Ethylene Glycol as submersive coolant

Toggle sidebar Toggle sidebar
daw123

daw123

Platinum Member
Aug 30, 2008
2,593
0
0
You can get Ethylene glycol must cheaper than what the OP linked in his post.

Based on a quick Google shopping search exercise for mineral oil, it (for example, Honda CX 500) costs £3.00+ per litre.

http://www.google.co.uk/products?q=mineral+oil&scoring=p

Using the same exercise, Ethylene glycol cost approx. £4.00+ per litre (for example, Bluecol 2 year Antifreeze 1L).

http://www.google.co.uk/products?q=Ethylene+glycol&scoring=p

So they are similarly priced and I'm sure you could get both cheaper if you shopped around or bought in bulk.

Any way, there must be some property, which the coolant has that prohibits it use or makes it less desirable for using in submersed PCs.
 
Last edited:
Throckmorton

Throckmorton

Lifer
Aug 23, 2007
16,829
3
0
Because it's extremely toxic. It would be better to use propylene glycol which is much less toxic (it's even used in some cosmetics etc)... assuming it doesn't cause any kind of corrosion etc.
 
P

Pneumothorax

Golden Member
Nov 4, 2002
1,181
23
81
My best guess is ethylene glycol is hydrophilic so eventually you're going to have water start to build up in it.
 
AdmiralFace

AdmiralFace

Member
Oct 13, 2010
72
0
0
mobpcp.blogspot.com
Throckmorton said:
Because it's extremely toxic. It would be better to use propylene glycol which is much less toxic (it's even used in some cosmetics etc)... assuming it doesn't cause any kind of corrosion etc.
Click to expand...

Pneumothorax said:
My best guess is ethylene glycol is hydrophilic so eventually you're going to have water start to build up in it.
Click to expand...

Im fine with the toxicity, its only really harmful when ingested directly.
But the water build up is more of a problem. Could distilling it help?
 
EarthwormJim

EarthwormJim

Diamond Member
Oct 15, 2003
3,239
0
76
AdmiralFace said:
Im fine with the toxicity, its only really harmful when ingested directly.
But the water build up is more of a problem. Could distilling it help?
Click to expand...

The water build up will occur just from the environment. Not necessarily because there was water in the ethylene originally.
 
Hacp

Hacp

Lifer
Jun 8, 2005
13,923
2
81
AdmiralFace said:
In submersed PCs, the coolant is normally mineral oil or vegetable oil etc.

Why not Ethylene glycol (liquid cooling coolant:http://www.amazon.co.uk/Thermaltake-...8458737&sr=8-2)

Its effective and can be in all sorts of awesome colours. Ethylene glycol is non-conductive, could it have adverse reactions with the components?
Click to expand...

Well, ethylene glycol can absorb water vapor from the air. How that will affect the electronics, I do not know. Generally, you want things that are hydrophobic so water won't be able to mix in there and cause electronic chaos.
 
Modelworks

Modelworks

Lifer
Feb 22, 2007
16,240
7
76
Fluorinert would be best but unless you have a few thousand $ to spend I doubt you want to use that.

There is a reason ethylene glycol is not used in cooling computers, it isn't inert. You need to test everything it will come in contact with before using it. Some plastics deform and some metals will leech into it on contact making the liquid itself conductive.


Vegetable oil is a bad choice because it will become rancid.

The best solution is mineral oil. Just make sure it doesn't contain additives. Check farming supply stores , it is usually cheaper there. Transformer oil would be even better. It is a form of ultra pure mineral oil that is used in the transformers you see connected to power lines.
 
AdmiralFace

AdmiralFace

Member
Oct 13, 2010
72
0
0
mobpcp.blogspot.com
Modelworks said:
Fluorinert would be best but unless you have a few thousand $ to spend I doubt you want to use that.

There is a reason ethylene glycol is not used in cooling computers, it isn't inert. You need to test everything it will come in contact with before using it. Some plastics deform and some metals will leech into it on contact making the liquid itself conductive.


Vegetable oil is a bad choice because it will become rancid.

The best solution is mineral oil. Just make sure it doesn't contain additives. Check farming supply stores , it is usually cheaper there. Transformer oil would be even better. It is a form of ultra pure mineral oil that is used in the transformers you see connected to power lines.
Click to expand...

Good thing i didnt use it the eh? my comp would likely be dead XD

Thanks!
 
M

MadScientist

Platinum Member
Jul 15, 2001
2,175
53
91
I use Dow's Syltherm 800 in my lab's oil baths and glass reactor jackets. A bit pricey though for a computer. The last time I purchased it the price was ~$200./gal.http://www.dow.com/heattrans/prod/synthetic/syltherm.htm

Legoman666 used transformer grade mineral oil, Shell Diala Oil AX for his mineral oil PC. http://forums.anandtech.com/showthread.php?t=2060807&highlight=

Because of its toxicity and its affinity for water, ethylene glycol is a poor choice for immersing a PC in. At higher temps it is also a very good solvent. You would also have to purchase a lab grade EG since industrial grade EG contains ~3% water.
 
