In a home freezer at the recommended −18 °C (0 °F), a plastic ice-cube tray of tap water freezes solid in roughly 2 to 7 hours, with about 3 to 4 hours a reasonable middle estimate, while a 500 mL (17 oz) bottle takes much longer: roughly 6 to 20 hours. These ranges are SciQuest estimates from a heat-transfer calculation, not laboratory measurements. The main factors are how much water you freeze, its shape and how strongly cold air moves around it.
Fresh water freezes at 0 °C (32 °F), but it only turns to ice when its surroundings are colder than that. Most of the waiting time goes into removing the hidden “latent” heat that water must lose to change from liquid to solid.
Key facts about freezing water
- Freezing point: 0 °C (32 °F) for fresh water at normal air pressure. Salt lowers it: polar seawater starts to freeze at about −1.8 °C (28.8 °F).
- Recommended freezer setting: −18 °C (0 °F), according to the US Food and Drug Administration.
- Latent heat of fusion: about 334 kilojoules must be removed from each kilogram of water at 0 °C to turn it into ice.
- Where the time goes: for water starting at room temperature, about 80% of the heat removed goes into the change from liquid to ice, not into cooling the water to 0 °C.
- Hot versus cold: claims that hot water freezes faster (the Mpemba effect) remain disputed.
How long does it take water to freeze in a freezer?
The table shows estimated times for water starting at 20 °C (68 °F) to become completely solid in a −18 °C freezer. Each column uses a different heat transfer coefficient, a number that describes how quickly moving or still air carries heat away from a surface. Engineering course notes give 3 to 20 watts per square metre per kelvin (W/m²·K) for air moving only by natural convection and 30 to 300 W/m²·K for fan-driven air.
| Container (water volume) | Still air, h = 5 W/m²·K (hours) | Moderate air flow, h = 10 W/m²·K (hours) | Brisk air flow, h = 20 W/m²·K (hours) |
|---|---|---|---|
| Plastic ice-cube tray (about 350 mL in total) | about 7 | about 3.5 | about 2 |
| 500 mL plastic bottle, lying on its side | about 20 | about 10 | about 6 |
| 1 L plastic tub, 15 × 15 × 4.5 cm | about 28 | about 14 | about 8 |
| 2 L plastic bottle, lying on its side | about 33 | about 18 | about 10 |
SciQuest estimates. Assumptions: freezer air at −18 °C; water starting at 20 °C; time until all the water has turned to ice (not until the ice reaches −18 °C). Cooling to 0 °C uses Newton’s law of cooling; the freezing stage uses Plank’s equation (as described by the UN Food and Agriculture Organization) with a thermal conductivity for ice of 1.6 W/m·K. The tray is treated as a thin slab cooled from top and bottom; bottles as cylinders; the tub as a slab cooled from top and bottom only. The estimates ignore supercooling, the container walls, contact with cold shelves, door openings and defrost cycles. Real times can differ substantially, so treat them as a guide.
The pattern matters more than any single number. Bigger containers hold more water for each square centimetre of surface, and heat from the centre must pass through a growing layer of ice. The FAO notes that doubling the thickness of a block being frozen can more than double its freezing time.
How much energy does it take to freeze water? A worked example
Take 250 mL of tap water at 20 °C. Its mass is about 0.250 kg, because water has a density of about 1 gram per cubic centimetre.
- Cool the water to 0 °C. Heat removed = mass × specific heat × temperature change = 0.250 kg × 4,186 J/(kg·K) × 20 K = 20,930 J, or about 20.9 kJ.
- Turn the water into ice at 0 °C. Heat removed = mass × latent heat of fusion = 0.250 kg × 334,000 J/kg = 83,500 J, or 83.5 kJ.
- Total to make solid ice at 0 °C: 20.9 kJ + 83.5 kJ = 104.4 kJ. The freezing step is 83.5 ÷ 104.4 ≈ 80% of the total.
- Optional: chill the ice to −18 °C. Using a specific heat for ice of about 2,050 J/(kg·K): 0.250 kg × 2,050 J/(kg·K) × 18 K ≈ 9.2 kJ more.
That is the key insight: changing water into ice takes four times as much heat removal as cooling the same water from room temperature all the way to 0 °C. If a freezer pulls heat out of this water at an average of 5 watts (5 joules per second), 104.4 kJ would take about 20,900 seconds, or 5.8 hours. At 10 watts, it would take about 2.9 hours. The heat-removal rate is the part that varies most from one freezer and container to the next.
Why doesn’t water freeze if the air is exactly 0 °C?
Heat only flows from a warmer object to a colder one. If water at 0 °C sits in surroundings at exactly 0 °C, there is no temperature difference, so there is no net flow of heat out of the water. Without removing its latent heat, the water cannot turn completely into ice, however long you wait.
That is why freezing speed depends on how far below 0 °C the surroundings are. While the water is turning to ice, it stays near 0 °C, so a −18 °C freezer gives a temperature difference of about 18 K. A colder freezer increases that difference and speeds freezing.
What makes water freeze faster or slower?
- Amount of water: more water means more latent heat to remove.
- Shape: a shallow, wide layer freezes faster than the same volume in a deep, narrow container, because it has more surface for its volume and thinner ice for heat to pass through.
- Air movement: faster-moving cold air removes heat more quickly, according to FAO guidance on freezing food. A crowded freezer can block air flow.
- Freezer temperature: a colder setting increases the temperature difference that drives heat out.
