A/C Efficiency--Air v. Water.....

[Naja002]

1 watt is approximately 3.413 BTU/h
1000 BTU/h is approximately 293 W


A btu is a btu regardless of aplication....I've got that much.

But A/C units are rated by how they perform with air. Water is much more thermally conductive then air, so I'm wondering:

Would an A/C chiller work "Better" with water then with air?

In other words, would it be able to handle more btu with water--even if its through "efficiency" (ie, faster thermal transfer)?

I know what I'm trying to ask, but not sure the above makes much sense.....If not, please let me know and I'll try to re-word it. :up:

[DetroitAC]

It's a bit too complicated a question to just give a simple answer.
What are you cooling? A/C stands for air conditioning, and chiller refers to cooling a liquid, so "A/C chiller" has me confused...

maybe you can elaborate more...

[OldChap]

I'm reading that as what would happen if instead of pushing air through an ac evap you cooled it with water (put it underwater?)...could be wrong though

[Naja002]

It's a bit too complicated a question to just give a simple answer.
What are you cooling? A/C stands for air conditioning, and chiller refers to cooling a liquid, so "A/C chiller" has me confused...

maybe you can elaborate more...


Ok, NP. A/C chiller is just taking an A/C and making a chiller out of it. Nothing new. Been done many times before. I thought that was the "standard" term for it, but maybe its called something else....:shrug: Like just a "Chiller".....:shrug: Apologies. :p:




I'm reading that as what would happen if instead of pushing air through an ac evap you cooled it with water (put it underwater?)...could be wrong though

Right. Not so much about pushing or moving water through the evap, but just a general chiller setup--evap in the res.




I'm really just wondering:

If a 5K btu A/C is good for ~1500watts with air......can it handle any more then that with water? Either through capacity or efficiency (ie, increased thermal transfer).

[Sam__]

So are you sugesting you buy a rad bigger than the one on the AC? pumping water round it and through the one on the AC to cool it more efficiently?

In my opinion i would say its not really worth it if you can just get a nice fan on the AC.

[repsol_23]

I have a setup similar to what you are asking about. I have modded an AC window unit to cool my water loop quite effectively I might add. Here are a couple of pics: first one is when I was just trying it out to see if it would work and the second is with a better cooler box. I have been quite satisfied with my results!! Here is a link to a thread I had started in the TEC section.

http://www.xtremesystems.org/forums/showthread.php?t=199415

Hope this helps.

edit: I forgot to mention that I am just cooling the air within the box and using the fans on my radiator in the box to chill my water. On a nice cool night such as last night I can get temps within the cooler down to @ -7C. With these temps my cpu at load never goes above 0C and my gpu never goes above 23C.


1 watt is approximately 3.413 BTU/h
1000 BTU/h is approximately 293 W


A btu is a btu regardless of aplication....I've got that much.

But A/C units are rated by how they perform with air. Water is much more thermally conductive then air, so I'm wondering:

Would an A/C chiller work "Better" with water then with air?

In other words, would it be able to handle more btu with water--even if its through "efficiency" (ie, faster thermal transfer)?

I know what I'm trying to ask, but not sure the above makes much sense.....If not, please let me know and I'll try to re-word it. :up:

[Naja002]

Ok, Everybody let's step back for a second. Its just an academic question really. It seems that Ya'll are thinking way too much into this and thinking more about what I want to Do then what I want to Know.




It's a bit too complicated a question to just give a simple answer.
What are you cooling?

maybe you can elaborate more...

All I want to do is take an A/C and make a chiller. No big deal....its been done many times before. :up: I currently have about 688watts being cooled by water now (=3x OCed quads + 4x OCed GPUs). That's a fair estimate of what's actually being cooled by water now. Its doesn't include heat from the PSUs, HDs, etc. My current setup is linked in my sig. Its the Multi-PC car rad setup that Ya'll may have already seen. Remove the 55gal barrel and the car rad.....Install 5-day cooler and chiller. I'm not looking for sub-zero temps and hope/expect to be able to use the thermostat that's on the A/C. I think 64F or above will suit my purpose--I'll just have to check that out after everything is setup.

One thing I don't want is to pay for an A/C to run constantly. I will still need to run the water pump...so that's a given. My current setup uses 2x 9" fans that total about 50w on low and 72w on high. I realize that an A/C will use considerably more electric, but setup correctly.....I think it will be worth the added expense to me.

