Two Liter Air Conditioning

The plan is to freeze two liter bottles full of water outside this winter.
Store them in an icehouse (a pit of sawdust).
Then next summer place them in this contraption.

Is there a reason this won't work?

Thanks,
Shawn

I see no reason why this would not work.
However...
The effectiveness of the device is limited to the ration of meltwater to airflow.
I assume this would be installed inside for better efficiency-recycle the already cooled air.  As the ambient temperature around the water cools, the rate of melting would decrease.   

On those really hot days, I think you would be expending much effort checking on empty jugs.

The threads on a soda bottle are not a good match for standardized pipe fittings.  It may be a challenge making the jug and pipe work together, but you would need a means of getting some air into the jugs to balance the air pressure inside the jugs or they will stop dripping.  Dont want to poke a hole in the jugs, they would be single use.

The number of bottles you would need to air condition a small space for a season would be...large.

May I offer some suggestions:
Run your ice house as you normally do, with big blocks
Rather than jugs, install a bathtub or other container which can be fitted with standard plumbing adapters.  This would allow easy work in hooking it all up.  A tub would allow a substantial amount of ice to be melting, and increase the surface area for faster melting.
Next, put in a chute from the icehouse to the tub.  On those hot days, grab a shovel or ice grabbing thingamajig and load up the chute.  Gives you extra capacity.  Saves a lot of hauling.

I've been thinking about a central cooling system which utilizes irrigation water:

Water comes out of the ground around here at about 65 degrees.  If it was directed through a radiator or heat transfer coils through which air is blown, the air would be cooled.  Same idea as yours, but with fresh running water in place of frozen jugs.  Not as cold, but what it loses in temperature it makes up for in volume and continuous operation.

Continuous operation would provide excellent cooling for a structure.  The problem is the volume of water required for continuous operation-about 2 gallons per minute (4 liters or so per minute).  If the water was being utilized, the idea would not be a waste of resources, so irrigation seems like a fine solution. 

The effectiveness of the system would not be as intense as a commercial air conditioner, but it would use a whole lot less energy.  I've done a little math and concluded that I could get about a 10 degree difference between indoor and outdoor temperatures in my current home.  Effectiveness could be improved with more water flow, multiple heat transfer coils, and insulate this drafty old place.

2 gallons/minute, 12 hours/day=1440 gallons per day
at 8 pounds per gallons, 11520 pounds.
If I can raise the temperature of that water by 10 degrees in its pass through the radiator (or multiple radiators), 115200 BTU of cooling.  About the same as a small window sized air conditioner running your electric bill up all day.
If I'm pumping irrigation water to begin with, the only additional energy to operate this system would be the air blower.

Ground water holds its temperature.  Being deep in the ground, the temperature is fairly constant and averages about 60 degrees everywhere in the world, at any time of the year.  Your ground water system and plumbing set up will make a big difference in the temperature where such a system as described above is installed. 

Another use for such a system would be for space heating.  Assuming you have a use for a high volume of water in the winter, this ground water radiator would heat a space, probably to within 10 degrees of the groundwater temperature.  40-50 degrees for an indoor space is not exactly toasty, but it sure would take the edge off.  Perhaps a structure housing animals would benefit best.

Keep us posted on your [s]contraption[/s] system concept.

Ken,

Thanks for the response.

I saw this idea on the net somewhere, a lady in Arizona was using her well water
through a homemade copper radiator to cool her house while watering her lawn.

I much prefer your idea of watering crops.
We are not allowed to dig a well here, so we can not use that idea.

I also like your bathtub and chute idea, unfortunately, living in a small home I need
something that takes up less space.  Also, I have no pond or lake to cut ice from.

So...since I have to set the water out to freeze, why not use two liter bottles.
They are easy to obtain, it keeps them out of the landfill, and my children can move
them.

Yes the cooler would be placed inside, maybe one in each room.

I was thinking of just setting the bottle ends inside a much larger fitting, the ice should
melt at a slow enough pace so leaking and pressure would not be a problem.
And I hope that would allow enough air to gurgle into the bottles to keep the water flowing.

