Physics Van 3-site Navigational Menu

Physics Van Navigational Menu

Q & A: How long for water to freeze?

Learn more physics!

Most recent answer: 01/07/2017
How long does it take for water to freeze?
- Dara (age 12)
Hi Dara!

The answer to your question really depends on three things: how much water you have, how cold it is to start out, and how cold the things around it are. Water actually freezes when it gets to 32 degrees Fahrenheit (0 degrees Celsius), but the time it takes to get there may be different.

Let's start with the first. If you take two glasses, and fill one with a tiny bit of water, and the other about halfway, then put them both in the freezer, the one with less water will freeze first (you can try this at home, but I recommend using plastic cups and not glass ones).

Now let's move on to the second part. Let's say you have two glasses, and you fill one with really cold water that has been in the refrigerator, and the other with really hot water from the sink. If you put both of them in the freezer, the one that started out colder will freeze first.

For the third part, let's imagine that you have two glasses with the same amount of water in them, and the water is at the same temperature. Imagine putting one outside on a really really cold day in Georgia, and having a friend in Alaska put one outside on the same day. Since it would be so much colder in Alaska, the glass of water there would freeze before yours.

So, if you took a tiny bit of really cold water in a glass, and put it outside on a cold day in Alaska, it would freeze a lot faster than a big glass of hot water outside on a cold day in Georgia.

Hope this helps!

(published on 10/22/2007)

Follow-Up #1: freezing and authority

This isnt a question, but in fact a correction to your answer, Sara. It has been known for centuries that hot water freezes faster than cold water. Both Aristotle and Francis Bean believed this to be true. It has been proven fact in several different experiments by many different scientists. You gave a good answer, and certainly helped out Dara, but I just wanted to make sure she got all her facts straight. Mackenzie
- Mackenzie (age 11)
Midland, MI, USA
Mackenzie- Thanks for your note. You're right to remind us that the common-sense result isn't always right. Sometimes the hot water freezes faster.
I would be a little more cautious than you about authority, however. Aristotle made many mistakes, even on simple questions like how many teeth women have. Francis Bacon (is that who you mean?) also  wasn't much of a scientist.

My own attempts to repeat this experiment have flopped so far. I always forget to look at the glasses until they're both frozen. You might try it yourself, using metal cups so they don't break, and being sure to put the same amount of water in each. Also, try it a few times switching the positions of the hot and cold cups, since freezers don't cool evenly.

 The most careful discussions of the subject (e.g.) suggest that which water freezes fastest depends on several factors, including whether it is allowed to evaporate, how well heat is conducted into it, how much gas is dissolved in it, etc.

I think Tamara has a good discussion on this site, under the name 'Mpemba effect', named after a high school student who had the courage to believe his observations rather than his teachers.

Mike W.

(published on 10/22/2007)

Follow-Up #2: what freezes first

Hot water does not freeze faster than cold water. That idea is ridiculous. It can be misconstrued as such when you look at it without a mind towards physics and chemistry. When a liquid is cooled, it may pass the freezing point and not appear to freeze. This is due simply to the fact that the molecules need additional energy or a solid to begin the crystallization process. If you put boiling water in the freezer along with cold water, the cold water would reach the freezing point of 0 C much more quickly, but it does not make the transition to ice until a force acts upon it. If you rap the glass you will see the liquid freeze in front of your eyes. The boiling water would not even have reached the previous temperature of the cold water. In conclusion, with both liquids under constant pressure, there is no way for hot or boiling water to "freeze" first.
- Cody (age 17)
United States
Your argument makes a great deal of sense, but subtle complications in the real world can give strange results. Here's an example. Let's say that you set the water in a glass in a freezer with a lot of frost. The hot water may melt the frost, and then make excellent thermal contact with the cold shelf. The cold water may sit on top of the frost in poor thermal contact. Then the initially hot water could cool more quickly and end up freezing first. In some cases it's necessary to actually do experiments to find out how things act.
Mike W.

(published on 02/09/2010)

Follow-Up #3: how long does it take for cocoa to freeze?