I may add another PC to my setup--down the road, so I would be adding and estimated 208w (OCed quad, NB and OCed GPU), totaling 896w actively being removed by water. I subtract 20% from the estimated total for things like PSU waste/heat.

So the max total for this setup would be:

1120w - 224w (20%) = 896w actively removed by water. That's IF I added another rig. Otherwise I'm looking at ~688w.

A 5K A/C is good for ~1,500w from what I understand. But with a good Coleman 5-day cooler of decent size (say ~70qt/17.5gal) for a res.....Would a 5K A/C be able to maintain a decent temp--without switching on/off, on/off. Or would I be better off with a 6K? 8K?

Some points about the res:

From what I understand the Coleman brand 5-day coolers actually maintain ice for 5-7days. Whereas other brands such as Igloo don't...even though they claim to be 5-day.

The purpose of the larger res is to run the A/C a bit longer when its running....and hopefully allow it to be off longer. I don't have any plans of going sub-zero or anything, so I can just use straight water. :up:

An 8K A/C is only about $40 more than a 5-6K A/C. My 5K uses right at 500w when the compressor is going. My 8K uses right at 700w.

The cooler itself will be ~$50 after tax. So, would I serve my purpose better by buying an 8K or a 5K and putting the $40 toward the cooler?


This is the long version......:p:

[Naja002]

Ok, in an effort to answer my own question....I did some math. What I come up with is: Its about the same.

The 5K that I have is about 4yrs old and I don't know what the energy-star rating is on it, but killawatt shows that it uses right at 500w.

The 8K that I have is new and has an energy star rating of 10.8 and killawatt shows that it uses right at 700w.

So, my math shows 707w more per day to run the 5K than the 8K. But if they had equal Enery-star ratings.....I would assume that the figure would be closer....ie, ~5000w/day useage to cool 688w load.

That sound about right?

[majestik]

If you don't want sub-zero (or sub-ambient) temps, i'd just stick with the car radiator. The whole point of chiller is for sub-ambient temps.
If you do want sub-ambient - I don't want to know how much insulation material you'll need :p:

I'm building a chiller out of a 6k btu unit right now. It pulls 3gal. of "winter window washing" fluid in a rubbermaid slim cooler to 0C VERY fast (like, 5 minutes). Then it slows down dramatically because compressor/condenser heat up and all. The fluid is rated for -40C and contains methanol - so probably just dilluted methanol with water. Basically, lower heat capacity than water.
I don't know how fast a 5k btu AC will pull down on a 50gal tank, but it will definately be pretty slow.

p.s. cooler maintain 5 day ice WITHOUT LOAD.

[DetroitAC]

If a 5K btu A/C is good for ~1500watts with air......can it handle any more then that with water? Either through capacity or efficiency (ie, increased thermal transfer).

Yes, it will handle more with water at your stated conditions than with air. The reason is that your stated desire is 64F. This is the largest factor affecting the compressor's cooling capacity. @64F you will have a much higher suction gas pressure, higher suction gas density thus more cooling capacity.

The fact that the evap is submerged in water will probably help a bit also, as the evap will be more effective than with air, but without knowing te flowrate through the evap (if it's just submerged, it's only free convection flow), it's really just speculation.

Would a 5K A/C be able to maintain a decent temp--without switching on/off, on/off. Or would I be better off with a 6K? 8K?
Your calculations were correct, 5k Btu/hour = 1465 Watts (You said 1500, which was close enough), and it'll have more capacity than that because it'll be running at a higher temperature. By looking at the datasheet of a 5k rotary, I'm guessing that it may have 7.5k Btu/hour capacity = 2200 Watts, that's just specualtion really, as I don't know any of your (albeit hypothetical) particulars.

so if your cooling load is only 688 Watts, it must switch on-off. Let's say the cooling capacity of the unit is 2200 Watts. It'll be running 31% of the time, and off 69% of the time. In reality, the capacity isn't a nice fixed number like that, it varies even on every cycle, but you get the idea...

The amount of time that it runs and is off depends on the size of your reservoir, the settings of the controller (hysteresis) and only to a tiny degree on the reservoir insulation (since the heat influx will be trivial compared to the heat load).

So, no, you don't need to get any bigger, you'd already be ~3X oversized. :up: plenty of room for future upgrades. You'll also be much louder than you need to be. I'd prefer a properly sized quiet little chiller myself...

[majestik]

Just to add an idea for you.