I believe that you are right, a tremendous amount of bottles will be needed.  But they are
free.

This is an attempt to lower power bills until the I can afford a geothermal system.  I might
not be able to cool the whole house all summer, but any savings helps.

Neat design and idea!

Thought: Run some pipe from a shady spot, through the ground, through the ice pit, and into your house. Put a  small DC fan on it. It should pull the air through the cool ground and Ice pit and into your house. Not as much surface area as what you're proposing, but it will definitely work.

In fact, just the pipes in the earth will get you some cooler air.

I've been keeping my 2L bottles to try something like this, however I'm not trying to do AC, but keep a cellar cool.  For my needs at least, there isn't any need to drain the bottles, just let the air flow past them when needed.  The water coming out of your bottles would still be very cold and it seems to me you might be waisting a lot of coolness.

There is a thread on Icehouses around here somewhere.  I live in Florida, I might see ice for a couple of weeks per year, but I have to be up early in the morning, looking in the right direction and get awfully lucky.  An ice house is not an option down here.

I grew up in Maine.  Ice can be made up there 4 months of the year.   I got to thinking about it...

A small form is pretty simple to come up with.  5 gallon buckets are cheap.  Fill them up at night, thaw them just enough the next day to get the 5 gallon block of ice out, roll them into an icehouse.  I watched videos on youtube of crews cutting ice blocks from lakes and rivers.  Cubes are handy in that you can store more ice in a given volume.  There is little wasted space, the ice will last longer.

I scratched my head, thought about it some more.  Ice is pretty easy to cut, if you have the right tools.  Instead of a form for a 40-50 pound block, why not build a giant, monolithic form in the icehouse, fill it gradually, leave the door open during those days when it is 20 below.  Cut it as needed.  Sure would save a lot of handling.  
All you would really need are posts.  The posts support a couple of courses of boards, perhaps lined with visqueen to prevent sticking.  As the ice freezes, the bottom course is removed and installed as the next top course.  

I ran some figures.  Did some long division.  Worked on the problem using that new math.
I came up with about 1000 sqft being possible with water flow at a rate of 150 gallons per hour.  This is the same rate of water flow as a 2.5 gallon per minute 'water saving' shower head.    The water would be able to spread out in a thin layer which I suspect would freeze rapidly.  Each foot in height would equate to around 7400 gallons, and take a couple of days to fill.  This gives you plenty of time to remove and raise boards.  At 10' high in a 12' high icehouse, you still have a little bit of headroom.  This would take about 3 weeks to fill and have a total mass of about 600k pounds.

At a buck for 20 pounds of ice, assuming half of it melts, the marketable ice would have a potential value of $15000.  Sure would be a cool way for a kid to make money in the summer!

Getting the ice to last, the monolith would need to be produced during the coldest part of the year.  I read somewhere that ice melt in a well built icehouse is around 50%, but don't ask me to back up that figure.  Half of the above monolith would leave about 300k pounds of ice and could provide a MASSIVE amount of cooling.  If consumed over 100 days, the monolith could offer 3000 pounds of ice per day.  Depending on how well the heat transfer coils conduct heat to the meltwater from this ice, it could quite easily cool a home of substantial size.

My experience doing something similar.... I have an 8 x 8 X 6(aprox) cellar. It is in the "crawl space" which is 6' high. Two walls and the roof have 4inch foam to insulate. The floor and two outside walls are bare concrete.... the bad part is that one of the outside walls is SE and has 2' above ground (painted black!) and works like a solar collector   Anyway, I have been freezing 4litre milk bottles about 2/3 full (so they don't split... I tried full) Anyway, I use about 5 of them and they go from fully solid to fully thawed pretty quick... The room is cooler, but not much.

