Efficiency of a homemade Air Conditioning unit

[lando45]

TL;DR Summary

Yesterday I built a homemade AC unit, it works quite well but I think I can increase its efficiency.

Hi guys,

I used to post on here quite a bit back in 2005-2007 when I was studying Physics in school, I remember it always being a great resource so hopefully I can get some similar help with this "real life" problem I'm having now!

We are having a mini-heatwave here in the UK the last few days, making it really difficult to sleep comfortably at night. I started looking into the idea of getting an AC unit installed in our bedroom, but as I was searching around online I saw this video pop up on YouTube:



So I thought, why not try to make that as a fun DIY project! I sourced all the parts and constructed it yesterday. Only difference to the one in the video is that instead of installing the air outlet pipe and the fan directly into the lid of the coolbox, I have instead got a bit of insulating board which acts as the lid and I have installed the air outlet pipe into the insulating board (this way, I can still use the coolbox as a coolbox when I need to). Below is a picture of my setup:

We used it last night for the first time, I put a big frozen block inside (approx 20litres). The air that came out was cool (definitely cooler than if you just had the fan aimed at you directly), but I wouldn't say it felt truly cold like a proper Air Conditioner. By the time we woke up this morning the ice was all melted, I don't know at what stage of the night that happened. I think there's room for improvement. Bearing in mind I haven't studied physics for 13 years, my knowledge of the theory is very rusty but below is a few of the questions I had!

1) What effect will increasing the amount of ice have? Would it a) decrease the temperature of the air output, b) keep the cold air blowing out for longer or c) both?

2) Is the "empty space" in the box a factor? The coolbox I have there is about 90litres I think, if I'm only putting e.g. 20litres of ice in, that leaves 70litres of empty space. If I used a smaller coolbox e.g. 40litres, would that have any effect?

3) Does the shape of the ice I put in make a difference? Last night I put in one big block. If I used the same amount of ice but split up into smaller cubes, what difference would that make? My instinct says it would make the air colder as the total surface area of the ice would be greater, but perhaps the ice would melt quicker which might cancel that out

4) Does the speed/power of the fan make a difference? Presumably it does, the fan I used there has 4 settings (Off, 1, 2 and 3) - I used 3. Would a stronger/more powerful fan mean more air output through the outlet pipe?

5) Does the positioning of the ice within the box make a difference? I feel like if there was just one big ice block resting on the base of the cool box, that wouldn't work as well as if the ice was elevated and the air from the fan could circulate all around the block. Perhaps if I rested the ice block on an elevated grille it would work better

Those were all the questions I had, would be great if anyone has any input. Any other ideas/suggestions of course also welcome! Thanks

EDIT: One other question that I forgot to include!

6) I saw some people in the comments of the YouTube video suggest that including salt with the ice would make it colder. How does that work? I think mixing salt to the water prior to freezing would lower the freezing point to below 0degC. So my freezer would take longer to freeze it, but then it would stay frozen for longer overnight (and at a colder temperature)?

EDIT2: And another..
7) Would this be more effective at night with the bedroom windows/door open or closed?

 

[jrmichler]

The rate of cooling depends on the amount of air flowing through and how well the ice transfers heat to the air. Air flowing around the entire ice block transfers more heat than air flowing over the top of the block. Breaking up the block into chips increases the heat transfer area, which increases the rate of heat transfer. The total amount of cooling is the same, but you can control the rate that you use it. The higher the rate of heat transfer, the sooner the ice is all melted.

A small window air conditioner is rated at 5000 BTU/hr. Smaller units are not made because they would be too small to be useful. Your 20 liters of ice weighs about 40 lbs. The latent heat of ice is 144 BTU/lb, so the total cooling capacity of that ice is 40 lbs X 144 BTU/lb = 5760 BTU. If that cooling is spread out over 8 hours, that's 720 BTU per hour.

You would get the most benefit from this small amount of cooling by running the cool air through a hose set to discharge under your blanket. That way all of the cooling would be going directly to you. Even simpler would be to snuggle up to a frozen water bottle.

Note that we call it heat transfer even when we are discussing cooling.

 

[lando45]

That's helpful, thank you. So for a given volume of ice, the total "cooling power" is fixed. But the rate at which it's used is dictated by (among other things) the size/shape of the ice. So one big block would cool at the slowest rate (but would last longest), whereas the same block broken into small chips would result in a stronger cooling effect, but it would run out sooner.I also gather from your response that elevating the ice (e.g. onto a raised grille) would increase the surface area of ice that the air is exposed to, so that should help.

What about the size of the cooler box, how relevant is that (if at all)? All other variables being equal, would there be any difference between using e.g. a 30 litre cool box (20 litres of ice and 10 "empty" litres) and a 90 litre cool box (20 litres of ice and 70 "empty" litres)?

 

[Averagesupernova]

Salt will lower the freezing temperatures of water. That's it.

 

[hutchphd]

Salt will lower the freezing temperatures of water. That's it.

