DIY Air Conditioner Chiller

Credit: Chris Paris

I occasionally get email requests for information on how to build your own aquarium chiller. My first words of caution are to forget about using a dorm fridge as the starting point for your chiller. A reasonable starting point is a window air conditioner. Eventually I would like to say more about building a chiller, but for now, here are some notes I sent to the last person who asked this question.

Building your own chiller is a very challenging project, and after trying it a few times, I found the $300 average price of a used chiller very inviting. That’s what I ended up doing; I bought a used chiller.

My final DIY chiller attempt, pictured above, was real close to being adequate. I went wrong in two places. First, the epoxy resin I chose was too brittle, so it developed small cracks over time. I don’t know if the cracks went through to the copper, but I didn’t like it. I used the 3:1 ratio epoxy from Fiberglass Coatings Inc.. If I did it again, I’d use their 1:1 ratio epoxy, which is more flexible, and in fact can be mixed deliberately off ratio to control the flexibility. I might also wrap the copper pipes in fiberglass cloth as a part of the epoxy coating stage. I’m talking about making a fiberglass-epoxy composite coating, which would have much more strength than epoxy alone. Fiberglass Coatings sells fiberglass cloth and tape (thin strips). Don’t make your coating thicker than necessary though, because epoxy (and glass) are not good thermal conductors compared to metals. I would try using a single layer of cloth no heavier than 4 ounces per square yard (that’s how fiberglass cloth is rated – in ounces per square yard).

My second mistake was trying to preserve the sheet metal end pieces of the heat exchanger. The copper tubes pass through these end pieces, and the end pieces serve to support the tubing. If I did this again, I would carefully cut away the end supports, and build some sort of support out of acrylic. Then my acrylic box would completely encapsulate the copper tubes and my new supports.

The metal end pieces made it difficult to get a good seal of epoxy between the sheet metal and the copper. After a few months of chiller operation, I noticed little bits of corrosion. Nothing bad happened in the reef, but I didn’t like the corrosion, so I just stopped using the chiller.

During its short service time, my DIY chiller was more effective than the commercial reef chiller that I use now. My DIY chiller started life as a 7800 BTU/hour air conditioner. A rating of 7800 BTU/hour is about what you get from a 3/4 HP reef chiller. But the 7800 BTU/hour rating of my air conditioner indicates how quickly it cooled air. It may cool water faster because of water’s greater thermal conductivity, or it may cool the water slower because the epoxy coating is an insulator. I didn’t do tests, so I don’t know which way it is.
I used a commercial (Goldline brand) digital temperature controller to control the reef tank temperature. It turns the air conditioner on as needed to maintain the setpoint.

Why a dorm fridge will not work as a reef aquarium chiller
It’s tempting to think that a dorm fridge will keep a tank cool. After all, it keeps beer real cold. The catches are that the fridge is insulated, and you are not constantly adding new warm beers to the fridge at a high rate. You wouldn’t use a dorm fridge to chill beer in a bar that wants to serve many cold beers throughout an evening. An aquarium is more like that. You’re constantly adding more heat, which comes from lights, pumps, and ambient air if the air is hot. You need a chiller that can remove heat faster than it’s coming in.

So rather than think in terms of how cold a fridge can get stuff, think in terms of how quickly it can remove heat. Air conditioners and commercial water chillers are rated in BTU/hour (in the United States), which is a measure of how quickly they can remove heat. A dorm fridge moves about 100 BTU/hour. A small window-mount air conditioner moves about 5000 BTU/hour (50 times as powerful as a dorm fridge). A 1/6 horsepower commercial water chiller sits in between these figures. It moves about 1500 BTU/hour. Therefore, it’s reasonable to conclude that we might succeed in building a reef aquarium chiller from an air conditioner, but not from a dorm fridge. The fundamental heat moving capacity is just too low in the dorm fridge. If you convert a dorm fridge to a water chiller, the fridge will run constantly without making a significant dent in the water temperature, even if you hope to achieve only a few degrees decrease in temperature.

You can get the heat into your dorm fridge in a number of ways, including coiled tubing, an open bucket of aquarium water stored in the fridge, or even by immersing the entire cold heat exchanger in a bucket of circulated aquarium water. It doesn’t matter how clever you are about getting heat into the fridge. The problem is that the fridge can’t remove it fast enough.

You can do some rough calculations to make my claim more intuitive. Suppose you have a 125 gallon reef aquarium. Neglecting your sump, your 125 gallon tank has about 1000 pounds of water in it. One BTU is the energy required to change the temperature of one pound of water by one degree Farenheit. If a dorm fridge moves 100 BTU/hour, then it would take 10 hours to drop the temperature of your tank by just one degree, and that’s if the tank has zero energy input, meaning no lights, no pumps, and perfect insulation from the room. Not very encouraging is it?
Another way to look at it is that one BTU/hour is 0.293 watts. Thus, a dorm fridge that can remove 100 BTU/hour can counteract a heat input of 29.3 watts. That’s good for one small powerhead, or a small fraction of a metal halide lamp.

Let’s run the same numbers with a 5000 BTU/hour air conditioner as the cooler. That could drop your tank temperature by one degree in 12 minutes, or five degrees in one hour. We’re still neglecting the very significant heat input, but these numbers are more promising. 5000 BTU/hour is 1465 watts, so the air conditioner would maintain a constant temperature in a tank (any size tank) that has a heat input of 1465 watts. That’s more like what we need for a reef tank.

Building a water chiller from an air conditioner is a tough task, but I have some notes about it here.

Occasionally I hear from someone who claims to have an effective dorm-fridge-based water chiller. I then start asking questions, and I usually discover that, at the same time as the dorm fridge went online, the user made one or more other changes, such as adding whole-room air conditioning, adding a fan to the aquarium blowing at or over the water, decreasing light use, and so on. Occasionally I find someone who believes that he has a working dorm fridge chiller but has made no other changes that would help to maintain a lower temperature. I suspect that these are cases of not having all the information, and that there is always some explanation other than that the dorm fridge really is doing what it cannot do.

Blane Perun

Diver - Photographer - Traveler

Whale in Ocean