Can someone answer my question?? If you put 200ml-500ml of *cold* cocoa in the any normal freezer, how long does it takes to freeze?
- Kavi (age 19)
Dear Kavi, 
The actual time depends on may factors:  thermal contact and quantity of cocoa, beginning temperatures of the freezer and coca, composition of the cocoa, etc.  The best answer is for you to do the experiment and find out for yourself.  Try it several times varying one or another parameter.  Keep a record and see if you can figure out a trend.
Let us know what you find.


(published on 03/04/2010)

Follow-Up #4: Time for water to freeze?

Can someone provide an answer on how long a specif volume of water will take to freeze at a typical residential freezer temp, i.e. 8 ounces of water starting at 38 degrees farenheight takes "X" minutes to freeze, or "x" minutes per ounce at "x" degrees?
- Peter Thompson (age 29)
Minneapolis, Minnesota, USA
No.  This sort of problem is unanswerable because it depends on too many variables that we don't know the values of. At best one can give scaling laws, for example twice the heat transfer rate will halve the freezing time or twice the volume, at the same transfer rate, will double the time.  The freezing time depends on: the air circulation rate, the surface and shape of the container, the amount of contact area, etc. The best way to answer this question is to do some experiments and vary some of the parameters.  Eventually you will get some empirical idea of what's going on.


(published on 06/13/2010)

Follow-Up #5: time and rate happeneth to all

Rate and time are being confused. RATE is DISTANCE/TIME. 100C to 0C / x = 100/x && 5C to 0C / x = 5/x. if x is 5 minutes then 100/5=20 > 5/5=1. 20>1 = FASTER RATE.Though related they are not the same thing. Hot water freezes at a faster RATE than Cold Water. The equal sized/surface/area/pressure/temp/volume of the container of water closest to 0C will freeze in a shorter period of time than the hotter container of water. Hot water is often used because it lacks the dissolved air which makes the ice look cloudy. Ice sculptures can be crystal clear like glass if the hot water is super cooled to freeze the water before much air gets caught in the water. Common Chemistry experiments are done by taking temperature readings of a solution while stirring it to maximize the solutions contact with the surface of the beaker which is submerged into a ice/salt bath. The ice/salt bath on the exterior provides an subzero environment to freeze the solution on the inside. A Computer is usually doing the sampling automatically and graphing the results in a 2d graph showing the temp reading on the y axis and the time frame on the x axis. typically the graph is logarithmic. you see the temperature of the solution plummet quickly, but the amount of energy required to break the phase barrier is exponential. So you see water go from 100C to 1C in a short period of time, then it takes a much longer period of time compared to the 100C to 1C to go from 1C to 0C. Thus, the misunderstanding is that hot water freezes 'faster' than cold water. The same can be said about boiling water. the amount of energy introduced to the system to achieve boiling point is exponentially larger than the quantity to heat it up to 99C from 1C. observe the graphs where f(x) = y, f(x) = log(x) and f(x) = x^2 or f(x) = x*x [pronounced x squared) then turn each one 90 degrees to see they're the same thing on a different axis. Maybe a more real world scenario is in order for clarification. Bob is standing at the beginning of a 100 meter dash. Walter is standing 1 meter away from the end of the 100 meter dash finish line. The whistle blows! Bob is off to a speedy start traveling at 10 meters per second. It takes Bob 12 seconds, accounting for the the time it takes to accelerate from 0 to 10 m/s, to finish the race. Walter, being 1 meter from the end took his sweet time traveling at at only 1 meter per second and crossed the finish line 11 seconds before Bob. Bob traveled at a FASTER RATE, but Walter still finished first. Who was faster? Bob was faster! Who finished in the least amount of time? Walter, because he finished in 1 second. To properly calculate the time it would take to freeze a substance from point x to point y... As many posts before said there is not enough information given. Glasses are typically cylindrical in shape. So lets start with the Surface Area of a Cylinder Equation: (2*pi*radius^2)+(2*pi*radius*height). The SUBSTANCE of the container is very important for heat transfer rates as well as the mass of the container. --hth
- WALTER (over 9000 years old)
Anchorage, Alaska, USA
Walter- I should defer to your advanced age, but I see no evidence that anyone was confusing time and rate. These messy practical heat flow problems cannot reliably be reduced to your simple equations.

