# Is there a way to make an air conditioner without venting heat outdoors?

Is there a way to make an air conditioner for when you can’t vent the heat outdoors?

Can you trap the heat in an object or convert it to another form of energy?

For example, could you transfer the heat through an electrical outlet?

I would be looking for replacing an in-window air conditioner for a room rather than providing AC for a house.

My situation is that I can't vent into ground or attic or window. But I appreciate learning if that possibility opens up.

• This is like trying to bail out a boat without throwing the water overboard.
– J...
Sep 26 '21 at 12:22
• Related: Are mobile air conditioning systems an inherently flawed design? on diy.SE discusses A/C units with a hose or two hoses that go out a window, instead of having to sit right in a window. (And the efficiency challenge of having the compressor motor inside the space you're trying to cool.) Sep 26 '21 at 13:33
• Related: "Waste heat recovery unit", Wikipedia.
– Nat
Sep 26 '21 at 17:15
– hazzey
Sep 28 '21 at 12:33
• @J... So, by boiling the water to steam, or electrolyzing it, or binding it in silica? Sep 28 '21 at 12:57

The possibilities of transforming heat into any other form of energy are limited by the second law of thermodynamics. Essentially, you cannot simultaneously lower the entropy in a closed system (e.g your house) by transforming heat into a different form of energy, and transferring heat from a colder object to a hotter one requires work, so that more heat is produced in the second object than eliminated from the first. This (heat transfer from a colder body to a hotter one) is exactly what an air conditioner does.

Dumping heat into an electrical outlet while cooling your house is what a perpetual motion machine of the second kind would do.

The ceiling fan doesn't cool down anything, it only produces additional heat (though not much). It gives you a cold sensation is because it promotes air circulation, which improves heat/water exchange with your skin. If the air is colder than 36°C or below 100% humidity, it will carry heat away. If the air is hot and humid, the fan will actually make you feel hotter.

• This isn't strictly correct. What about using a chemical reaction that's endothermic? Not practical, but it does reduce the temperature of a sealed space.
– Drew
Sep 26 '21 at 23:19
• @Drew The OP could also subject their house to adiabatic expansion or put a black hole in it :) Sep 27 '21 at 8:33
• I disagree with the last sentence of your answer. Converting heat to electrical energy is exactly what a peltier element does. And it is by no means a perpetual motion machine. Converting heat to electrical power is also, by definition, what a thermal power plant does. Of course, both of these examples are of no use to OP, and what he asks is next to impossible. But I still think you rephrase (or remove) that last sentence of your answer because it is incorrect. Sep 27 '21 at 14:55
• @Opifex but doesn't a Peltier device work by moving the heat back across the divide? So now you're back where you started. Sep 27 '21 at 15:21
• @Opifex A Peltier element which produces current also produces heat on the cold side. A Peltier element which cools the cold side consumes current. You can't have both current generation and cooling. Sep 27 '21 at 18:33

Watercooled , if you don't count sewer as "outdoors".

• So you transfer heat into water or ceramics or both (like a coffee mug) or down the drain? Sep 25 '21 at 14:54
• I like the simplicity of this answer but the maths in my answer suggests you would be wasting hundreds or thousands of litters every day, which isn't economically nor ecologically sound in most places.
– Pere
Sep 25 '21 at 22:13
• I look ed at doing something like this, using a closed coolant loop and a heat exchanged/radiator at each end. The hot end would have been sunk in a stream that ran through the property. The maths worked out, but the parts cost was high, and payback window was many years. And it also turned out to be illegal in my area, without an expensive Resource Consent which would have make it all uneconomical. Sep 25 '21 at 22:41
• @Mark I am not an expert, but I suspect that if you can run a pipe to a river you can figure a way to vent outdoors for much less - so piping to a river would be wasteful. Sep 26 '21 at 2:17
• @Mark - You can say "use" if you want, but in a lot of parts of the world water is a more or less scarce resource that shouldn't be used and dumped in large amounts when avoidable. Furthermore, if the only available water source is tap water (quite usual), that would be expenxive.
– Pere
Sep 26 '21 at 7:44

Use a (largish) water tank.