- Starting temperature: warmer water has more heat to lose first. The FAO notes that chilling a product before freezing shortens freezing time.
- Lids and trapped air: air is a poor conductor of heat. In one FAO example, removing a box lid cut the freezing time of smoked fish from 15 hours to 8.
- Container material: aluminium conducts heat roughly 250 times better than ordinary glass. A metal tray resting directly on a cold metal surface can therefore help, although in our estimate the slow transfer of heat into still air is usually the bigger bottleneck.
At what temperature does water freeze?
Fresh water freezes at 0 °C (32 °F) at normal air pressure. Dissolved salt lowers the freezing point. According to the US National Snow and Ice Data Center, each increase of 5 practical salinity units lowers it by about 0.28 °C, and seawater in polar regions, at a salinity of about 32, begins to freeze at about −1.8 °C (28.8 °F).
Water also expands as it freezes, which is why ice floats. Leave space at the top of any container, and avoid freezing a sealed, completely full glass bottle, which the expanding ice can crack.
What is supercooling?
Water does not always start freezing the moment it reaches 0 °C. Ice needs a starting point, called a nucleation site, such as a speck of dust, a scratch in the container or a tiny ice crystal. Without one, water can cool below 0 °C and stay liquid; this is called supercooling.
How far water supercools varies from sample to sample. In clouds, tiny pure droplets can stay liquid to very low temperatures: atmospheric models typically assume they do not freeze on their own until about −40 °C, though research suggests some freezing begins closer to −30 °C.
Does hot water freeze faster than cold water?
Usually not, and the evidence that it ever does is disputed. The idea is called the Mpemba effect, after E. B. Mpemba, who described it with D. G. Osborne of University College, Dar es Salaam, in Tanzania, in a 1969 paper. The claim is that water starting hot can sometimes freeze before water starting cold.
A 2016 study by H. C. Burridge and P. F. Linden reviewed earlier evidence and ran their own controlled cooling experiments. They concluded that there was no evidence supporting meaningful observations of the effect. Other researchers point to supercooling as a possible explanation for occasional reports: in a 2010 preprint, James D. Brownridge reported that hot samples froze first when they started freezing at a higher temperature than the cold samples. A 2025 preprint by Andrei Klimov and Alexei Finkelstein, not yet peer reviewed, reported the effect only when the freezer was very close to the temperature at which ice nucleates and attributed it to the random timing of ice formation.
Separately, physicists have demonstrated a Mpemba-like effect, in which a system starting hotter cools faster, in a carefully controlled colloidal system (Kumar and Bechhoefer, 2020). That result is about a microscopic particle in water, not about freezing a glass of water. The simple energy budget points the other way: in the worked example above, starting the same 250 mL at 80 °C instead of 20 °C adds about 62.8 kJ of heat to remove. For a deeper look, see SciQuest’s article on whether hot water freezes faster than cold water.
What these estimates can’t tell you
Freezer performance varies, and air flow can differ from one freezer, or one shelf, to the next. The heat transfer coefficient, the most uncertain input, could fall outside the range in the table, and the model ignores supercooling, which can delay the start of freezing.
The most reliable answer for your freezer is a measurement. The FDA recommends an appliance thermometer to check that a freezer actually holds −18 °C (0 °F). Note the time you put water in, check it every 30 minutes and record when it is frozen all the way through.
Curious why water looks blue in deep ice and oceans? Read why water is blue.
Related reading
- Will hot water freeze faster than cold water?
- What is underneath the ice in Antarctica?
- Why do water bottles have an expiration date?
Update note: This article was substantially revised on October 12, 2026, to correct the claim that water freezes in surroundings at exactly 0 °C (it cannot, because no heat flows out), correct the definition of the Mpemba effect (the claim is that hot water can freeze first) and label it as disputed, remove the unsupported 30-minute freezing time and contradictory statements about container size, add supercooling and salt-water freezing points, and add a worked energy calculation and a table of estimated freezing times with stated assumptions. It was drafted with AI assistance and checked against the sources listed below.
Sources
- US Food and Drug Administration, Are You Storing Food Safely?
- HyperPhysics (Georgia State University), Phase changes and latent heat
- HyperPhysics (Georgia State University), Specific heat and table of specific heats
- HyperPhysics (Georgia State University), Thermal conductivity table
- HyperPhysics (Georgia State University), Density of water
- University of Waterloo, ECE 309 course notes, Convection Heat Transfer
- Food and Agriculture Organization of the United Nations, Freezing and refrigerated storage in fisheries: 5. Freezing time
- National Snow and Ice Data Center, The science of sea ice
- Herbert, Murray, Dobbie & Koop (2015), Sensitivity of liquid clouds to homogenous freezing parameterizations, Geophysical Research Letters 42, 1599–1605
- Mpemba & Osborne (1969), Cool?, Physics Education 4, 172
- Burridge & Linden (2016), Questioning the Mpemba effect: hot water does not cool more quickly than cold, Scientific Reports 6, 37665
- Brownridge (2010), A search for the Mpemba effect: when hot water freezes faster than cold water, arXiv preprint
- Kumar & Bechhoefer (2020), Exponentially faster cooling in a colloidal system, Nature 584, 64–68
- Klimov & Finkelstein (2025), The Mpemba effect in pure water has a stochastic origin, arXiv preprint (not peer reviewed)
Discover more from SciQuest
Subscribe to get the latest posts sent to your email.

Leave a comment