Leaving the grid entirely is a winter problem, not a battery problem

The grid is doing a job that is easy to overlook

A connected house treats the network as an infinitely large battery that never runs out, accepts surplus whenever there is any and supplies whatever is needed at any hour. Nothing in a domestic installation replaces that, and the standing charge that pays for it is often described as a cost of doing nothing.

It is not nothing. It buys the ability to be wrong about your own consumption, every day, without consequence. Removing it means every mismatch between what the roof produced and what the house wanted has to be absorbed by hardware you own.

That is the whole difficulty, and it is a seasonal one rather than a daily one.

Autonomy is measured in the worst week, not the average

Sizing a grid-connected battery against a typical evening is reasonable, because a shortfall simply means importing. An off-grid system has no such fallback, so it has to be sized against the run of dull short days that occurs somewhere in every winter, and the array has to be large enough to refill it afterwards.

The relationship is brutally non-linear. Covering the average day is straightforward, covering a normal bad week costs a great deal more, and covering the rare fortnight of persistent cloud costs more again for capacity that sits idle for years at a time.

Consumption has to be examined just as carefully, because an off-grid household cannot treat any load as a small percentage of a large total. Every appliance is being sized against a finite supply, so the design almost always pushes hard towards efficiency before it pushes towards capacity, on the sound principle that a unit never needed costs nothing at all to store.

This is why off-grid designers talk about days of autonomy rather than capacity, and why the honest answer to how many days is enough is usually more than you were hoping.

The seasons pull in the wrong direction

At higher latitudes, the array produces a small fraction of its summer output in midwinter, while heating and lighting demand peak in exactly those months. Solar generation and household demand are close to inverted across the year, which no reasonably sized battery can bridge.

Storing summer surplus for winter would require capacity orders of magnitude beyond anything domestic, because the losses of holding energy for months are not the main obstacle. The volume of it is. Seasonal storage is a genuine engineering problem and it doesn’t have a household answer today.

Nearer the equator the seasonal swing is milder and the picture improves considerably, which is one reason off-grid living is far more practical in some places than others.

Which is why almost every off-grid system has a generator

The standard solution is a second source used rarely: a fuel generator, sometimes a small wind turbine where the site suits it, occasionally a connection to a neighbouring supply. It runs for a handful of days a year and it is what allows the battery and array to be sized sensibly rather than absurdly.

That changes the character of the project. There is fuel to store, a machine to maintain, noise to consider and emissions to account for, and the resulting system is a hybrid rather than a self-sufficient one. Many people find that disappointing after the fact.

It is, though, the arrangement that works, and a design without a backup source is usually a design that hasn’t confronted its worst week.

Where it makes sense, and where it plainly does not

Off-grid is the right answer when connection is genuinely unavailable or the cost of extending a supply to a remote site is prohibitive. In that situation the comparison is against trenching a line across a field, and the storage system wins easily.

For a house that already has a working connection, disconnecting is very rarely justified. The equipment costs far more than the standing charge it avoids, the household loses the ability to export surplus, and the reliability of a domestic system is not comparable to a network with people paid to maintain it.

A partial version exists and is frequently the better outcome. A house at the end of a weak rural line, where outages are frequent and long, can install generation and storage sized to carry it through those interruptions while keeping the connection for the months when the roof cannot cope. That house gets most of the resilience without taking on the seasonal problem at all.

There are non-financial reasons people choose it, and those are theirs to weigh. But as an economic decision for a connected house, it does not stand up, and it is worth saying so plainly rather than leaving the arithmetic to be discovered later.