I've never seen a water tank used as a heat sink for an air conditioner, but they were used for stationary engines some time ago.

Typical air conditioners can produce about 2 kW of heat. If that heat is released in a 500 litter water tank:

$$\frac {2 kW}{500 kg}\cdot\frac{1 kJ/s}{1 kW}\cdot\frac{0.24 kJ}{1 kcal}\cdot\frac{0.24 K\cdot kg}{1 kcal}\cdot\frac{3600 s}{1h}=\frac{2\cdot 0.24\cdot 3600}{500}K/h=3.456 K/h$$

If my maths are right, a 500 litter tank temperature will increase less than 5 ºC per hour when the air conditioner is working at full range. That makes for several hours before it becomes too warm to be a problem. If it isn't enough you just need to make the water tank larger.

Of course, you end with a large tank of warm to hot water inside your home and you need to move the heat and water outside. The obvious solution is to place the water tank outside, but other ways might fit better your circumstances - for example, taking the tank outside every evening using a pallet jack.

Another answer suggests using water and discharging it into the drain, which is quite equivalent to emptying the tank, but that would waste an insane amount of water - unless there is plenty of free water at your place.

And as an end note, I would forget about extracting energy from the stored heat unless you have a practical use for warm water, but you may enjoy the pleasure of a warm bath after a long day working under the air conditioner.

• Ideally you could use this to feed a hot water heater, so you have a natural way (showers, dishes) to use up heated water and make room for new water to enter the system from the city water supply. (I'm picturing two tanks: one at near-ambient temp that your heat pump can pump into efficiently, and a separate normal hot water heater tank (or tankless natural gas) that pulls from that.) Sep 26 '21 at 13:30
• Sounds like a good way to heat a hottub. Oct 14 '21 at 20:18

This question reminds of the concept of a heat bank, a form of heating I experienced in the 1970s. The type of heat bank I am familiar with consisted of a metal box which contained an electric radiator and large quantities of steel which were heated by the radiator.

The concept was the radiator was energized during the night when cheaper off peak electricity rates applied and heat was stored in the steel plates. During the day, the electric radiator was turned off and heat from the metal plates radiated into the room with the heat bank.

The waste heat from an air conditioning unit could be stored like this, but the amount of heat produced by an air conditioner would be small by comparison. The other question that arises is what is then done with the stored heat. In a hot climate it would not be easily used.

• These are also known as "night store heaters/radiators" and may have ceramic/thermal bricks inside instead of steel plates. Sep 25 '21 at 22:38
• The other problem is that as the heat bank heats up, the refrigeration cycle gets less efficient. Sep 26 '21 at 1:32
• @Criggie - we call these things night storage heaters in the UK - they take advantage of a cheaper electricity tariff (one-third of daytime rate) offered by suppliers between certain hours typically midnight until 07.00 in winter, and from 01.00 to 08.00 in summer. Older ones used to be a pain if the weather during a day is unexpectedly sunny; you have to open lots of windows to avoid being stifled. Modern models are more controllable. Sep 26 '21 at 11:02
• What about a heat safe? Likes reverse refrigerator that taps heat inside of it. At some point you could display it out let out cool down when the device is turned off? Sep 27 '21 at 7:02
• The trick to modern night storage heaters (@MichaelHarvey) is insulation/sealing to keep the heat in unless the vents are open. Old ones had crude vents on cruder controls Sep 27 '21 at 15:01

Nobody specifically mentioned geothermal yet but this is exactly the principal used here.

Rather than exchanging heat with the outdoor air, heat is exchanged with the ground a meter or two below the surface.

I suppose one could make the argument that this is still "outside" but the heat does not make it to the atmosphere and if you have a heating/cooling geothermal unit, the heat that is stored in the ground during the summer cooling season will be used when heating is needed in cooler weather.

I have never seen a geothermal window unit but it would be possible in theory.