If I am interpreting correctly there are quite a number of ice cream aficionados who might take issue with this. Water ice in salt water will reach a temperature below 0 C.

 

[Averagesupernova]

Your ice is already as cold as it will get with a given freezer. Why allow it to be water instead of ice at that same temp?

 

[lando45]

Your ice is already as cold as it will get with a given freezer. Why allow it to be water instead of ice at that same temp?

Yep I'm with you there, having thought about it I don't see any benefit to adding salt

I'm still interested in whether a smaller coolbox would make a difference, and also whether keeping the bedroom door/windows closed would help

 

[Dullard]

Interestingly (or maybe not):

A 'Ton' of refrigeration (A.K.A. 12,000 BTU/Hr) is what is required to melt a ton of ice in 24 Hrs. You can use that relationship along with the quantity/rate of ice melt to compare to a more conventional HVAC option.

My company has a small turboprop aircraft - we use one of these for cabin cooling on the tarmac. It's amazing how fast ice can melt. I can't imagine trying to feed one (producing useful room cooling) for 8 Hrs.

If you turn your bed into a well-insulated tent (and draw/return air from inside), I'd expect that you could be reasonably comfortable while melting around 7 LBs of ice/Hour (per person). Getting the fan speed right is the difference between heatstroke and frostbite.

 

[hmmm27]

Buy a case or two of bottled tapwater (of quarter or half-liter disposable plastic bottles) and don't even open the bottles : just freeze and throw into the box... over and over, again. The only mess you have to deal with is condensation (as opposed to 20L of loose water).

In combination with the bottles, the design you have would benefit greatly from a vertical partition, dividing the cooler into two chambers, all the way to the top (lid) and sides, but leaving an inch or so at the bottom for airflow and clearance for condensation collection.

 

[lando45]

Buy a case or two of bottled tapwater (of quarter or half-liter disposable plastic bottles) and don't even open the bottles : just freeze and throw into the box... over and over, again. The only mess you have to deal with is condensation (as opposed to 20L of loose water).

In combination with the bottles, the design you have would benefit greatly from a vertical partition, dividing the cooler into two chambers, all the way to the top (lid) and sides, but leaving an inch or so at the bottom for airflow and condensation collection.

Thanks. I will look into a vertical divider

Regarding the plastic bottle suggestion, will the cooling effect of the ice not be quite significantly reduced if the ice is all enclosed in plastic?

 

[hmmm27]

Thanks. I will look into a vertical divider

Regarding the plastic bottle suggestion, will the cooling effect of the ice not be quite significantly reduced if the ice is all enclosed in plastic?

Yes, heat transfer will be slower (no idea how much slower) from the intervention of the thin plastic shell and, as time goes on, be slower still (the difference between naked ice shedding meltwater, which surface is always at 0C until it's gone, and a chunk of ice inside a layer of water inside a bottle).

Still I'd think the added surface area more than makes up for it.

 

[Spinnor]

A 'Ton' of refrigeration (A.K.A. 12,000 BTU/Hr) is what is required to melt a ton of ice in 24 Hrs. You can use that relationship along with the quantity/rate of ice melt to compare to a more conventional HVAC option.

Do you have that backwards, a ton of refrigeration required to produce a ton of ice as refrigeration generally cools things?

https://www.google.com/search?q=A+'Ton'+of+refrigeration&oq=A+'Ton'+of+refrigeration&aqs=chrome..69i57j69i64l2&sourceid=chrome&ie=UTF-8

 

[Lnewqban]

Summary:: Yesterday I built a homemade AC unit, it works quite well but I think I can increase its efficiency.

EDIT2: And another..
7) Would this be more effective at night with the bedroom windows/door open or closed?

I would not worry about the efficiency of this machine.
Consider it to be an enhanced personal fan, in which the heat generated by the little fan's motor works against the ice duration and against your comfort.

Point the airstream directly to your body, which generates around 200 BTU/h of sensible heat (dry heat due to your body's natural temperature) and about 130 BTU/h of latent heat (vaporization heat of your sweat).

Open doors and windows, if safe.
Even better, try some forced ventilation from one side to the opposite one of your room.
You want the make up air for your machine to be as cool as possible.

Heat that accumulated during the hot hours of the day keeps flowing into the room's air during the evening.
If air does not move inside the room, the hottest (less dense) air will accumulate as high as possible while any cooled volume of air (heavier) will sink as close to the floor as possible.
Airflow over your skin always improves the cooling by transpiration (unless upcoming air is as hot and humid as your skin).

 

[Dullard]

Do you have that backwards, a ton of refrigeration required to produce a ton of ice as refrigeration generally cools things?

https://www.google.com/search?q=A+'Ton'+of+refrigeration&oq=A+'Ton'+of+refrigeration&aqs=chrome..69i57j69i64l2&sourceid=chrome&ie=UTF-8

LOL. I wondered who would feel the need to split that hair. As melting ice is the source of the 'refrigeration' in this discussion, I'll stand by my definition.