very interesting how the utility just throws out all these rates, then basically says, "go pick one" when basically few if any customers have REALLY figured out how to best profit from these rates. They should provide a "roadmap" of how to best use each rate, but they don't.  They will tell you what you'd pay under each rate but only for a full year and after-the-fact. 

 

[apostolakisl]

Hello ano,
 
Let me start by saying that I think your pre-cooling approach has value.  However, you should not be viewing the outdoor temperature as the main culprit.
 
Given that you have a well sealed (limited infiltration) home, your main enemy is the Sun's solar radiation.  The majority of the heat flux will be delivered through the walls and windows.  For vertical surfaces (wall and windows) the solar radiation will be at a maximum at sunrise and sunset (sun angle near 90 degrees to the surface).  The outdoor temperature lags your wall/window temperature by many hours since it is principally influenced by horizontal surface heating.
 
I generated the plot below from one of the Phoenix area WeatherUnderground stations that measures both Air Temperature and Solar Radiation (Horizontal).  I generated the Vertical radiation curve based on the observed 1000 W/m^2 maximum and the sun's angle above the horizon. 
 
As shown, you will begin taking serious heating from the sun's Rs immediately after sunrise on your South and East exposures.  This heat will take time to make it through walls due to the thermal resistance of your insulation.  The heat path through windows will be much faster (even Low E windows) - draw the blinds on these exposures.
 
I've taken a lot of liberties in the vertical calculation below.  There are other terms (reflected and diffuse radiation) that will predominate when the sun's angle of incidence is near 0 (i.e. at noon you may still be taking significant Solar Radiation on vertical surfaces).
 
I'm making this distinction because your vertical surfaces are heating far earlier than the outdoor temperature would indicate.  Windows on the exposed walls could be a significant contributor to your heat loading even though the outdoor temperature is cool.
 
If you're interested in the radiant heat load calcs (with examples) : http://nptel.iitm.ac.in/courses/Webcourse-contents/IIT%20Kharagpur/Ref%20and%20Air%20Cond/pdf/R&AC%20Lecture%2032.pdf
 
 

 
Mike,
 
That is a great article.  Tell me, where did you get these graphs?  Are these your own measurements at your house?  If so, what did you use to make the measurements?  I see that dip in the red line, I assume clouds?
 
 
I did some much less complete stuff measuring the temp of the exterior surface of my house and measuring the interior temp of those same spots.  It was a testament to how well foam insulation works.  For example, at 5pm last summer the spot surface temp of my roof was 145 degrees and the surface temp of the insulation on that same spot underneath was 88.  That is a huge delta.  This particular patch of roof was East facing, high pitch no less, but the sun was high enough that it still was hitting it, albeit at a rather sharp angle.  West facing roof just can't be gotten to without serious effort and risk (and extreme discomfort from ridiculous temps).  My attic stays in the 80's even when it is 105 and sunny outside.  The attic has no vents, just the cooling that occurs by virtue of the fact that there is no insulation between it and the conditioned house underneath.
 
Regarding the window gain.  Of course this depends on what type of glass and frames you have.  The various low-e glass and glass with various types of reflective materials applied to the interior and exterior have a huge impact.  The window companies design different windows for different latitudes, specifically blocking or allowing IR in vs out and high angle vs low angle passage of IR.  I have also had some more vulnerable windows tinted such that IR is not passed through (at the expense of visible lights as well).  But it does have a fairly obvious reduction in the temp of objects the light strikes inside the house.  Just as an interesting note, I can't get grass to grow under my windows.  At first I didn't get it, thought the irrigation was missing.  Then I figured out it was the IR reflection off the glass when the sun was higher in the sky.  The visible light really did not reflect much, you couldn't see that it was brighter in those areas.  But if you laid on the dead grass and looked up into the windows, you could see the reflection of the sun in the glass.  As you moved around, you could see that dead grass corresponded very well to the areas that received reflection.