Quick answer

How long does a flask stay hot in winter UK? A half litre vacuum flask that only just passes BS EN 12546-1 must still be at 70°C or more six hours after a full, preheated fill in a 20°C room. On a 2°C morning, a simple cooling estimate puts that flask at 70°C after about four and a half hours and near 64°C at six, still above the 60°C that one flask maker counts as hot. How full it is matters more: in one published kitchen test of a 500 ml flask, it was 18 degrees cooler after three hours holding 140 ml than when full.

Table of contents

You filled it at seven, poured a cup at eleven, and got something warm rather than hot. The question is which of the things you control went wrong.

Four checks on your flask and your morning

None of these needs a thermometer.

1. Is it a vacuum flask, or only double walled?

Thermos Brand publishes a home check: "The sweat test is a reliable quick check: hold a cold-liquid-filled bottle. If moisture forms on the exterior, the bottle is not true vacuum insulated. In a properly vacuum-insulated bottle, the outer wall stays completely dry because no heat is transferring through the vacuum to cause condensation." Thermos Brand's food jar guide adds a test with something hot inside: "if the exterior is warm to the touch, the vacuum seal may have been compromised". Feel the body rather than the neck.

BS EN 12546-1:2000 treats vacuum and non-vacuum flasks as two products. Clause 2.1.1 defines a vacuum insulated container as an "insulated container in which the insulant is a vacuum", and clause 2.1.2 a non-vacuum insulated container as an "insulated container in which the insulant is not a vacuum". For 401 to 600 ml, Table 3 wants the vacuum version no lower than 70°C six hours after filling. Table 4 asks 40°C. Both get sold as double walled.

2. What the box claims, and what it leaves out

BS EN 12546-1 never defines hot, so each brand sets its own cut off. Stefan Marusic, a product development engineer at Sigg, told Galaxus that Sigg stops counting below 60°C: "But that's not standard. If a competitor wants to test down to 45 degrees because they define that as hot, they'll get a better result - on paper."

Read the conditions. Thermos UK badges its Stainless King Flask 470ml "18 HOURS HOT" and gives no ambient temperature, start temperature or fill level anywhere. Decathlon does the opposite for its 0.5 litre Quechua flask: "We've tested this flask in the lab according to standard EN 12546-1, at an ambient temperature of 20°C while filled to capacity. The time it takes to drop from 95°C to 50°C (drinks temperature considered warm enough to drink) is 14 hours."

3. How full you are filling it

Fill level is the check with the clearest measurement behind it. First Steps Nutrition Trust published a February 2016 statement using one 500 ml flask, warmed for a minute with boiling water, at three fill levels, ambient about 19°C. Filled, the water was 92°C after thirty minutes and 76°C after three hours. With 140 ml in it, 72°C and 58°C: 20 degrees behind the full flask within half an hour.

Full means the standard's nominal capacity, clause 2.3's "a level 10 mm below the lowest part of the inserted closure", not the brim. A Thermos care and use sheet makes the same point as a warning: "DO NOT overfill your flask. Leave sufficient space in the neck to fit the stopper."

4. Where it will be, and for how long

Heat leaves at a rate set by the gap between drink and air: the laboratory starts at a gap of 75 degrees, a 2°C morning at 93. If you will finish a vacuum flask within four hours of filling, the six hour figure was never your problem. To learn what yours does, fill it the way this page describes on a cold morning, leave it outside and pour a cup at the time you would usually stop.

October to January, month by month

Figures below are Met Office 1991-2020 averages for two very different stations, Heathrow and Capel Curig No 3.

The adjustment rule is that gap. Subtract the air temperature from roughly 95°C and compare it with the 75 degrees the laboratory starts from. A 2°C morning makes the gap about a quarter bigger, and on the simplest cooling model, Newton's law of cooling with nothing else changing, a half litre flask that just meets the standard reaches 70°C after about four and a half hours instead of six. At Capel Curig's October mean minimum that is just under five hours, and at its January one just over four and a half, so the calendar moves the answer by about twenty minutes; fill level moves it by far more. Treat all of that as an estimate: we could trace no published test of a consumer flask below about 13°C, and no source supports a fixed percentage penalty for cold.

October

Little has changed. Heathrow's October mean daily maximum is 15.81°C against a mean daily minimum of 8.78°C; Capel Curig runs 12.77°C and 7.09°C. At a stated 13°C, every flask on Outdoors Magic's page with a published result was above 62°C after six hours, opened twice before the reading, so a full, preheated vacuum flask as good as those, filled at eight, should still pour hot at two.