Here's some reasons why hot water could actually freeze sooner, surprising as that is.

1. There's a burst of initial evaporation. That leaves less water in the hot container, so that less latent heat needs to get dumped to the surroundings. If that's the reason, there would of course be evidence: less final ice.

2. There's less dissolved gas. That means that the freezing point is higher. If the surroundings are just a little below 0C, the time it takes to freeze is very sensitive to even small changes in the freezing point. For example, if there are enough solutes to lower the freezing point below the surrounding temperature, the time becomes infinite.

So the freezing  time has a complicated dependence on the temperature of the surroundings, the way in which heat is conducted in, the relative humidity of the surroundings, the amount of dissolved gas..... As a result, you can occasionally get the weird situation is which the hotter water freezes sooner, at least according to some experimental reports.

Mike W.

(published on 12/08/2010)

Follow-Up #6: how long for water to freeze?

I don't know why you couldn't give a rough answer to the initial question. Would it take minutes or hours for a room temperature cup of tap water to freeze in a typical residential freezer? That was all the little girl was asking. You didn't answer her question, you told her to find out herself. Mythbusters answered the question about beer. Try to help, okay? My guess would be about an hour. UIUC Engineering Grad & Mother
- E Wickliff (age 40s)
Wheaton, IL
I've made a guess as to which question your comment was intended to follow-up.

The original Q&A on this goes back to before my time on the site, so I'm not sure why an approximate answer wasn't given.  It's not hard to imagine a reason, though. It's rare for the turn-around time on this site to be less than a day. Often it's weeks. Sometimes years.  The time to freeze a cup of water in a home freezer is less than a day, so just doing it gets the answer quicker than asking us, and avoids having to rely on authority.

My guess would be maybe three hours to freeze if you really wanted it frozen or maybe half an hour to freeze if you were just trying to cool it and the glass would break if you forgot to take it out in time.  But children shouldn't be exposed to that kind of superstition.

Mike W.

(published on 04/07/2011)

Follow-Up #7: philosophy of freezing water

Like E. Wickliff pointed out, its really funny that no one really answered the girl's question. But moving forward, here's a link to why hot water DOES freeze faster than cold... Basically put, the hot water won't create an insulating layer of frost on the top and it contains less gas bubbles due to evaporation which in turn causes the hot water to cool more quickly as it evaporates more quickly than cold water. Faster cooling means it will eventually become colder than the initial temperature of the cold water" And age is mostly irrelevant to knowledge.
- Peter Ryan (age 31)
Vancouver, BC, Canada
I looked at your link about hot water freezing faster. It basically speculated about the same reasons as we had discussed, plus adding another speculation about how the lack of bubbles in the hot water would promote supercooling. That's a little odd, since by definition that effect would make the freezing of the hot water slower.  They then add a speculation that since the cold water would start freezing sooner, it would develop an icy crust that would slow further freezing. However, they really have no argument as to why a similar crust wouldn't form on the hot water when it starts to freeze.

The main thing I hope students would take from this is that each of these speculations can be tested. If the theory is that some of the hot water evaporated, you can check to see if the amount of ice it makes is noticeably reduced. If the theory is that the dissolved gases in the cold water lowers its freezing point significantly, you could try using cold water that had been degassed, maybe with a vacuum pump or other means. And so forth- each idea should be checkable.

This brings us back to the simple question of about how long it takes to freeze water in a home freezer. Why would anyone ask us that when it's so easy and quick to check directly?

Mike W.

(published on 09/09/2011)

Follow-Up #8: testing freezing times

Yes, we can be able to test it. But what is the correct temp for the freezer? And why would it be better?
- Tina (age 12)
Wesley Chapel, FL, USA
There's no particular "best temperature" for the freezer for this experiment.

So if you were looking for the strange effect of the hot water freezing sooner, you might test with the temperature just a little below 0C, say -5C. That's because dissolved gases (more likely to be in cold water) lower the freezing point, and that's very noticeable if the freezer temperature is close to the freezing point. If you were just trying to see what happens, without any particular favorite result, you could do that at any temperature, with the results maybe depending on temperature.

