Fun with ethanol cooling.

GruntBlender

Prominent
Mar 26, 2017
49
0
560
I built a cheap custom loop not long ago and decided to run ethanol as my coolant. (Reference)
Well, while it has its advantages, you gotta be more careful than I was when doing something this stupid. Learn from my mistakes, people!

First, let's look at what it did to the acrylic reservoir:
NV773X9.jpg
MngxgFP.jpg
Tx9IKWT.jpg
fzJoLfs.jpg

Ethanol doesn't like plastics.
While the tubes and metal components held up unscathed, the coolant has also caused my pumps to fail by swelling their internals a bit to the point where the rotor would catch the pump body and stop. The plastic also became brittle and deformed under stress (Particularly under hose clamps)
A7q1Crk.jpg


The last victim was the drain plug. While it didn't cause a leak during operation, when it came time to drain the loop the plug had shrunk and became stuck to the screw lid, giving me a surprise squirt when I undid it.

Lastly, an amusing observation is that after I drained the loop, I didn't flush it, just filled it up with water (there's only a few percent ethanol in there now, so I don't expect trouble.) Somehow, clear water + clear ethanol + clear colouring produced an opaque solution. I'm sure that's just bits of my reservoir floating around in the water now, but it looks cool (until it fails again).

In conclusion, if you're planning on using ethanol as a coolant for some reason, make sure none of the components will react with it. Get a glass or metal reservoir, and a metal pump body (or at least one with enough clearance between parts to not be bothered by a bit of swelling.) Avoid acrylic water blocks and flow indicators.
 

GruntBlender

Prominent
Mar 26, 2017
49
0
560


Nope. Ethanol performs worse than water. Its advantages are the lack of corrosion, it's non-conductive, and it's a biocide. Downside is higher temps and melted plastic components.
 
Ethanol doesn't like plastics.
Yeah, alcohols make very good solvents. In a relatively strong solution of ethanol, acrylic will literally dissolve into nothingness. You could use other plastics like HDPE or PET (see here).

I would also expect ethanol to perform significantly worse than water - across the board, it's thermal characteristics (specific heat, thermal conductivity, etc.) are roughly half the value of those for water. Not sure what "advantages" people are spouting off but scientifically there is no evidence I see to support it. Interesting experiment though.

 

GruntBlender

Prominent
Mar 26, 2017
49
0
560


Yeah, it was fun. Resource you linked only shows mild interaction at worst for Ethanol though. It'll warp acrylic and can crack it under some conditions, but I don't think it'll dissolve it completely.
Its thermal characteristics are pretty bad compared to water as you mentioned. I'm seeing about a 6 degree difference on my GPU when mining. about the same with stress tests. Though that probably could be overcome with a stronger pump.
 

rubix_1011

Contributing Writer
Moderator
We've had others ask and test with this before and chemistry can provide you the answers before attempting this.

This is why water is the preferred method of liquid cooling for PCs - primarily because of the delicate nature of what the components are built from.