Dig a 10ft hole in the ground, put a 10gal drum in it, fill it up with coolant of choice, run a powerful pump and bury it. There we go - cold and 0$ running cost (other than the pump).

[Naja002]

Thank You, DetroitAC! That's the kind of info that I was looking for....:up:






Just to add an idea for you.

Dig a 10ft hole in the ground, put a 10gal drum in it, fill it up with coolant of choice, run a powerful pump and bury it. There we go - cold and 0$ running cost (other than the pump).

Honestly, Instead of the barrel I would be more inclined to do something like this:

Ground loopin' - Look ma, no fans or rads. (http://forums.extremeoverclocking.com/showthread.php?t=135691)

[[XC] Hicks121]

See Naja, you were right in line with your thinking! :up:

[Xeon th MG Pony]

No it will handle the rated btu regardless of air or water the only thing will improve is the efficiency of your transfer fluid and the fact you can focus the cooling.

The compressor will only move so much as will the condenser, water is not a magical substance it will not increase the capacity, but it will allow you to bank reserve capacity by using a storage tank of chilled water for when you need to sink a larger load then the system alone can handle.

Will it improve efficiency? Yes (Becuase we are focusing where we are pulling the heat we get less unwanted btus in out coolant medium).
Will it improve capacity? No (In order to take in more btu's with out our water warming, we must reject more heat & pump more refrigerant, for that need bigger compressor & condenser!)
Is it a workable project? Yes (It has been don quite succefully many times here on this forum and the main part is all ready prebuilt & balanced for you!)
Is it a good way to make a cheap but effective chiller? Yes (As above the parts matching has been don for you, all you need to do is "assemble" the kit)

[Xeon th MG Pony]

Ok, in an effort to answer my own question....I did some math. What I come up with is: Its about the same.

The 5K that I have is about 4yrs old and I don't know what the energy-star rating is on it, but killawatt shows that it uses right at 500w.

The 8K that I have is new and has an energy star rating of 10.8 and killawatt shows that it uses right at 700w.

So, my math shows 707w more per day to run the 5K than the 8K. But if they had equal Enery-star ratings.....I would assume that the figure would be closer....ie, ~5000w/day useage to cool 688w load.

That sound about right?

An ac uses most energy when starting, so if any thing you want it sized to be on more then off for best power utilization. If you over size it you want a large thermal bank to allow a good long worth while off cycle!

Another idea is allow it to build a large level of ice that way you can store a good amount of thermal storage with less space.

[Naja002]

An ac uses most energy when starting, so if any thing you want it sized to be on more then off for best power utilization. If you over size it you want a large thermal bank to allow a good long worth while off cycle!

Another idea is allow it to build a large level of ice that way you can store a good amount of thermal storage with less space.

Ok, I understand what You are getting at, but pursuing ice changes all of the variables for me. I have no need or desire for temps that low....or the complications that come with it. :up: With what I am interested in doing: I don't need to insulate the PC hardware or use anything other than distilled. I have no fear of those things....just no need. Maybe this will lead to that, but right now I would just prefer to keep it simple. :up:





No it will handle the rated btu regardless of air or water the only thing will improve is the efficiency of your transfer fluid and the fact you can focus the cooling.

The compressor will only move so much as will the condenser, water is not a magical substance it will not increase the capacity, but it will allow you to bank reserve capacity by using a storage tank of chilled water for when you need to sink a larger load then the system alone can handle.

Will it improve efficiency? Yes (Becuase we are focusing where we are pulling the heat we get less unwanted btus in out coolant medium).
Will it improve capacity? No (In order to take in more btu's with out our water warming, we must reject more heat & pump more refrigerant, for that need bigger compressor & condenser!)
Is it a workable project? Yes (It has been don quite succefully many times here on this forum and the main part is all ready prebuilt & balanced for you!)
Is it a good way to make a cheap but effective chiller? Yes (As above the parts matching has been don for you, all you need to do is "assemble" the kit)


Ok, I understand what you are saying, but I think we are all on the same page in that the unit "performs" better with water than with air....whether its through capacity or efficiency. That's why I asked in the OP about either one or both. I am looking for the bottomline of performance, but I do appreciate understanding the more technical details that make-up the answer. So, Thank You! :up:




The amount of time that it runs and is off depends on the size of your reservoir, the settings of the controller (hysteresis) and only to a tiny degree on the reservoir insulation (since the heat influx will be trivial compared to the heat load).



What would be a reasonable assumption be for the average window-type A/C? Meaning how many degrees between on and off.