Mike W.

(published on 10/06/2011)

Follow-Up #9: time to freeze water

1 cup of cool water (in a glass tumbler) took about 3 hours to freeze in my freezer. ;)
- nnugles (age eh?)


(published on 08/01/2012)

Follow-Up #10: How long to freeze water?

I did this for my science fair project and it took 80 minutes for 4 ounces of room temperature water in a thin disposable cup to freeze in my freezer.
- Xander (age 6)
Cave Creek, AZ, US
Xander- Thanks, some of our readers will like this information.

Mike W.

(published on 02/17/2013)

Follow-Up #11: wisdom of babes

I'm so glad the 6 year old answered the question that it takes 3 hours. I'm making pina coladas and want to chill the pineapple juice and canned pineapple, but not freeze them. Getting an answer of 3 hours means that my timer of 1 hour will do it safely. I respect that it is better to give an ESTIMATE of an answer, than to plead ignorance.
- Peter W Lindner (age 63)
United States

I hope he sees your thanks.

Mike W.

(published on 08/30/2013)

Follow-Up #12: experimenting on freezing water

i am goin to be doing my junior cert this year and have a few questions.... For my science i am doing a project which is based on which temperature of water freezes the fastest an which one melts fastest at room temerature.. i will be doing it more then once with varius sized containers to get an accurate result. the junior cert is an important set of exams and leads to how well we will do and which college we will go to. so fr ive alot written down on that topic but am struggling. My teacher has told me to find out a method for how long the water takes to freeze. so basically if i go to the freezer and both are frozen i don't know how long it took to froze. so how often should i check the water and also how do i know if there is a difference from the last time i checked? do i just check the temperature or is there something im missing? please help and thanks :)
- aoife (age ...)

I can't follow what you're comparing here, but at any rate it seems that you want to compare freezing times for several samples.

Unless you can get some sort of automatic data logger you should check pretty often, maybe every 15 minutes or 1/2 hour. I grew up on experiments like that where we had to keep an eye on things almost continuously.

Maybe now you could do something clever, like conduct the experiment in a styrofoam cooler with a viewing port. Cooling maybe could be supplied by some dry ice/alcohol slurry. Each sample could have a thermometer in it. You could take pictures with a digital camera, maybe even with a timer. To avoid fogging, the viewing port could just be a long styrofoam tube with open ends, also serving to vent the CO2 from the evaporating dry ice. A little LED light could be turned on inside when you wanted to view or take a picture.

Whatever you end up doing, a crucial part will be to write it up clearly so people can know why you did it and what you did. What are you trying to find out?

Mike W.

(published on 09/09/2013)

Follow-Up #13: looking at freezing water

its saying i can't check every 15 mins cause it might influece the freezing by opening the freezer
- aoife (age 14)

Sure, opening the freezer every 15 minutes will change the freezing time. It will change the time a little for both the water that started out warm and the water that started out cold. This is just like using a slightly different freezer. What's wrong with that? It's not as if you started out knowing exactly which freezer with exactly which properties was the one you wanted to do the test in. 

The first thing to do here is to write up carefully what you're trying to find out and why. Without that there's not much point in worrying about the details of what to do.

And who is the "it" that's telling you what to do?

Mike W.

(published on 09/16/2013)

Follow-Up #14: Teach a man to fish

I could not agree more with the earlier comment that it is incredible that this question was either not provided with a solution or an estimate. With all due respect, Mike W, you have directly contributed to this thread being ridiculously long - my son almost gave up twice wanting to find the answer within the thread because there is so much 'hot air' and speculation here without an answer. In the future Mike W. I might suggest instead of pontificating about how one might go about finding the answer, that it might be better to stay off of these answer boards unless you actually have the answer. Good Day.
- Mark (age 40s)

It's a little ironical to get your suggestion now. Our group recently had a discussion about how we'd like to do more to help people figure out their own answers and test answers that they might hear, rather than just passively write down 'facts' from the Web. You know that old saying about teaching a man to fish.