Last edited:
AdmiralFace

AdmiralFace

Member
Oct 13, 2010
72
0
0
mobpcp.blogspot.com
EarthwormJim said:
Yes exactly, so I don't really see how distilling would help. Unless you plan on removing the glycol from your computer periodically, distilling it, then pouring it back in.

Seems like an awful lot of work.
Click to expand...

Well I wasn't sure where the water was coming from, so if it was present in the liquid, distilling would remove it. More water would hopefully not build up as the system would be sealed. But, nonetheless, it's still unsuitable due to its reactivity with some metals and plastics.
 
AdmiralFace

AdmiralFace

Member
Oct 13, 2010
72
0
0
mobpcp.blogspot.com
I bought a small bottle to test it out and instead of containing ethylene glycol as advertised, it contains propylene glycol... would this work? is it inert/hydrophillic?

I've just tested the solution with my oh-so=handy multimeter, its got a resisteance of around 11mega ohms... that should be high enough to insulate any current flowing through the component on the mobo...
 
Last edited:
Modelworks

Modelworks

Lifer
Feb 22, 2007
16,240
7
76
AdmiralFace said:
I bought a small bottle to test it out and instead of containing ethylene glycol as advertised, it contains propylene glycol... would this work? is it inert/hydrophillic?

I've just tested the solution with my oh-so=handy multimeter, its got a resisteance of around 11mega ohms... that should be high enough to insulate any current flowing through the component on the mobo...
Click to expand...

Now you need to test everything it will come into contact with. Make sure nothing reacts with the glycol. It shouldn't but you never know with Chinese manufacturing now. Also when measuring resistance how far apart are the probes ? closer = less resistance. Contacts on a typical pc can be as close as 1mm. At voltages of modern pc that use 3.3v logic levels it doesn't take much leakage to cause problems.
 
Last edited:
M

MadScientist

Platinum Member
Jul 15, 2001
2,175
53
91
If you are hell bent on trying this then PG would be a better choice than EG due to its much lower toxicity.
I would be more concerned with its compatibility with computer components, tubing, and tubing connectors, especially at higher than room temps.

Cole-Palmer's chemical compatibility table is a great resource.
http://www.coleparmer.com/TechInfo/ChemComp.asp

PG
Material Compatibility
304 stainless steel B-Good
316 stainless steel B-Good
ABS plastic B-Good
Acetal (Delrinr) B-Good
Aluminum B-Good
Brass N/A
Bronze A-Excellent
Buna N (Nitrile) A-Excellent
Carbon graphite N/A
Carbon Steel B-Good
Carpenter 20 A-Excellent
Cast iron A-Excellent
Ceramic Al203 A-Excellent
Ceramic magnet A-Excellent
ChemRaz (FFKM) A-Excellent
Copper A-Excellent
CPVC C1-Fair
EPDM A-Excellent
Epoxy B-Good
Fluorocarbon (FKM) A-Excellent
Hastelloy-Cr B-Good
Hypalonr A-Excellent
Hytrelr N/A
Kalrez A-Excellent
Kel-Fr N/A
LDPE B2-Good
Natural rubber A-Excellent
Neoprene C-Fair
NORYLr N/A
Nylon A-Excellent
Polycarbonate B1-Good
Polyetherether Ketone (PEEK) N/A
Polypropylene A2-Excellent
Polyurethane B-Good
PPS (Ryton®) N/A
PTFE A-Excellent
PVC C1-Fair
PVDF (Kynar®) N/A
Silicone A-Excellent
Titanium A-Excellent
Tygonr N/A
Vitonr A-Excellent

Explanation of Footnotes
1. Satisfactory to 72°F (22° C)
2. Satisfactory to 120°F (48° C)

Ratings -- Chemical Effect
A = Excellent.
B = Good -- Minor Effect, slight corrosion
or discoloration.
C = Fair -- Moderate Effect, not recommended
for continuous use. Softening, loss of strength,
swelling may occur.
D = Severe Effect, not recommended for ANY use.
N/A = Information Not Available.
 
Last edited:
AdmiralFace

AdmiralFace

Member
Oct 13, 2010
72
0
0
mobpcp.blogspot.com
MadScientist said:
I would be more concerned with its compatibility with computer components, tubing, and tubing connectors, especially at higher than room temps.
Click to expand...

Six hours a chip, a diode, a capacitor, rubber band, polyphenelethene strips and some PCB in a beaker of PG. So far no change, still non conductive. I'll test at higher temps tomorrow.

MadScientist said:
Cole-Palmer's chemical compatibility table is a great resource.
http://www.coleparmer.com/TechInfo/ChemComp.asp
Click to expand...

Very handy thanks :)
PG seems to be relatively inactive in the presence of metals and plastics.
0.6mA @3v, 40degC
2 hours later:
0.6mA @3v, 50degC
 
Last edited:
You must log in or register to reply here.

TRENDING THREADS

COMPANY
RESOURCES
FOLLOW
Top Bottom