I want to keep the concrete walls for mass, but I will probably have to loose the SE wall by insulating it both outside and in... That one wall feels warm in the afternoon. My water supply goes through that room and I am thinking after I do some insulating, to get some pex and form aluminium fins on it to use the incoming water to cool the room if it os cooler than the room. (this means being able to bypass the pex) Then, if we are still here long enough, I had thought to use reverse PAHS to cool the ground under and to the side of the cellar in the winter and use an insulative layer to keep it cool in the summer. I would use salt water that when close to the surface would loose its heat to the cold winter snow (and maybe melt it from our sidewalk) and then sink to the bottom where the warmer earth would warm it back up... thermosiphon style.

Anyway, as with solar heating a home... I am finding I need to first make it require less cooling. I go through way too much ice for not much effect. I need to insulate first. Second, My bottles sweat a lot... they were leaving puddles and now I have them in pans to catch the water. I am not sure what this is doing to the humidity in my cellar. My first thought was that I was making it wet in there, but on second thought i may be drying it out. The air close to the bottles cools and losses its water to the bottle as sweat, I remove that water when I empty the pot... less water in the room. So one way or the other the room humidity is being effected.

Please let us know how it goes... what steps you take... what works/how well... what doesn't.

I have an high pressure air conditioner and it was mounted in space that was the attic. The problem I have is the vibration it creates in the ceiling below. Unless your air compressor is very smooth running all of that vibration is going to be transferred to your ceiling. Possibly after a few days of running your shop will be covered with all the dust that was clinging to the ceiling. Heat issues would rule it out for me.

Len Ovens wrote:the bad part is that one of the outside walls is SE and has 2' above ground (painted black!) and works like a solar collector

$20 in cheap white paint will fix that. It'll take a couple days for the concrete to cool after you paint it, but you should feel a noticeable difference.

Just run air over the bottles. They will transfer the heat exchange and condensation will form. No need for a radiator. A simple cheep cooler will do for a box. Lots in the resale shop.

Then be real smart. Get a solar powered freezer and freeze the bottles. Repeat.

Put a huge water container with a mix of water/alcohol in it inside the super insulated ice house. Open the ice house during the cold winter nights so the antifreeze gets as cold as possible. Close up the ice house for summer. Circulate the antifreeze through a radiator inside during the summer, then back into the ice house, in a loop. Blow air over the radiator and into the room for cooling. Insulate the crap out of the incoming and outgoing pipes. Put something to catch/collect the water that drips from the radiator during summer operation. Water your plants with that water. Or, pour water into a Berky water filter and drink it. This is assuming you keep your radiator clean and there are no toxic metals involved. (congrats, you're also making water from air, no well or city or rain water necessary)

I assume the volume of antifreeze (water/alcohol mix) in the icehouse would dictate how far into the summer the liquid would remain cold enough to cool the room. Flow rate and the temperature of the liquid would determine how cool the room gets. If you have a solar powered freezer, pipe the liquid through that before it returns into the huge tank in the ice house. Run the pump off the same solar panels/deep cycle batteries you're using for the fridge.

Perhaps circulate the antifreeze from the huge tank in the ice house through a very long, metal pipe or hose laid out in the yard during the winter months, looped so it goes back into the tank, to increase the surface area exposed to the cold. That might help to get the liquid in the tank colder, faster during the winter. Blow air through the line to get the liquid back into the tank before the weather warms. Put the hose away so your yard doesn't look too redneckish for summer.

I'd try to line the outside walls of the ice house with stacked sandbags full of sawdust, in addition to the sawdust insulating the tank inside the ice house. Sort of like building an earth bag structure around the outside of the ice house. If the bags aren't too heavy, maybe cover the roof as well. That would provide another foot of insulation, assuming you have access to that much sawdust.

I have no idea how much liquid you would need to last through the summer or how much this setup would cost to install. Logic says it should work, unless the tank doesn't stay cool for some reason. when not in use, blow air through the line to put all the liquid back into the tank, so none sits in the lines warming up when the flow is stopped.

That's how I'd try it, at any rate. It would sure save a lot of work adding and removing frozen bottles constantly---in theory, which is all this is.

Daniel Morse wrote:Just run air over the bottles. They will transfer the heat exchange and condensation will form. No need for a radiator. A simple cheep cooler will do for a box. Lots in the resale shop.