• Would having a heated metal rod that went right into the ground inside the house to the same thing? Since heat would travel along the route the ground would then dissipate the heat. It might cook some ants though or warm their home. Sep 27 '21 at 20:17
• That would indeed transfer some heat into the ground. It would be terribly inefficient though. Sep 27 '21 at 20:36
• You need a lot of contact between the cooling loop and the ground. Either that means a long trench or deep-drilled holes when you don't have the space for a long enough trench. Such units use less power than traditional air conditioners, but the cost of burying the cooling loop generally makes it uneconomic. Sep 27 '21 at 22:41

Surprised they haven’t been mentioned yet because they seem to be all the rage in the spammy ads you see on a lot of sites, but depending (very heavily) on the general atmospheric conditions, an evaporative cooler (sometimes called a swamp cooler) might work.

Evaporative cooling in general is based around the idea of latent heat of vaporization. In short, evaporation requires energy to happen, and that energy is almost always in the form of heat absorbed from the surrounding atmosphere. The actual amount of energy required is inversely correlated with how readily the liquid evaporates, and water actually takes a pretty significant amount of energy (roughly 2.25 kJ/g), so there’s a non-trivial cooling effect on the surrounding air just from water evaporating.

An evaporative cooler takes advantage of this by utilizing a reservoir of water and some sort of apparatus to provide a high surface area for the evaporation to happen (most comercial designs use specialized wicks), and often some mechanism to increase airflow as well. Under ideal conditions, a good evaporative cooler can easily translate to a cooling effect of multiple degrees celsius.

The big downside to an evaporative cooler is that ‘ideal conditions’ part. For it to work well, you need to be in a relatively dry climate. In the continental US for example, they work great in most parts of Arizona, Nevada, Utah and New Mexico, but are kind of useless in places like Florida, Georgia, or Louisiana due to the relatively high average humidity. You also have to keep it topped up with water, and generally have to make a point to keep it clean (it’s stagnant water, so you will get microbial growth if you don’t clean it regularly, and it will tend to smell rather bad).

• If your heatpump is OK with 100C cooling side, you can just use boiling water, no matter of climate. Sep 26 '21 at 6:07
• If you do this purely indoors, without venting the evaporated water, it will re-condense inside the room and dump its heat of vaporization back into whatever cool surface gets wet. Or just bring the room air's humidity up to equilibrium if there's no heat-pump cooling the room and heating water. Sep 26 '21 at 13:23
• I updated my answer with a device I bought a few months ago. I don't have conclusive evidence it works. Although I believe it works if I would freeze one part of the component. Sep 27 '21 at 20:19
• @PeterCordes While true, that querent's question wasn't "how do I get oxygen into my hermitically sealed house" suggests that he's got normal gas exchange going, and just doesn't want to run new pipes. :P Sep 28 '21 at 13:02
• @PeterCordes It really depends on the outdoor absolute humidity (unless you’re actively running a humidifier or dehumidifier, indoor absolute humidity tends to track outdoor absolute humidity pretty closely) and how much gas exchange there is in and out of the building. A ‘normal’ house actually has a nontrivial amount of gas exchange (hence why humidity tends to track so closely), which can be enough to make a swamp cooler function well. The scam aspect is people selling them as cooling devices in humid climates where they don’t work at all. Sep 28 '21 at 19:19

There are phase change materials (Similar to the cold pack for injuries) that at some point in time was the all the rage. Basically what they did was storing and releasing heat energy. However, they have a limited capacity.

Figure: Phase change material in the form of a cold pack The idea is (or was because it never caught on) that the phase change materials are introduced into the walls/structural elements of a house, and they passively collect and release temperature around the set temperature.

Figure 2: how a phase change material works (source: steemkr.com)

Regarding the problem at hand "Is there a way to make an air conditioner for when you can’t vent the heat outdoors?", IMHO the answer is no - at least for applications which are the size of the house, and for large periods of time.

Update : regarding whether the heat can be converted to electrical energy and released to the environment

The problem is that you need to convert the heat energy to another form. To do that you need some kind of thermal machine. The efficiency of thermal machines is bound theoretically from the Carnot efficiency which is depended on the ambient temperature and the medium.

Since you are talking about a house, I'll assume a temperature of 23 °C as the desired and 33 °C(ambient) then the carnot efficiency is at best:

$$n_{th} = 1 - \frac{273+ 23}{273+ 33}= 3.27\%$$

So only a small portion of the energy can be converted and the rest will remain heat.

