Sign in with
Sign up | Sign in
Your question
Solved

Mineral Oil vs. Water in Liquid Cooling System.

Last response: in Overclocking
Share
February 8, 2010 12:08:31 AM

Okay guys I need your help as this question has been rattling in my brain for a week now and I am yet to settle on the best course of action. First of all humor me as this is an experiment nothing more at this stage.

The Scenario:

Running a standard PC liquid cooling system connected to CPU, Videocard, Ram and Hardrives in overclocked system.

Modification:

Running a 1 gallon reservoir/radiator submerged in 10F/-12C Bath.

The Debate:

Running Mineral oil in cooling reservoir & tubing versus water w/ antifreeze in cooling reservoir & tubing.

Control:

Water or Mineral in reservoir are Chilled to 10F/-12C before system is started.


My current thoughts and considerations:

- Water should absorb heat faster but that means entire system cooling water (1 gallon +) may loose its cooling temperature and heat up faster after some use.

- Mineral oil holds its temperature better, therefore continues to cool longer as the chilled oil in reservoir will not succumb to heat as fast as water will however will it transport heat from the component blocks as efficiently? or rather retard heat due to it's maintaining its cold better than water?

- Running water with antifreeze gives added potential to run system at -40F but again due to thermal conductivity I get the feeling water/antifreeze will only best in short applications

- Mineral oil's lowest usable temperature is 10F however if everything I am reading is accurate the oil won't heat up or loose it's cool in reservoir for some time which makes it ideal for longer sessions? (I believe Puget ran their fully submerged system for 8 hours in uncooled still oil)

All comments, advise, and thoughts are welcome I will post/update results as soon as I have completed this experiment I just want to make sure I have thought of everything before doing this, you know...measure twice cut once and all that.

Thanks in advance for your insights.


More about : mineral oil water liquid cooling system

a b K Overclocking
February 8, 2010 12:12:54 AM

Mineral oil is a terrible thing to use in a system like this.

The reason being is that...

1. It is a bad heat conductor compared to water

2. It has a high specific heat capacity meaning that it will hold more heat, and will not release it very easily when it gets to the radiator.
m
0
l
February 8, 2010 12:20:24 AM

I hear what your are saying but let me add this factor to the equation for the sake of argument. Lets assume the radiator is submerged in the same 10F/-12C bath for right now, no fan air across the radiator.

Another question and I am going to look this up now, heat can be conducted away from the heat producing component true but can't thoroughly chilled liquid also transfer cold over to the hot component forcing it's temperature down?
m
0
l
Related resources
Can't find your answer ? Ask !
a b K Overclocking
February 8, 2010 12:25:51 AM

Ok, if you are going to submerge the radiator (or have a chilled res), then water would work much better.

Obviously with Sub zero temps you cant run with water. So acetone or antifreeze would be your best bet. I doubt the water will stay sub zero anyway.

How do you plan to chill the liquid in the res?
m
0
l
February 8, 2010 12:53:40 AM

In a medical/lab freezer modified to allow in and out hoses after they run through bath located inside of it.
m
0
l
a b K Overclocking
February 8, 2010 1:08:52 AM

You will burn out the compresser. It is not ment to handle an active heatload. DO you know what the HP rating for the compressor is?

10/1 it wont survive very long
m
0
l
February 8, 2010 8:44:19 PM

Thanks for the response I will look into its specs tomorrow. I am also toying with the idea of running an air fanned radiator before it enters the cooling stage to perhaps shed some heat. I also have access to a thermo-electric cooler which may eliminate the compressor issue but I need to educate myself on that technology and its limitations.

Another thing I was wondering about was condensation. I guess I will find out when I put this thing together but I am worried that with water condensation may form and leak onto PC parts which it not a good thing, that's why I find myself wondering about mineral oil again. Thoughts?
m
0
l
a b K Overclocking
February 8, 2010 9:19:39 PM

You will have severe limitations witha TEC. It will be to much work and your TEC probably cant handle the load.

Read this article that i wrote for sub zero insulation. If you are running 24/7 you will also want to put some dielectric grease in the socket. And maybe paint the MOBO with nail polish before starting the steps i outlined here:
http://ovaclockers.com/e107_plugins/forum/forum_viewtop...
m
0
l

Best solution

a b K Overclocking
February 9, 2010 1:25:13 AM

z1nomad said:
I hear what your are saying but let me add this factor to the equation for the sake of argument. Lets assume the radiator is submerged in the same 10F/-12C bath for right now, no fan air across the radiator.

Another question and I am going to look this up now, heat can be conducted away from the heat producing component true but can't thoroughly chilled liquid also transfer cold over to the hot component forcing it's temperature down?