Then be real smart. Get a solar powered freezer and freeze the bottles. Repeat.

What you have to remember here (adopting my chemistry professor persona) is that water has a much higher heat of fusion than the specific heat of either liquid water (1 calorie per gram) or ice ( half a calorie per gram). This means that most of your cooling effect is from the melting of the ice, and that Bill's suggestion above to use antifreezes to lower the temperature of the ice is not going to buy you that much more cooling power. If you just let the cold water drain away and not blow any air over the radiator, you wouldn't lose that much cooling potential compared to the mass of the melting ice.

Now one additional way to cool would be to take the cold water dripping off the system, direct it to some absorbent material and blow a fan over that, making an evaporative cooler. That way, you have two phase transitions working to cool things: (1) the melting ice and (2) evaporating water. Of course for the second to work well, you have to be in a dry climate. Evaporative coolers work well in places like Albuquerque, where the summer dew point rarely goes up into the 50s, but they fail miserably in southeast New Mexico and West Texas, where the summer dew point will spend a lot of time in the 60s.

Well, this is an old thread. Seems like there might be some updates. I was immediately interested because I am tyring to keep my cheese "cave" cool. My first thoughts were that if I were to try it I'd have to freeze a LOT of bottles to get from the cold season to the hottest part of the summer when I would need them. The construction to store those bottles would be close to what it would take for me to put in the cooling channels through the earth under my lawn perhaps.

I do need ideas for cooling, so I hope someone will post something here for me to learn from, and get ideas.

This give me an Idea! But I have to go through a full story to get to that idea. So here goes:

I have a wood stove in my basement. Not connected to a central heating system. In the winter, I need to get the basement up to 30C to get the main part of the house comfortable. So I ran a piece of ductwork from close to the stove to my living room. At the stove end I have a small exhaust fan sucking air into the ductwork and thus blowing it into my living room. The fan is plugged into and extension cord I made run through a light dimmer switch, so I can make the fan run at what ever speed I want. Works wonders for heat transfer. I kept the living room at the same temperature as the basement.

It's getting warm here now. 24C in my living room.... I went to the basement and found it was only 17C there! Hmmmm. I could turn on that fan and blow cool air directly upstairs!

So then we get this thread in the dailish email. It mentions ice house. I have what I call a cold room in the basement. It is insulated to the point that, with 30C basement in the winter, that room was like a walk in fridge! If it had a drain and the walls were water proof, I could fill that room up with ice!

So here's My idea. If you have a basement with an empty room having a drain, turn that into an ice room. The other option would be to build an ice house close enough the your house to be able to run ductwork or dryer hose into a window. Hook up a fan with a dimmer switch to blow the cold air from the ice room to your living space. Turning up the dimmer increases the speed of the fan which increases to volume of cold air transfer. This increase in air transfer means more warm air is being sucked into the ice room and speeds up the melting of the ice. It would work exactly like a commercial air conditioner excepted limited by the volume of ice you have stored.

I am going to throw a monkey wrench in this idea for AC use. Air conditioners can be rated in tons. That is tons of ice melted in a typical 24 hour period to produce the cooling. A SMALL residential cooling unit will be rated at 3 tons. That means the heat transfer it produces is equivalent to melting 3 tons of ice over the course of a day. 2000 lbs / 62.4 lbs/ft^3 = 32 ft^3 per ton. That equals 907 liters of water per day per ton or 2700. Since you see 10% of the iceberg is the quote lets say that is 1000 liters per ton to allow for ice expansion.. Say you needed 3 tons of cooling so that is 3000 liters per day. That is 1500 2 liter pop bottles per day. Need AC for 30 days now you are looking 45,000 bottles. Then your icehouse needs to keep them cool for that time period. Say your loss rate there over the season is 90%. If you did it all in 2 liter pop bottles that is 450,000 bottles.

For comparison sake the AC units on typical equipment like tractors and combines are usually rated at 6 to 8 tons. So a 3 ton unit is actually very small. Ice houses work because of their small losses. With AC your losses are high and the numbers can't be supported unless you can build huge piles of ice.