November

Heathrow falls to 11.47°C and 5.26°C, Capel Curig to 9.75°C and 4.50°C.

December

Heathrow averages 8.79°C and 3.08°C with 7.17 days of air frost; Capel Curig 7.58°C and 2.40°C.

January

Heathrow sits at 8.42°C and 2.68°C, Capel Curig at 7.03°C and 1.91°C. Between these two stations the bigger January difference is wind: Capel Curig No 3, at 216 m above sea level, has a January mean wind speed at 10 metres of 15.52 knots (about 18 mph) against Heathrow's 8.15 (about 9 mph), while its January mean daily minimum is only 0.77 degrees lower.

A vacuum defeats conduction and convection through the wall, so wind does not act on a flask the way colder air does: the outer wall already sits near ambient, and moving air over it shifts little heat.

Wind acts on what the vacuum does not cover, the cap, the neck, the collar and the open mouth while you pour. We could trace no published test of a flask in wind against still air, so treat that as reasoning, though it costs nothing to keep the flask inside your pack and to pour out of the wind, away from your lap. Feels like temperature is no help: the Met Office "calculates wind chill by measuring how much heat is lost from a person's bare face at a walking speed of 3mph". It describes you, not the flask.

Filling a flask so it holds

Preheat it, and argue about the size of the gain later

Clause 6.4(e) of BS EN 12546-1:2000 lists this among the statements a conforming flask's instruction leaflet must contain: "Pre-heat or pre-cool the container with water to ensure efficient use and to reduce the risk of breakage of the glass filler. Never use microwave or conventional ovens for this purpose". Half that reason is protecting a glass filler, which a flask that is stainless steel inside does not have, so if yours has glass inside, never skip it. Zojirushi and Stanley 1913 say five minutes.

How much it buys is disputed: hours against minutes. Thermos Brand's blog puts the gain at one to three hours for a preheated food jar. The one independent measurement we could find, a 2013 undergraduate project at the British Columbia Institute of Technology by Chu, logged 250 g of macaroni cheese thirty times in each of three food jars, preheated and not, and called the gain "only 10 to 15 minutes", though statistically significant. Both concern food jars. Preheat anyway, and boil the kettle again while the flask stands, so what goes in the moment you empty it is straight off the boil.

Fill it hot, and fill it full

Fill it as late as you can: the standard's six hours run from the moment the water goes in. Clause 5.4 tests from water at 95°C or above, so every retention figure from the standard assumes water straight off the boil. Anything cooler starts you below the line the number was measured from, which is why a brew that has stood never behaves like the box.

Then fill to the nominal capacity, about 10 mm below where the bottom of the stopper will sit, and never to the brim. Contigo's AUTOSEAL travel mug instructions add a caution: "If filling with boiling liquids, allow liquid to cool down for at least 3 minutes before sealing with the lid." Contigo gives no reason, though the same sheet says "The button should be depressed to release pressure, each time the lid is screwed on". Follow your own flask's sheet, especially if it has a push-button or sliding drinking lid.

Seal it, carry it shut, pour into a cup

If your flask came with two stoppers, check which one the maker means for heat: Outdoors Magic lists the Primus Trailbreak EX 1 Litre with an "Additional standard stopper for optimum heat retention included". Stanley 1913's FAQ, on the moment after filling: "secure the stopper and lid as quickly as possible to avoid heat loss."

Then leave it shut. Thermos Brand's food jar guide: "Every time you open the lid, you release heat and allow cooler air in." We could find no measured figure for one pour. Each cup you pour also leaves the rest in a less full flask, which on the First Steps Nutrition Trust numbers cools faster, so the last cup is the coolest. If the last cup matters, fill a second flask and leave it shut until then.

Pour into a cup rather than drinking from the flask. The Thermos care and use sheet quoted in check 3 gives the reason: "To prevent scalding always pour contents into a cup. DO NOT drink directly from the flask." If the cup has to wait, a lidded insulated mug slows the cooling; our review of an insulated camping mug with a spill proof lid explains why you leave its slider open for a minute before closing it over a hot drink.

What goes wrong, and what to do

Warm at eleven

Start with how full it was (check 3), and remember that every cup you pour leaves the rest less full. Then ask what went in, how long before the walk you filled it, and how often it was opened. If a flask that used to hold heat has stopped, check the stopper's seal is in place, then run the hot-water test in check 1, and if it points to a failed vacuum, replace the flask. None of it is fixable on the hill.