• I updated to account for you comment. I think you should put that into the question, if you want an answer on that specifically. Sep 25 '21 at 7:35
• There's not a lot of well-evidenced data out there on the size and growth rate of the market for phase change materials, but this peer-reviewed paper and this commercial report suggest that 'never caught on' is not entirely fair. Sep 27 '21 at 12:58
• @danielhatton I shouldn't have made this personal assessment. Although the scarcity of growth or market size data is not a good omen. Sep 27 '21 at 13:12
• Carnot cycle only works with a real temperature difference. It will not work with desired temperature, otherwise you could decide you desire to reach absolute zero, and suddently you can transform heat into work with 100% efficiency. Sep 27 '21 at 18:43
• @DmitryGrigoryev a fair point that the house would indeed heat up by the thermal machine (by a little), however I was addressing a comment (that is now deleted) that asked about converting the heat into electricity. In that context I used the thermal machine and the maximum efficiency of the Carnot cycle. Sep 28 '21 at 10:16

Heat can be converted to electricity only according to the second law of thermodynamics: you need in this case a hot reservoir and a cold reservoir. The heat engine takes heat from the hot reservoir and puts it into the cold reservoir. That's the way all heat-based electricity generation works. In gas-fired generation, the hot reservoir is very hot (it's actually inside the turbine/engine and can be several thousand Celcius) and you get a good efficiency of 50% if you put a steam turbine after the gas turbine (combined cycle generation). In coal power, it's 600 degrees Celcius and you get an efficiency of 35-40%. In light water nuclear reactors, the temperature is 300 degrees Celcius and you get an efficiency of only 30%. These efficiency figures assume the heat can be rejected to a cold reservoir of 0-30 degrees Celsius, which is usually possible. A very hot day on a very hot climate could reduce the efficiency somewhat.

If you want to convert heat inside your house to electrical energy, you need two conditions to be true:

• The outdoors needs to be colder than the indoors
• You need to vent heat outdoors (exactly what you wanted to avoid)

In this case, typical temperature differences are below 10 degrees Celsius compared to 300-1000 degrees Celsius as used in commercial energy generation. So the efficiency would be so poor it's not worth pursuing. A fan is a better and cheaper option. Although a fan uses little bit of energy, it's far more efficient in directing the heat outside the house. Natural circulation would be horribly slow. An electricity generator operating on the temperature difference would probably be even slower.

If the outdoors is hotter than the indoors, then you need active energy to vent the heat outdoors. That's the electricity input to your air conditioner. Sometimes, even if the outdoors is colder than the indoors but not by a very large margin, you will want to use the air conditioner anyway as it's faster to cool than a fan.

You could in theory trap the heat to ice, but then you need a constant supply of ice, and you need to somehow get rid of the melt water, possibly through the drain, which would effectively be your way of venting heat outside.

However, it could be theoretically the case that disposing of heat by directly venting it outdoors as hot air is impractical but disposing of melt water not so impractical. Unfortunately, getting all that ice will be difficult. Ice was used to cool houses before the advent of air conditioning, and air conditioning replaced ice due to being far more practical.

Most houses have forced air ventilation. It might be feasible to get rid of the rejected heat through that ventilation system.

When situated in the old industrial building, once we had a "cooling system" that was simply sucking in the air through the hole in the floor from the accessible there labyrinth of the industrial underground tunnels. I do not know how productive this would be in a large scale but worked well enough for a small server room, using electricity only to drive the fan.

Edit 1 based on the current version of the question "Where can heat be vented from an air conditioner that isn't outside, in the attic, or into the ground?"

Answer: It is possible to put the heat into something like a large water tank. The problem is that you have to take the heat out of the water tank eventually. This is the heat battery concept. In a practical sense, when your available resources is that you have a room you can't even fit a regular air conditioner to, you're just going to be hot.

You can also try to use an evaporative cooler. These are sometimes called swamp coolers. Here's the issue: These only work in very dry environments. In most of the USA (assumed from your use of degrees F), using one of these will just turn your room into a sauna.

New question 2: How does a ceiling fan fit? A ceiling fan, being an electrical device, will add some amount of heat to the room. This is probably insignificant compared to the environment and other loads. What a ceiling fan does is makes it feel cooler in the room because moving air cools you down, at least as long as you can sweat. There are edge conditions in the southwestern US where blowing very hot air across your skin actually heats you up, but in most places, blowing a fan at you makes you feel cooler.