There are answers for smaller scale things like refrigeration but swamp coolers, insulation, geothermal and thermal mass are the only real and fairly cheap answers to AC. If you have reliable water power you might look into using a mobile AC compressor to power a stationary AC system. Be aware you need a lot of water or fall height for this to work.

Just something I started when the power was out for a week a few years ago, and a couple of years before that. I have always been extremely careful of water use and long ago turned off the water inlet to the toilet. Shower water is scooped into buckets shortly after showering. Water is used for toilet flushing, watering trees/bushes/compost,etc. So I started to think about how to heat only what I need for dishes, my own shower, and will closed bottles of water retain household heat a bit in the winter or cool the house down a tiny bit in the summer? I understand that several gallon jugs is not much in the spectrum of the quantity of air in the house but having a room temp supply of water to wash dishes, clean myself, or use for cooking seems to be better than getting cold water from the tap and heating it from say 50 degrees F. Am I saving anything over several years? At least when the power is out I have an emergency supply for a few days.

Even before learning about Paul's lightbulb and heat the workspace not the room space podcast and talks, I have used small refrigerated water bottle or frozen icepack under my feet to cool down on a hot day. I thought the two liter air conditioning would be a two liter bottle with cold water/ice so individuals could cool their bodies, not the rooms!
What a savings of energy that could be! How many people could be cooled by a freezer full of two liter bottles!

A PS on this. If you intend to freeze large volumes of water there are better techniques than just pouring the water out and letting it freeze. Large scale ice house that froze the water in place used fogger nozzles to make fine droplets for maximum surface area to promote heat transfer from the air and therefore freezing. This also gives surface area for evaporation for added cooling. This also gives heat transfer to the air making warmer air. So a riser chimney would likely help fan the air through the system.

At UC Davis, they've been playing with solar refrigeration for years. You could build a similar off-grid refrigeration unit to make ice to keep your food cold, or rigged to cool the air in your abode.

Page 27 of this PDF states: "The SAAR is a small scale, sustainable solar refrigerator system that can operate in tempered regions of the world and be duplicated using common materials found in developing communities."

Powerful. Sustainable. Clean.

Find it here: http://www.solaripedia.com/files/1113.pdf

The Visitor Center at Zion National Park in Utah has the best passive cooling system I've ever seen. It is desert, so the air is dry. They have a cooling tower, and gravity flow water wets fins. So, think of a chimney with wet insides. As the air cools it falls, cooling more and drawing more air into the down chimney. The cold air that flows out over cools the large public space, they have to partially block the flow. It puts out way more cool than a "swamp cooler". For people who live in arid climates, I recommend looking into it!

C. Letellier wrote:I am going to throw a monkey wrench in this idea for AC use. Air conditioners can be rated in tons. That is tons of ice melted in a typical 24 hour period to produce the cooling. A SMALL residential cooling unit will be rated at 3 tons. That means the heat transfer it produces is equivalent to melting 3 tons of ice over the course of a day. 2000 lbs / 62.4 lbs/ft^3 = 32 ft^3 per ton. That equals 907 liters of water per day per ton or 2700. Since you see 10% of the iceberg is the quote lets say that is 1000 liters per ton to allow for ice expansion.. Say you needed 3 tons of cooling so that is 3000 liters per day. That is 1500 2 liter pop bottles per day. Need AC for 30 days now you are looking 45,000 bottles. Then your ice house needs to keep them cool for that time period. Say your loss rate there over the season is 90%. If you did it all in 2 liter pop bottles that is 450,000 bottles.

