This is bugging me to the point that it almost enrages me.

A split-unit air conditioner would typically cause condensate to form in the indoor unit. This water can be simply drained with gravity having a pipe running to a sewer.

However in most cases there is no nearby sewerage pipe or the installation would be too much of a hassle. Usually the water is pumped all the way back to the outdoor unit (because there are already pipes going through walls that way) where it is just dumped on the roof, in a gutter or elsewhere (sometimes even causing problems)...

I don't get it... the manufacturers are not oblivion to the cooling power of evaporating liquid (the whole AC device works on that principle) so why don't they add a contraption to sprinkle the radiator with the condensed water (like dripping it onto a slanted surface for instance)? The radiator is already weather proof since it should be able to handle rain and snow blowing in and it would also get wet with condensation when working in reverse (heating) mode. Condensed water will not build up calcium or lime scales as opposed to tap water (unless one suffers from severe indoor dust storms).

Example of a typical ductless AC installation

Corrosion of the radiator may be a concern, but when used in reverse as a heat-pump, the radiator would be wet any way (in the wetter parts of the year when not freezing) water will condensate on the outside radiator. That is a reason why the radiator should not be designed in a way that it would hold water in order to evaporate it. When used in reverse as a heat-pump it would be ideal to let that water run off before having a chance to evaporate. In this mode there will be no water coming from the drain so it would not interfere.

Here's a schematic illustration of a most simple implementation:

Schematic illustration of a simple setup

The rate at which condense is supplied may be greater than what can be evaporated this way (between 5 and 20 gallons a day).

An active way to evaporate the water in a controlled way is to run the hot gas trough a flattened pipe that is cooled by sandwiched Peltier elements that bring the water up to boiling point and dramatically cool the gas with a comparatively small amount of energy (using Peltier elements without the evaporating water would not be efficient at all, but boiling water should extract a huge amount of heat).

Here's a rough illustration of the idea:

Illustration how to utilize condensate2

I cannot find much info about how hot the refrigerant is at the start of the radiator, but I think a TEC2-19004 should be able to get it to boil. It can handle 4 Amps at 16 Volts so that would be 64Watt max to get a difference of ≥80°C according to their claims. Running at max capacity is not most efficient but should not be needed, we don't want the refrigerant to freeze in the radiator either...

However we may need a few of these to consume all the water. The water will not start to boil until the whole body of water has reached boiling point. Without a hot radiator it takes about 2400kW to evaporate 2.4 liters per hour in an even more wasteful evaporation pan. In this use-case a Peltier solution could be allot more efficient than most of the other inefficient ways that Peltier solutions are usually known for.

The way condensate currently just gets dumped right under the radiator baffles me, this idea seems so obvious, I cannot believe that I'm the first to come up with it. So my question is why doesn't any AC unit on the market have a built-in option to use drained condensate for cooling by re-evaporation?

  • 1
    I'm guessing the reason is that the efficiency benefits don't outweigh the maintenance and cost concerns (i.e. for the amount of efficiency you'd gain, it would be cheaper/easier to just spec a more efficient unit). But I'll be interested to see what answers come.
    – LShaver
    Commented Jun 4, 2022 at 17:35
  • I have come up with a potential issue with sudden temperature changes when the first cold water drops hit the hot radiator, it could cause the metals to wear faster due to expansion and crimping. But how much difference would it case during the life span? I still think the energy savings will far outweigh the material damage. Commented Jun 8, 2022 at 7:17
  • Anything involving a Peltier is going to be really inefficient. Commented Jun 9, 2022 at 8:48

2 Answers 2


Yes, indeed, all AC units have condensing water on the evaporator. One way to get rid of it is simply a pipe where it is discarded to a sewer or discarded outdoors.

In units that are not designed to be permanently installed but rather moved on wheels, where the single unit contains both the evaporator and condenser in the same unit (unlike split units where evaporator and condenser are different), there is typically a system that lets the condense water from cold evaporator to drip to the hot condenser.