Mike W.

(published on 11/24/2013)

Follow-Up #15: philosophy of teaching

I am a retired secondary science teacher and applaud "teaching a man to fish". Understanding the context of what we are trying to do is critical to plausibility and efficacy. Where might we now be in our troubled, perilous world if, before we began doing all the things we've done, we had first devoted an inordinate amount of time determining if we ought to be doing it at all? Simply demanding a conclusion represents a detrimental amount of arrogance and disrespect for the elegant systems, of which we are only a part.
- Robert (age 62)
Oconomowoc, WI

We're happy to pass those thoughts on.

Mike W.

(published on 01/25/2014)

Follow-Up #16: Does warm water freeze faster?

If two glasses of water are placed in a home freezer, one at room temp and the other at say 150*, I believe I was taught that the warmer water would freeze first due to the fact that the warmer water had more stored heat energy than the cooler glass of water. The cooling would proceed at a faster rate in the glass of warm water. Am I correct? If not,please correct my view. Thank you.
- Marty Shows (age 71)
Houston, TX

From what we hear, which freezes faster depends on detailed conditions. The reasons are discussed in the earlier part of the thread. The reason that you suggested though, is not one of them. More stored heat by itself would mean that it would take longer to get rid of the heat. It's other factors, such as less dissolved air, that can make the warm water freeze first.

Mike W.

(published on 02/22/2014)

Follow-Up #17: experimental results on freezing hot and cold water

After reading all of the variables (Answers) to this Q posted on the Physics Van, I decided to turn this into an experiment, to introduce my 6 yr old daughter to the Scientific Method. So... We took 2 samples each of boiling and cold tap water (N= 4). Each sample was 4 oz (120 mL). 1 boiling 4- ounce sample was placed into a bowl. 1 boiling 4 ounce sample was placed into a COVERED plastic bottle. 4oz Cold tap water was placed into an identical plastic bottle, and 4oz cold H20 into an identical glass bowl. The 4 containers were placed outside in Alaska on a 10 degree F (-12.22 Celsius) winter afternoon.. My daughter won this guess- the glass bowl of COLD water froze first, at 35 minutes. The plastic bottle of Cold water froze second, at 41 minutes. The glass bowl of hot water froze third, at 44 minutes. The plastic bottle of hot water finished last, and (uncertainly) froze at 87 minutes (I'm not sure if the floating ice formed on the side of the bottle, or not). All times are for the formation of ice crystals in/on the liquid (not entirely frozen). I came to the Van with a question; you provided the research and here is an answer. Hope this helped. -Carl Anchorage, AK
- Carl F (age 28)
Anchorage, AK, USA

Many thanks for these carefully reported results!

Mike W.

(published on 01/27/2015)

Follow-Up #18: freezing milk and water

How hard can this really be?? Wait it I'd hard with all the variables that go into it, from temperature of water to the humidity. There is just so many variables that go into this. So lets ask give Mike a nice round of applause for trying to help! Plus he is a savage! Would milk and water freeze at the same time if they had all the same variables down to a dot?
- Hugh G. Rexcion (age 37)

In equilibirum, the water freezes at slightly higher temperature, because the sugars and salts, etc. in the milk slighty lower it's freezing temperature. (See .) Since in practice solutions can supercool (see ), i.e. stay liquid a bit below the equilibrium freezing point, I can't guarantee that the milk won't freeze first, because the fat droplets might reduce the tendency to supercool. We'd love to hear the results of the experiment.

Mike W.

(published on 03/23/2015)

Follow-Up #19: pure heat conduction

So say we have this magical freezer it is always -20 degrees Celsius (The temperature is never changing) In this magical box there is no gravity and it is a vacuum. So we add a floating sphere of water. Its has a radius of 1cm so its Area is 12.57 cm2 this is about 4.19cm3 The ball of water is 10C. Now lets say this magical ball of water as well because it spreads its temperature through its self evenly (Can be treated as 1 System evenly distributing its energy). Water having a heat capacity of 75.375 0.05 J/molK and the density is 1g/cm3. Now is there any kind of calculation you can do to determine roughly how long it will take for that water reach -20C? Also if there are any more assumptions that need to be done to determine this what would they be? (Assume that the pressure is 1 Atm and constant).
- Mike Watson (age 23)
Vicotira, BC, Canada

Let's forget about the part about magic water that keeps its internal temperature uniform. That removes the only real physics of the problem, which is thermal diffusion. You've made our life easier by removing gravity, which causes convection, a much mesier process.

