Three Pathways and a Counterfactual: the British Home to 2050

Compare NESO’s three pathways to net zero for home heating, electricity demand and vehicle-to-grid flexibility. These are models, not forecasts or measurements of what has happened.

Three pathways to net zero

The three pathways that reach net zero by 2050 are Holistic Transition, Electric Engagement and Hydrogen Evolution. They make different assumptions about how homes stop burning gas and how people use electricity.

There is no central case, so the pathways are not averaged. Falling Behind is shown separately because it is a counterfactual in which the net-zero target is missed.

Gas boilers

The number of homes with a gas boiler, at the end of each year. This is the measure the three pathways disagree about most starkly, because almost everything else in a home follows from how it is heated.

Air source heat pumps

The same count for air source heat pumps, the technology every pathway expects to do most of the replacing. Ground source heat pumps, hybrids and heat networks are all counted separately by NESO and are not included in the figures below.

Hydrogen boilers, in one pathway of three

Hydrogen for home heating is the clearest illustration of what a pathway is. It is not a thing that might happen a bit in every future: NESO models it in one pathway and publishes no series for it at all in the others. A table with a single column is the honest shape of that, and it is the reason the column is named.

What homes draw from the grid

Total residential electricity demand across Great Britain, in gigawatt-hours. It rises in every pathway, because electrifying heat and driving moves energy onto the grid that used to arrive as gas or as petrol — and it rises by very different amounts, because that is what the pathways are about.

NESO’s years here are financial ones, April to March, so a column headed 2030 is 2030/31. The release also breaks this total into appliances, lighting, heat pumps, electric vehicles and the rest, and those parts are the total divided up rather than added to it — which is why only the total is shown.

What a household could give back

An electric car is a large battery that spends most of its life parked. Vehicle-to-grid lets it push power back out at the moment the system is tightest, and NESO models how much of the winter peak that could take off. The figures are published as negative numbers because they are subtractions from peak demand, and we leave them that way.

Read them as a modelled potential rather than as a plan. They depend on how many electric cars exist in that pathway, how many are plugged in at the wrong time of a winter evening, and whether their owners have any reason to let the grid use them.

Falling Behind: the future nobody is planning for

The three pathways above all arrive at net zero by 2050 and disagree about how. Falling Behind is the fourth thing NESO publishes and it is not a fourth route to the same place: it is the counterfactual, the version of the next twenty-five years in which the transition loses momentum and the target is missed. It leaves the country still emitting on a large scale in 2050.

It is here because it is the most useful comparison on the page. What separates a home in Falling Behind from a home in Holistic Transition is roughly the size of the job still to do. What it is not is one of the futures being planned for, which is why it has its own table rather than a column beside the three — a fourth column would make the set look like a range, and a range has a middle.

If the name is unfamiliar, the 2024 edition called it Counterfactual. It is the same thing under a new label.

None of this has happened yet

Everything above is modelled. It is worth saying plainly, because the figures look exactly like the measured ones elsewhere on this site: the same units, the same technologies, the same country.

The measured picture — heat pumps and solar actually installed, grants actually paid, and what the grid actually did — is on solar and heat pump deployment and live electricity. We keep them apart on purpose. A modelled 2050 figure and a counted 2024 one on one axis would draw a line between them, and that line would not be a trend: the two are built on different definitions, different counting bases and different vintages.

For the same question at a local scale, from a network operator’s own modelling rather than this one, local energy projections answers council by council. Each council there is answered by the one network operator that serves it, on that operator’s own model and assumptions, so those figures are not a share of these and are not added to one another either.