While a lot of water would still be needed, I think you may be over estimating the amount of ice needed because air conditioners are rated by their output, but don't actually run 24/7. For example, a ton of cooling capacity is equal to 12,000 BTU per hour, so a 3 ton unit would be 36,000 BTU per hour. Melting a pound of ice absorbs 143 BTU, so that works out to the equivalent of melting 250 pounds an hour (56 Two-liter bottles an hour). If the formerly frozen water is brought up to 60F, you would get another 28 BTU per pound, which would bring the total number two-liter bottles down to about 48/Hr. Yesterday in Indiana it was hot (high over 90) and muggy. The total run time on my AC was about 4 hours to keep my house comfortably in the 70s, so that would have been ~1000 pounds of ice (roughly half what can fit in an IBC tote). Put another way, to get 100 days of equivalent cooling, I would need ~50 frozen IBC totes.

Of course, there will be losses due to melt. 100 years ago, ice houses lost 20-50% of the ice between harvest and delivery. Since there would be no delivery, and our insulation has improved a bit, a reasonable guess would be a 10% loss due to melt. Adding in another 5 IBC totes for melt loss, the system still seems doable...probably several times more expensive (at least up front) than a conventional system, but doable so long as we a using something much larger than a two-liter bottle.

John Wolfram wrote:

C. Letellier wrote:I am going to throw a monkey wrench in this idea for AC use.   Air conditioners can be rated in tons.  That is tons of ice melted in a typical 24 hour period to produce the cooling.  A SMALL residential cooling unit will be rated at 3 tons.  That means the heat transfer it produces is equivalent to melting 3 tons of ice over the course of a day.  2000 lbs / 62.4 lbs/ft^3 = 32 ft^3 per ton.  That equals 907 liters of water per day per ton or 2700.  Since you see 10% of the iceberg is the quote lets say that is 1000 liters per ton to allow for ice expansion.. Say you needed 3 tons of cooling so that is 3000 liters per day.  That is 1500 2 liter pop bottles per day.  Need AC for 30 days now you are looking 45,000 bottles.  Then your ice house needs to keep them cool for that time period.  Say your loss rate there over the season is 90%.  If you did it all in 2 liter pop bottles that is 450,000 bottles.

While a lot of water would still be needed, I think you may be over estimating the amount of ice needed because air conditioners are rated by their output, but don't actually run 24/7. For example, a ton of cooling capacity is equal to 12,000 BTU per hour, so a 3 ton unit would be 36,000 BTU per hour. Melting a pound of ice absorbs 143 BTU, so that works out to the equivalent of melting 250 pounds an hour (56 Two-liter bottles an hour). If the formerly frozen water is brought up to 60F, you would get another 28 BTU per pound, which would bring the total number two-liter bottles down to about 48/Hr. Yesterday in Indiana it was hot (high over 90) and muggy. The total run time on my AC was about 4 hours to keep my house comfortably in the 70s, so that would have been ~1000 pounds of ice (roughly half what can fit in an IBC tote). Put another way, to get 100 days of equivalent cooling, I would need ~50 frozen IBC totes.

Of course, there will be losses due to melt. 100 years ago, ice houses lost 20-50% of the ice between harvest and delivery. Since there would be no delivery, and our insulation has improved a bit, a reasonable guess would be a 10% loss due to melt. Adding in another 5 IBC totes for melt loss, the system still seems doable...probably several times more expensive (at least up front) than a conventional system, but doable so long as we a using something much larger than a two-liter bottle.

I may be over estimating a bit but probably not as much as you think.  3 ton / 12,000 BTU is rated to cool 500 to 550 square feet with decent insulation.  Many homes need way more AC because of size or because of poorer insulation and then there is climate.  And I only figured 30 days run time which compensates for the parts of the year where it runs less somewhat.  Lets take your numbers for a bit and round the 48 per day up to 50.  Times 4 hours per day gives 200 bottles per day and over 100 days that is 20,000 bottles.  Needing a bit more cooling for harsher climate, larger home or poorer insulation could easily push those up a bit.   So I will say 45,000 bottles is likely not off as much as you think.(either way I think we are of the right order of magnitude as to number of bottles to this point)