The way this works is that these mobile units on wheels have the evaporator on top of the condenser so the dripping part is easy. However, this is not enough. The water doesn't evaporate fast enough, so it may drip to the bottom of the condenser. In this case, the condenser needs a large water container in the bottom of it, and a pump to re-raise the water back to the evaporator above the condenser, where it may start the dripping and evaporating process again.

Furthermore, in some climates where the weather is very hot and humid, it is possible the condenser is unable to evaporate all condensing water and the water container needs to be manually emptied. So there needs to be means to get rid of the water in cases where the condenser can't get rid of it.

I suspect the reason why split AC units don't do the same is caused by these:

  • You need a pump at the condenser to re-raise water from bottom of condenser to top of condenser give water multiple chances at evaporating. The pump adds cost and complexity.
  • You need means to get rid of the water in cases where it accumulates faster than it evaporates. If getting rid of it doesn't mean dripping it on the ground, there needs to be some means of containing the water and letting the user empty the accumulated water. If the condenser is installed for example in an apartment building, how do you empty the container? You would need ladders potentially ten meters long or something similar to be able to reach the condenser. Imagine doing this every day! Of course it might be possible to install Yet Another Indoor Unit that absolutely has to be installed lower than the condenser and would contain just a pipe ending at a shut-off valve that would allow draining the water from indoors. Of course this would require Yet Another Hole At The Wall...
  • If the condenser is installed higher than the evaporator, the single pump at the condenser isn't enough, you need a second pump at the evaporator to be able to raise the water higher than at the evaporator

So the expense of a pump or two and the inconvenience of draining the container if water doesn't evaporate fast enough isn't probably worth it.

It's true that the temperature at the condenser is critical to AC system efficiency. There are two ways it can be improved:

  • Make the fan at the condenser rotate faster to cool down the condenser
  • Evaporate the condensed water to cool down the condenser

The manufacturers probably have though that if condensed water can be gotten easily and cheaply rid of by dripping it to the ground, the expense of potentially two pumps and the inconvenience of getting rid of the too fast accumulating water in some climates isn't worth it to slightly improve the system efficiency, because the manufacturer can improve system efficiency by making the fan rotate a bit faster, or to install a better (larger) fan.

The self-contained AC units on wheels that can't get rid of water via a pipe because the entire unit is contained indoors and hot air is blown out of a window using a duct, do evaporating not because it greatly improves system efficiency (it improves efficiency only slightly), but rather because that's the only option they have. They have to have a pump and means of shutting down the device when water accumulates too much (to let the user solve the situation).

Also I wouldn't be surprised to find out if the evaporation of water in self-contained AC units only works because they're so terribly inefficient in cooling the hot condenser (they do it via a duct). Split AC units have a very very large fan at the condenser. Spit AC units are also more efficient. I suspect the very very large fan can do a far better job at cooling the condenser, improving system efficiency. I'm not sure if the well-cooled condenser is hot enough to actually evaporate the water. If not, then it's not actually true that evaporation of water improves system efficiency. To evaporate, you need hot enough condenser and that alone may mean the system efficiency has to be low.

  • For indoor units I understand, they usually have a bucket with a float switch. However These could also use Peltier elements to boil it away into the hot air duct. I added a photo of what I have often seen here in the Netherlands, these units are often labeled as A+++ for energy efficiency, but could raise the bar allot more with a minor tweak. I wouldn't give the water a 2nd chance since it will not get much cleaner. This company even has heated pans to evaporate the water. (scroll down to the bottom). Commented Jun 6, 2022 at 21:02

I considered a water drip on my outside condenser but the possibility of corrosion was likely to more than offset any efficiency gain. Yes, the outside unit gets rained on occasionally but that is not the same as usually wet.

  • If the AC is also used as a heat pump during the winter (which is becoming common), then there will be a few months where it will be "usually wet". When it is freezing the air is dry but between 0 and 20 Celsius the radiator is bound to get allot of dew. Commented Jun 8, 2022 at 7:06

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.