The water will gradually approach -20°C. If by "reach -20°C" you mean get very close, you have to wait several times the thernal diffusion time, the time it takes for heat to diffuse a distance of about the radius of the sphere.

So you need the thermal diffusion constant, which is the thermal conductivity divided by the heat capacity per unit volume. So that's (0.60 W/m-K)/ 4.2*106 J/m3-K =1.4*10-6 m2/sec. WIth your 1 cm radius, the typical diffusion time is about 100 sec. The exact solution for the time-dependent temperature can be obtained by decomposing the spatial dependence into a bunch of patterns each of which relaxes exponentially toward equilibrium. The pattern with the longest spatial wiggles has the slowest relaxation.

Mike W.

(published on 03/16/2015)

Follow-Up #20: physics, banking, and figuring things out

found the article trying to figure out how long it takes water to freeze and decided to read it. I am very impressed with Mike. He is patient in trying to teach and unfortunately there is not always a straight answer to many things in life due to variables. I am a banker and I see this a lot in banking. Sometimes you have to give a general answer to the public and let them learn the variables for themselves. And you get the response "but, you said!" What I said was "generally". I appreciate Mikes patience and his positive manner. Great thread.
- anna (age 42)
Rochester, NY

We really appreciate these comments!

Mike W.

(published on 06/25/2015)

Follow-Up #21: does cold water freeze faster?

The answer given here about cold water freezing faster than hot water is flat-out wrong and disregards the Mpemba effect.
- Jay (age 33)

The site you refer to has nice discussions of the limitations of some common explanations of the effect whereby sometimes hot water freezes before cold water. Unfortunately the explanation they give, that the heat changes the molecular bonds between the water molecules so the resemble those of ice, is pure baloney, far more than the ideas they criticize. The relaxation time for those little local modes in the water is somewhere crudely in  the neighborhood of 10-12 s, not exactly long enough to help the cooling water remember anything about its past.

The main explanation that they leave out is probably the best. Heating causes dissolved gases to leave the water. Dissolved gasses lower the freezing point. So the heated water has a slightly higher freezing point.

We discuss these issues in another thread.

Mike W.


(published on 08/14/2015)

Follow-Up #22: stop the van!

Mike W. Please stop answering questions.
- Kyle (age 30)

Any special reason? What direction do you want to see the site go?

(published on 08/12/2016)

Follow-Up #23: hot or cold water freezes faster?

It's not baloney to think that when a really hot substance is introduced to a really cold environment that the molecules will behave differently. It's called sublimation, and it is one of the ways that matter changes it's states. That is why boiling water will turn to snow if thrown into negative degree temperatures. Can't do the same with cold water. Now this does not prove that under normal circumstances would hot water freeze faster than cold water, but it does prove that hot water does react differently to freezing temperatures than cold water. I would like to hear you rebuttal my good sir.
- Ricky (age 20)
Corbin, Kentucky

We're basically in agreement that various complicated effects can cause hot water to freeze faster than cold under some circumstances. Perhaps the most common of those is found when the cold surroundings are just a bit below 0°C. Then the freezing rate is very sensitive to the exact freezing temperature, which depends on dissolved non-water molecules. Typically, hot water has fewer dissolved air molecules than cold water. That can make the initially hot water freeze before the initially cold water. Another reason can be that more of the hot water evaporates, leaving less behind. The smaller amount left can cool more quickly than a larger amount. These effects can both operate together. 

You're pointing to another mechanism. That evaporated water gets into very cold air, causing it to condense and then freeze. In effect, the initial heat helped get the water out into better thermal contact with the air, in the form of vapor.

So yes, rates can be complicated.

Mike W.

(published on 01/07/2017)

Follow-up on this answer.