Some good questions here, but let me answer a couple of them, since I did take a few years of trade school refrigeration and air conditioning.
1. Antifreeze does not just lower the freezing point of water. It also contains additives that prolong the life of bearings, seals, and hoses, as well as rust inhibitors. The chemicals in antifreeze also allow the water to be a better thermal conductor when mixed properly than with clean water alone. A 50/50 mix is the normal "best for everything" suggestion.
2. There is no such thing as "cold". You cannot transfer "cold" to anything, anywhere in the universe. It doesn't work that way. There is only heat, or the absence of heat. When something is lower in latent heat than our body temperature, it feels what we call "cold" to us. But more accurately, we should say that it feels like it has less heat than we desire for comfort. The reason it feels "cold" is because the heat is leaving your hand when you touch something that does not contain as much latent heat as your hand. The larger the difference, the faster the heat transfers from your body, giving you the feeling we call "cold".
Just as an air conditioner does not create "cold" air in a technical sense. It simply moves the latent heat in the air from an area where it is undesirable, and moves it to an area where it is more desirable. Hence, the hot air exhausting outside. That hot exhaust is the latent heat that has been removed from the air inside the room or house. Hence, the heat pump. Basically a 2 way air conditioner. It can move heat from the inside to the outside, or from the outside to the inside. Even if it seems very cold outside, there is still latent heat in the air that can be removed and moved into the house.
Any air conditioner is a actually heat pump, but a standard air conditioner is only a 1 way heat pump.

Share
February 9, 2010 9:51:09 PM

I love it! learn something new every day. Thank you for your insights and I welcome more of them as I am ordering the parts and putting this together I can always benefit from the experiences and education of others. It's funny there is no Guidebook for projects like this and one ends up having to learn about liquids, refrigeration, heat dynamics, condensation; each of which is an ocean of research by itself.

Decided to move forward with water/antifreeze. Now starts the fun of acquiring all the bits and pieces and setting up shop. Updates/comments/questions as this develops.
m
0
l
February 9, 2010 10:33:45 PM


Definitely water! We make air coolers out of copper because they move heat better, but run it without a fan and yes they get hot. Make an air cooler out of styrofoam and the cooler will stay cold, but your chip will cook.

The mineral oil may stay colder longer - but that is actually bad, because its pulling less heat away from the parts. plus circulation issues at temp and just the pain of working with it compared to water.

m
0
l
February 9, 2010 10:34:54 PM

suddenstop said:
Make an air cooler out of styrofoam and the cooler will stay cold, but your chip will cook.

styrofoam will melt lol
m
0
l
a b K Overclocking
February 9, 2010 10:46:30 PM

A great book if you wan to get into refrigeration is "modern refrigeration and air conditioning" it can be picked up online cheap.

There is another great book written by gary lloyd, a guy on xtremesystems, it outlines some of the more advanced types of refrigeration. If you really want a subzero rig, i would buy a phase change. They will get real low for about the same price of an extreme wc loop.
m
0
l
February 9, 2010 11:25:37 PM

Thanks for the info. I have also been looking at this and the possibility of integrating a peltier into my project somewhere. I was wondering what you can you tell me about how freezers/fridges might handle an internal heat load on a continuous basis. Limitations, considerations e.t.c. anything you can offer would be helpful and will help me better design this thing.
m
0
l
a b K Overclocking
February 10, 2010 12:15:35 AM

z1nomad said:
Thanks for the info. I have also been looking at this and the possibility of integrating a peltier into my project somewhere. I was wondering what you can you tell me about how freezers/fridges might handle an internal heat load on a continuous basis. Limitations, considerations e.t.c. anything you can offer would be helpful and will help me better design this thing.


Peltiers just arnt that good unless you can get a high wattage. Even then you will have a tuff time breaking sub 0 on an i7.

The thing about phase change is that it will handle a continuous heatload at a very low temp. A fridge is not ment to handle an active heatload, it is ment to cool a isolated area down to temperature then keep it that cold. This takes very little work from the compressor. You will want atleast a 5000 BTU A/C unit if you are going to do any kind of chilled water unit.

I would not buy the cryo-z. It was never ment to take the heatload of an i7. It was ment for C2D chips, even quads will get that unit to positive temps. I would suggest buying a unit from piotres or little devil. They will make a phase change that will keep it alot lower than the positive 20c seen when overclocking an i7 w/ a cryo-z.
m
0
l
February 10, 2010 1:21:27 AM

Aha! Now you got me thinking. A combo, what about freezer/fridge for reservoir with upside down heatsink submerged in it fused to sealed opening at top, 220W Peltier on top at the bottom of heatsink, then a righside up heatsink on top of peltier sandwiching it with heatsink having 2 fans in a push/pull. Fridge for reservoir, Peltier to draw heat heat load out of fridge. I know its roughly laid out but you sere where I am going?
m
0
l
a b K Overclocking
February 10, 2010 1:25:37 AM

Or how about just a phase change? You would get better temps than with all that stuff your talking about.
m
0
l
February 10, 2010 11:40:49 AM

Researching this today. Thanks.
m
0
l
February 24, 2010 3:36:37 AM

If you wanted to actually cool the liquid, and you don't think a medical freezer could handle it, you could try an aquarium chiller. They are expensive, but they are made to cool down large amounts of liquid. You could probably get very low temperatures in your lines if you ran your liquid through one of those.
m
0
l
a b K Overclocking
March 17, 2010 7:54:41 PM

Best answer selected by r_manic.
m
0
l
!