Power outage planner: how long food, heat and comms last

About four hours with the door kept shut and a room at normal temperature — that is the official figure. But it depends heavily on conditions: in a cool room it roughly doubles, in a heatwave it roughly halves, and every door opening lets in a volume of warm air worth about a quarter of an hour. The tool applies those corrections instead of averaging them away.

A full one for about two days, a half-full one for about one. The difference is not the motor but the latent heat of fusion: melting ice takes several times more energy than warming the same mass by a few degrees, and a full compartment simply contains more ice. That is also why loading the freezer before an expected outage works — even with bottles of water.

No, and the calculation explains why better than any warning. Two healthy burners in a forty-cubic-metre kitchen with sealed double glazing settle at roughly 68 ppm of carbon monoxide. A certified household alarm is required to sound at seventy, and the NIOSH threshold of 35 is crossed in under an hour. That is a healthy hob; with a yellow sooty flame the concentration rises tens of times. A hob is designed for cooking with the extractor running, not for hours of heating.

For carbon monoxide, barely: even six candles produce single-digit ppm, which is nothing. The danger of candles is entirely elsewhere, and half the advice on the internet frightens people about the wrong thing. Candles are a fire risk — knocked over in the dark, near curtains, left burning overnight. If you use them, stand them in a glass jar on a non-flammable surface and put them out when you leave the room.

An attached garage counts as indoors. By the calculation, an ordinary petrol generator in a closed volume of about forty-five cubic metres drives the air to a lethal concentration within minutes, and the equilibrium runs into thousands of ppm. This is not "poorly ventilated", it is incompatible with life. A generator runs outdoors only, well away from windows and doors.

The decisive factor is not the walls but the neighbours. If the heating is out for the whole building, a mid-floor flat at minus ten outside drops to sixteen degrees within hours and keeps falling. If only your flat lost heat, the neighbouring flats keep warming you through the walls and the same flat stays comfortable far longer. A detached house cools faster than any flat: all of its surfaces face outdoors, not one wall and some windows.

When the equilibrium temperature — the value the building is heading towards — drops below roughly three degrees. The tool calculates it separately from the cooling curve, because what matters is not the speed but where the movement stops. If equilibrium is eight degrees, freezing is not a risk however long the outage lasts; if it is minus two, it is only a matter of time.

Because 20,000 mAh is quoted for internal cells at 3.7 volts, while the bank delivers five volts through a converter with losses, and the phone itself gets warm while charging and loses more. Together, about 80% of the original energy reaches the phone battery. That is not a manufacturer's trick, it is physics that nobody prints on the box.

Base stations typically run on batteries for four to eight hours; larger sites last longer and often have generators. The practical conclusion is one: make the important calls and download offline maps in the first hours, not when the phone and the network run out together. A call costs less battery than hunting for a weak signal, which is why a phone in a fringe area drains three times faster than usual.

Roughly three litres of drinking water per person per day, four in heat, and about fifteen litres for all uses. For three days and two people that is eighteen litres of drinking water and ninety in total. The easiest reserve is made in the first minutes: fill the bath and every container while there is still pressure in the pipes — in a block of flats the pumps stop with the lights.

Anything that kept ice crystals or stayed below four degrees can be refrozen — quality suffers, safety does not. Anything above four degrees for more than two hours goes in the bin, and smell proves nothing: the dangerous bacteria have no smell. Above thirty-two degrees the two-hour rule shortens to one.

It is an order-of-magnitude model, not a measurement. Precision is unreachable: every home has its own windows, wind exposure and floor, and every fridge its own age of door seal. But the decisions here are made by order of magnitude — "sixteen degrees by morning" and "four degrees by morning" are different decisions, and the model tells those apart confidently.