Heat transfer - water vs. 50/50 water/antifreeze



Which transfers heat more efficiently - water or 50/50 water/antifreeze? I’ve Googled this endlessly and no one seems to have a definitive answer.


50/50, because it can be used at higher temperature, increasing temperature drop, which increases heat transfer out of the radiator.


You are correct - regardless of climate, everyone should use 50/50 because of a higher boiling point. But ignoring that - which transfers heat more efficiently?


Once you are boiling, you aren’t transferring heat effectively. Therefore your question has be answered regarding heat transfer at the normal operating temperature of warmed up motor. At 200 degrees 50/50 is better. At room temp perhaps water is better, but not likely.

Since using straight water would ruin a modern cars cooling system, why are you asking?


My answer stands. You need a smaller radiator to use 50/50, that spells ‘efficiency’ to me. If, on the other hand, you want the heat transfer coefficients for water vs. 50/50 in something like BTU/ft2/minute/degree F, I don’t know. Here’s some info, but not that:


The folks who make Water Wetter claim straight water is a better heat transfer agent than 50/50 eth. glycol+water.

(Naturally, they also claim their product increases the heat transfer of both when added…)

I am inclined to give this claim (the first part) a modest amount of credibility.


Good question! I’m trying to get a theoretical answer from a physics standpoint, not from an automobile application standpoint.

Note: water wouldn’t boil in a car radiator at somewhat above boiling temperature anyway because it’s under pressure.

Also, if someone really, really wanted to use water as a coolant in a modern engine, he could do it by adding both a water pump lubricant and an anti-corrosion additive.


At 15 psig water boils at about 250F, 50/50 at about 265F.


Water does have a higher specific heat capacity i.e. the calorie per degree gram is higher than ethylene glycol. But as has been noted water vaporizes to steam on a hot surface thus reducing the ability of heat to transfer from hot metal. You can raise the total pressure; use a mixture of water and ethylene glychol; or use both higher pressure and a mixture to raise this limit to heat transfer.

Ever wonder why noone uses pure ethylene glychol as a coolant or even oil. It is because the specific heat capacity of glychol or oil is about a third that of water. Even though its boiling point would be higher the heat transfer surfaces would have to increased to get the heat from the engine hot spots; to the radiator; and then out to the air.


Just for the sake of the quiz I would guess that pure water would be the more efficient coolant. But rusty water is a very poor coolant and that’s what you get in a short time without the anti-freeze.


Also, the freezing point for 100% ethylene glycol is 10F, compared to -60F for 70% EG.


Pure water can transfer more heat than 50-50 mix…It’s a matter of how many BTU’s it takes to raise the temperature of water one degree… Pure water can absorb and transfer more BTU’s than most fluids…


I wish I could remember where I got the information, but way back when, the story was that you should not use plan water because it does not transfer the heat as well as the mix. BTW the same holds true for straight coolant.


See Table 2 on the first page…



I found this info at engineeringtoolbox.com concerning heat transfer. Note! The specific heat capacity of ethylene glycol based water solutions are less than the specific heat capacity of clean water. For a heat transfer system with ethylene glycol the circulated volume must be increased compared to a system with clean water.

We all know that clean water would become rusty almost immediately in most engines so the 50/50 mix is here to stay…at least until it’s replaced by something better.


I think it’s circulation rate, volume/minute, right?.


Probably. I think they were leaning toward water being a better heat conductor but I really don’t know. I have not seen a cut and dried answer for this question. It should be a simple experiment to prove one over the other. I don’t have the time, the lab or the real impetus to do this on my own.


Yeah, I too interpret this as circulation rate. The chart is saying the a mix would have to circulate at a greater rate to dissipate the same amount of heat as pure water. That roughly translates to it being less efficient at doing so.

Unfortunately, this information has no meaningful value in automobiles. Elevated boiling points, corrosion inhibitors, and lubricant for the water pump make a coolant mix the only real option anyway.


Does anyone here follow the 24 hours of Lemons or Chumpcar series? These are endurance races for cars that cost less than $500 (other than the cost of safety equipment) where basically POS beaters get run as fast as possible around a racetrack for hours upon hours.

I mention this because the rules dictate that the cars run straight water so as to avoid messing up the track if (when?) there’s a crash or a leak. This kind of racing is probably about the worst kind of abuse you can do to a cooling system (and many of them are not exactly in tip-top shape), and yet most cars don’t have any problems with overheating or boiling over.

  1. I doubt of they run 195 degree thermostats.
  2. those cars won’t have to go another 100,000 miles on the same water pump.
  3. nobody cares if they rust internally after.

In decent weather without longevity expectations and with no need to run at high temps for emissions, water will work great.