What this calculator does#
It splits the job ice does into its two real parts and prices each one: cooling your food and drink down to temperature once, then absorbing the heat that leaks in every hour afterwards. The result is a mass of ice, an ice to contents ratio, and the ceiling on how long this particular box can work in this particular weather.
How the maths works#
Wall leak. Heat flow through the walls is conductivity divided by thickness, times area, times the temperature difference. We use 0.026 W per m per K for closed cell foam (0.007 for vacuum panels), and the wall thickness implied by your cooler type: 12 mm soft bag, 18 mm basic hard cooler, 30 mm mid-range, 57 mm rotomoulded. Area is the effective conduction area, midway between the inner and outer surfaces, estimated at 7.4 times the two-thirds power of the volume in cubic metres. For a 50 litre box that is about 1.0 square metre, against an outer surface of roughly 1.2, and using the outer figure is why many published estimates run high.
Sun. Direct sun and a parked car do not change the air temperature but they do change the surface temperature of the box, so we model them as an increase in effective ambient: shade 0, part sun 3 degrees C, direct sun 8, inside a parked car 12.
Pull-down. Cooling the load once takes its mass times about 3.6 kJ per kg per K times the drop to 5 degrees C, the temperature a cooler should hold. Pre-chilling the contents and the box overnight removes about three quarters of it.
Lid openings are charged at 8 kJ each, which accounts for the contents surface re-warming rather than for the small mass of air exchanged.
Ice mass is the total energy divided by 334 kJ per kg. Block ice gets a 15 percent slower melt rate for its lower surface area.
What the model leaves out#
Lid seal quality and the drain plug, which are real and vary between coolers. Warm food added mid-trip. Direct radiative heating of a dark lid, which we approximate rather than model. Melt water that you drain away, which was still doing useful work as cold mass.
Treat the answer as a floor and round ice purchases up. If the answer says 6 kg, buy 8.
Frequently asked questions#
What is the right ice to food ratio?#
Around 2 to 1 by mass for multi-day trips in warm weather, less in the shade or for shorter trips. The calculator gives you the number for your own conditions rather than a rule, because the difference between shade and direct sun can be several kilograms.
Should I drain the melt water?#
Usually not. Cold water still absorbs heat and it fills the air gaps that would otherwise convect. Drain only if packaging is at risk of getting soaked, or if you need the space.
Is block ice better than cubes?#
Block ice melts more slowly for the same mass because it has less surface area, so it wins for duration. Cubes chill a warm load faster because they make better contact. Carrying both, blocks at the bottom and cubes packed around the food, gets you both effects.
Does pre-chilling really matter?#
Yes, and it is free. The pull-down job can be a third or more of the total ice budget on a hot day. Chilling the box and the contents overnight in a fridge removes most of it, which is often the difference between three days and four.
When is a 12 volt fridge worth it over a cooler?#
Roughly when the trip is longer than four or five days, when ice is inconvenient to buy, or when you are already carrying a battery for other reasons. The energy figures and the crossover are in 12 volt power and fridges.
Why does my expensive cooler still lose ice quickly?#
Almost always sun exposure and lid openings rather than the box. A rotomoulded cooler in direct sun with the lid opened twenty times a day can lose more ice than a mid-range cooler in the shade that stays shut.
Standards, sources and further reading
- Latent heat of fusion of ice, 334 kJ per kg, and the thermal conductivity of closed cell polyurethane foam, approximately 0.03 W per m per K.
- Steady state conduction, q = (k / thickness) x area x temperature difference. Surface area is estimated from capacity as described below.
How this page is made. Every number here is either a published standard, a physical constant, or arithmetic we show in full so you can check it. Read our evaluation method and editorial standards, or tell us we got something wrong.
Last reviewed and updated 9 September 2026.