The 24 hour claim

Outdoors Magic's flask page, bylined Matt Jones, describes its method: each flask primed five minutes, refilled at 90°C, left outdoors at a stated 13°C with wind speed recorded as zero, and reopened four times over eight hours. Its Thermos Ultimate 500ml, a flask Thermos now markets as 24 hours hot, read 50.9°C at 24 hours, below Sigg's 60°C line for hot. Read an hours claim as that brand's own idea of hot, not a promise and not a ranking: Galaxus found that Sigg's 60°C limit "means that some bottles will look worse than the competition." On the estimate above, a flask that only just passes the standard would be down to about 43°C twelve hours after a bedtime fill on a 2°C night.

Scalds, soup and the 63°C figure

The performance you are maximising is the hazard, and Decathlon says so on its product page: "Warning: the flask is very insulating, the liquids will remain hot for a long time." The Child Accident Prevention Trust says "A hot drink can scald a baby even 15 minutes after it's been made." A full flask pours at about 90°C for the first hour and was still at 76°C after three hours on the First Steps Nutrition Trust numbers, so keep it away from children, and cool a child's drink with cold water and test it yourself before you hand it over.

If it goes wrong, NHS guidance reviewed 31 March 2026 says to treat all burns and scalds straight away: "Hold the burn or scald under cool running water for 20 minutes. Do this as soon as possible, within 3 hours of it happening. If there's no running water available, you can use cool bottled water or a wet towel." NHS inform in Scotland adds: "Never use ice, iced water, creams or greasy substances like butter on a burn." and "Keep yourself or the injured person warm. You can use a blanket or layers of clothing to do this. But, avoid putting these on the injured area." The NHS page also says to "Remove any clothing or jewellery that's near the affected area - but do not remove anything that is stuck to the affected area." Then: "When the burn or scald has cooled, lay cling film over it if you can. Do not wrap the cling film around it."

The NHS says "Call 999 or go to A&E if" the burn or scald "is very large or deep", "is on your face, genitals or bottom" or "has been caused by an acid or chemical, or by electricity", and to get help from NHS 111 if "you or your child has a burn or scald and you're not sure what to do". It adds: "Call 111 if you need advice for a child under 5 years old." For A&E: "Do not drive to A&E. Ask someone to drive you or call 999 and ask for an ambulance."

The 63°C figure you may have seen is a rule for food businesses. In England it is paragraph 6 of Schedule 4 to the Food Safety and Hygiene (England) Regulations 2013, which applies to a person acting "in the course of the activities of a food business" who keeps food "at or in food premises" that "is for service or on display for sale". A walker with soup is not that person, but the rule exists for food that "needs to be kept at or above 63 °C in order to control the growth of pathogenic micro-organisms or the formation of toxins". Chu's food jars stayed above 60°C for less than three hours, so bring soup to the boil before it goes in and eat it at your first stop rather than saving it for later.

Frequently asked questions

Does cold weather actually make much difference?

On the physics it should: Newton's law of cooling says heat leaves faster the bigger the gap between drink and air, and on the estimate above a half litre flask that just meets the standard reaches 70°C after about four and a half hours on a 2°C morning instead of six. We could find no test of one flask at two air temperatures, so treat that as an estimate. Thermos Brand's blog says "Real-world performance often exceeds the published figures, particularly in cooler environments or when the bottle is pre-treated before filling." It never says whether that means hot contents or cold ones, for which a cooler room does help.

Is a bigger flask worth carrying?

Yes, if you will drink it. Table 3 of BS EN 12546-1:2000 asks 60°C at six hours from a vacuum flask of up to 200 ml and 80°C from one over 1200 ml. First Steps Nutrition Trust measured a full 1 litre flask at 86°C after three hours, against 76°C for a 500 ml.

Can I put tea with milk or hot chocolate in it?

Not recommended. Clause 6.4 of BS EN 12546-1:2000 says a container "shall be supplied with an instruction leaflet" that includes, at (c), "In order to minimise bacterial growth, do not use to keep warm milk products or baby food". Zojirushi's FAQ for its vacuum mugs goes further: "It is not recommended to fill the vacuum mug with any type of beverage containing dairy or carbonated drinks", because dairy "can cause curdling and leave residual odors". Carry the milk separately, and for hot chocolate carry hot water and stir the powder into the cup. Carrying boiled water for a baby's formula is a different job with its own rules, and the 70°C estimates on this page are no guide to it, so follow the NHS advice on formula feeding away from home.