Using oil as liquid cooling?

[Chiyawa]

Hi,

 

I wonder if I'm way too crazy to think of using oil instead of distilled water for liquid cooling. I remember substation use oil to cool down the transformer.

 

Yeah, we might need to modify the pump because of the higher viscosity of the liquid, but... Maybe this will solve lots of problem that normal water liquid cooling has, such as corrosion and gunk build up.

 

I don't know, but hey, Linus Tech Tips, maybe you can experiment if my crazy theory makes sense?

 

Regards,

Chiyawa

[Jumballi]

Been done, tested, and it doesn't work
Distilled water is much much much much cheaper and less corrosive

For the same value as oil, you can use liquid compounds which are much more thermally conductive and are not corrosive

[Guest]

15 minutes ago, Chiyawa said:

I remember substation use oil to cool down the transformer.

Isn't this mainly due to the oil used not being electrically conductive?

Also, water is more thermally conductive when compared to oil.

[Chiyawa]

9 minutes ago, Jumballi said:

Been done, tested, and it doesn't work
Distilled water is much much much much cheaper and less corrosive

For the same value as oil, you can use liquid compounds which are much more thermally conductive and are not corrosive

Hmm... Can you maybe point me to this article? I haven't come across any thing related to liquid cooling using oil.

[Chiyawa]

2 minutes ago, lewdicrous said:

Isn't this mainly due to the oil used not being electrically conductive?

Also, water is more thermally conductive when compared to oil.

Yeah. I mean, in case of leaks, we don't have to worry our electronics being short circuit.

[Dedayog]

12 minutes ago, Chiyawa said:

Hmm... Can you maybe point me to this article? I haven't come across any thing related to liquid cooling using oil.

Oil works as a coolant in large systems with heavy torque loads, like transmissions, differentials, etc.  But oil doesn't remove heat like water or other liquids, so it's not as effective on sensitive items.  

[SenKa]

I mean, oil is a broad term.

 

Motor oil? Mineral Oil?

 

The latter has been done plenty of times to cool PC's, with very good results. However, mineral oil does nasty things to plastics sometimes. I believe Linus had an issue where after he disassembled his old mineral oil PC the DIMM slots on the motherboard broke apart.

[Tarbot]

Your pump will thank you for not using oil...  The flow rate and effort to pump it would kill most pump heads.  All I use is distilled water, biocide, and an anti corrosive...  No dies or funky shimmering premix (They clog up micro fins).  I can also see what is going on in my loop as far as any growth or discoloration (Use non plasticiser tubing).  The biggest thing is maintenance...  Once a year drain, clean, and rebuild.  Good time to upgrade anything you want as well!

[Guest]

3 minutes ago, Chiyawa said:

Yeah. I mean, in case of leaks, we don't have to worry our electronics being short circuit.

You have to keep in mind the trade-off in thermal capacity/conductivity.

Without going down the extreme cooling routes (LN2, dry ice + isopropyl alcohol, etc.) I think water is the best you can get, imo at least. If, however, you're referring to extreme cases (extremely high/low temps.), then there are better fluids like actual coolants, instead of pure water, but this introduces problems of its own.

 

2 minutes ago, SenKa said:

However, mineral oil does nasty things to plastics sometimes. I believe Linus had an issue where after he disassembled his old mineral oil PC the DIMM slots on the motherboard broke apart.

This as well as being really hard to clean; they kept the parts that they used in their mineral oil build in a bucket for a while and they still had oil residues when they used them after storage.

[Mick Naughty]

You won’t have an issue with distilled water if it leaks either. Unless it’s been in there for a year. 
System shouldn’t be leaking if you have a system that has oil in it. 

[akio123008]

8 hours ago, Chiyawa said:

I wonder if I'm way too crazy to think of using oil instead of distilled water for liquid cooling. I remember substation use oil to cool down the transformer.

 

8 hours ago, lewdicrous said:

Isn't this mainly due to the oil used not being electrically conductive?

Also, water is more thermally conductive when compared to oil.

The oil used here isn't primarily for cooling; it's used to prevent electric arcs between conductors because oil has a higher breakdown voltage than air. (because voltages in substations are very high) Actually, modern substations don't use oil, but instead rely on some gas I can't remember the name of.

[Guest]

1 minute ago, akio123008 said:

Actually, modern substations don't use oil, but instead rely on some gas I can't remember the name of.

This is news to me.

Looking into it a bit, I think it might be sulfur hexafluoride.

[akio123008]

Just now, lewdicrous said:

This is news to me.

Looking into it a bit, I think it might be sulfur hexafluoride.

Yes that must have been it. I only recalled the "hex" part.

[Lipe123]

The easiest two disqualifiers for oil/glycol are:

• Cant pump it properly with a D5/DCC.
• It's doesn't have the same thermal conductivity as water, see: http://www2.eng.cam.ac.uk/~mpfs/papers/articles/WTC2005/pdfs/t-3/WTC2005-64316.pdf

Thermal conductivity is a non issue, no one builds a loop to leak all over.

Leaks are not common and only happen when you mess up the build or something breaks.