  As for the ice house my numbers there came from a small farm sized ice house.  That is something that economies of scale apply to in serious form.  The site you are looking at is pointing to numbers for large commercial sized outfits.  Every time you double the size of an icehouse you increase the amount of ice you can store by a factor of 8 while increasing surface area by a factor of 4.  Since surface area is proportional to heat loss/gain this is why bigger eventually gets big enough to work.  I agree modern insulation is better and we can probably do better that way.  But we are also talking a relatively small ice house which means higher losses.  90% may be off.  I don't have good numbers to base that on.  I know smaller farm type ice houses mostly didn't used to make the year.  The number I did have was from a well research historical novel on an estate on the east coast that said their losses for the year were 90%   While I may easily be high I am guessing 10% is low.  Without building one to see I don't know.  The best other info I had was an article on a farm that had a root cellar that the froze the ground around it, filled with ice and built a straw stack on top of for insulation late winter.  Most years they made it back to semi freezing weather on the other end but there was almost no reserve and that was built with modern insulation.  Any way you look at it you are talking about a lot of ice.

C. Letellier wrote:I may be over estimating a bit but probably not as much as you think.  3 ton / 12,000 BTU is rated to cool 500 to 550 square feet with decent insulation.  Many homes need way more AC because of size or because of poorer insulation and then there is climate...So I will say 45,000 bottles is likely not off as much as you think.(either way I think we are of the right order of magnitude as to number of bottles to this point).

I totally agree that moving around 2-liter bottles is a no-go, and in my example 50 IBC totes is equivalent to 26,000 two-liter bottles...so we aren't that far off on the numbers. If someone were to actually build this type of system with buried/insulated IBC totes, my guess would be that they would way oversize the number of totes because the cost of an IBC tote would be trivial compared to the insulation/plumbing/excavation costs.

Broken chest freezers can usually be had for the fuel and labor required to retrieve them, have fairly significant capacity and are fairly well insulated and water-tight by default. [Be sure to check for leaks before transport.]

In the right climate, you can install one of these in an unused room on the shade side of the house and use the cold of the winter nights to fill the whole thing with ice.

You want to install some form of drainage under the freezer. After you've filled it with ice you will want to open the drain plug to allow the meltwater to drain away.

Open windows and thermally isolate the room from the rest of the house and freeze a few inches per night [closing the freezer and throwing a few layers of blankets over it between freezing events] until full.

How you feed the heat of the house through that room to be absorbed by the ice during the summer months is a decision that varies depending on home design and available resources.

Continental climates may have an easier time filling freezers with ice but may need more than one of them to keep a home cool during the summer.

Does this work effectively?

The weather here in Southern Arizona is getting warm.  I used to cool my house with an evaporative cooler, but last year I installed a 12,000 btu (1 ton) mini-split heat pump.  I mostly wanted it for heating during the winter, but I figured I'd see how well it works for cooling.  I have a ~1500 sqf house that is pretty well insulated, not passive hause standards, but pretty good.
The daytime temps have been getting into the high 80s low 90s F  (+/- 32 C) recently.  

Anyway, it's been using just under 5 kWh a day to keep the inside of the house down around 74-75.  It runs constantly about 18 hours a day, but it has an inverter compressor so it can vary the speed of the compressor and therefor adjust how hard it works.
According to the specs it uses around 960 watts when providing approx 12,000 btus of cooling, but it can drop as low as about 200 watts (3500 btus) or surge up to 1300 watts (17,300 btus).
The highest I've seen it go so far is about 550 watts, most of the time it's using less than 300 watts.

Anyway, I figure it's providing the equivelent of 400-500 lbs of ice per day with temps around 90 in a small-medium size house.  When the temps climb into the high 90s, I'll bet I use the equivelent of 1,000 lbs or so of ice per day.

We usually use some form of cooling for about 5 months per year (beginning of May to end of September).  If we assume 600 lbs of ice per day, that works out to around 90,000 lbs of ice (45 tons), so not practical for me.

Whether or not it's practical for others would depend on their locations, etc.

Still, considering you could buy the heat-pump and enough solar panels and inverters to run it for around $1500,  I doubt you could build an effective ice-house plus the rest of the system for that price.  

And besides, the heat pump can be used to provide heat during the winter too, hard to do that with an ice-house.