Let's examine what we're trying to do.

An air conditioner is a device that pumps heat from cold to hot, getting by the whole heat-flows-only-from-from-hot-to-cold thing. The cool thing about it (no pun intended) is that you can move more heat than the energy you put in to the device.

Now, let's discuss what the heat is we're moving. Heat is typically divided into latent (humidity) and sensible (temperature) heat. Your heat in your room comes from several sources:

• environment sensible heat (it's hot outside!)
• environment latent heat (humid air infiltration)
• civilization-induced sensible heat. Pretty much every amp of electricity that goes into your house gets turned into heat. The biggest in America tends to be the refrigerator. But every single thing that uses electricity turns it into heat.
• biological-induced sensible and latent heat. Humans, dogs, and cats are hotter than a cooled space, and thus produce heat. Their breath and sweat also produce latent heat.

So, what can we do with this heat we are removing? Practically, almost nothing. The expense to turn that low-temperature heat into work that you or someone else could use is more than it's worth. Other answers note the idea of heat batteries. If you happen to live in one of the places where it is hot during the day and cold at night you could store the heat and let it out at night. The thing is, it's probably easier to just let the hot daytime heat up that battery anyway (desert communities have done this through building materials for millennia). Adding complexity to electrical/mechanical systems costs more for minor paybacks.

We could do things for fun, like Peltier devices that make enough electricity to maybe charge a phone, or Stirling engines for the fun of it, but these aren't scalable solutions.

So, in summary, just enjoy the cool and worry about other things.

Theoretically? Sure. From a practical standpoint? No.
At least, it's pretty hard to see that working in any feasible way. Regardless of what you do with the waste heat, you would need an inexhaustible supply of fresh air to cool the hot refrigerant. So the question is what to do with that hot air. As someone suggested, you could cool that with, say, water cooling by creating a 2nd heat exchanger in your air exhaust. Provided the back pressure was low enough that the performance of the A/C was not impacted, which is doubtful. So you would still be exhausting air, even if cool air, except be paying a water bill for constant water circulation. So if you're talking about cooling a real room, it's hard to see it being feasible without simply exhausting hot air like normal.

I bought this in a hardware store a few months ago and tried to use it in my car bc my AC would take 20 mins to cool the interior (98' F weather).

The air was sortof cool but it seemed like it wasn't any cooler than what I had in my car and it said it works better when you freeze some of the components.

It sounds like it is using ice to "store cold" and you're sinking heat into it during the melting phase.

I pulled it out of storage last night and it didn't seem much cooler in the room.

But it was worth it for the coloured LEDs? (They should have been a clue.)

A friend once said that if I don't have an AC unit the ceiling fan would generate more heat than it would be at cooling me down. I know heat rises so it makes sense but is this statement true or false?

Bringing in electrical electrical energy to run the fan will result in heating the room by a small amount. Fans do not change the temperature of the air. They provide increased cooling to skin by increasing the airflow across the skin which in turn increases the heat transfer rate from skin to the air. But cooling can only happen if the air temperature is less than body temperature (37°C) by at least a few degrees.

• @1.21gigawatts, I rejected your edit because I wasn't sure what you were trying to say. "Actually, this is only true for cases where the relative humidity is above 60% or so, which it often is with air conditioning. If the humidity is 45-50% (or lower), the skin temperature of blowing dry air may cool the skin by as much as 20 degrees, well below ambient." Blowing dry air at what temperature? Oct 14 '21 at 19:26

I have a water cooled air conditioner. It takes cool water in and returns hot water to the floor drain, cooling the room next door.

I'm surprised nobody has mentioned infrared radiation. Supposedly somebody has invented a paint that reflects so much energy that the self-radiation of its surface is greater than the incoming radiation, and it cools down below ambient temperature.

https://www.purdue.edu/newsroom/releases/2021/Q2/the-whitest-paint-is-here-and-its-the-coolest.-literally..html

Heat's gotta go somewhere? One idea would be to deposit underground in geothermal coils.........