Norfolk off-grid heating map

The headline figures

Those figures come from the Department for Energy Security and Net Zero’s subnational estimates of properties not connected to the gas network, published 18 December 2025 using 2024 data. The full attribution is at the foot of this page and it is a licence condition rather than a courtesy.

What the data actually is, and its limits

It matters to be clear about how these numbers are produced, because it determines what you can legitimately say about them.

1. They are estimates, not a census. DESNZ derives them by subtracting the count of domestic gas meters from Valuation Office Agency domestic property counts. There is no definitive register of off-grid properties anywhere.
2. The publisher states plainly that LSOA figures must not be aggregated to local authority, regional or country level. District figures therefore come from the local authority dataset, never from summing smaller areas. We hold both datasets separately for exactly this reason.
3. A substantial number of Norfolk LSOAs are suppressed for disclosure control. Suppressed areas must be shown as ‘no data’, never as zero – treating a suppression as a zero would understate off-grid density in precisely the sparse rural areas where it is highest.
4. The whole back-series is revised annually. These are 2024 figures published in December 2025, and DESNZ revises historic figures each year. They should be rebuilt when the next release lands rather than left presented as current indefinitely.
5. No gas connection does not mean oil. A property off the gas network may be heated by oil, LPG, electricity, a heat pump, solid fuel or biomass. The data tells you what is absent, not what is present.

What it means for heating in Norfolk

A property with no gas connection faces a different set of decisions from one on the grid, and those decisions are what shape most of our work.

• Fuel has to be delivered and stored. That means a tank, siting rules, containment requirements and the risk of running dry – see oil tank installation.
• Fuel price exposure is different. You buy in bulk, at a moment, rather than paying a metered rate.
• Servicing matters more. An oil boiler drifts out of tune in a way a gas boiler largely does not, and nobody is going to notice until it fails – see what actually happens during an oil boiler service.
• The heat pump question is live, because there is no cheap incumbent fuel to compare against. It is also harder, because much of this housing stock is older and solid-walled – see oil vs heat pump for an off-grid Norfolk home.
• Emergencies are worse. A rural property with no heating in February, at the end of a lane, is a different situation from a terraced house in a town.
• The trade base is different. Fewer engineers specialise in oil than in gas, because in most of the country there is far less of it.

That last point is why we built the business around oil rather than treating it as a sideline. In a county where more than a third of homes have no gas, oil is the mainstream rather than the exception.

What the district picture shows

The pattern within Norfolk is what you would expect from the geography, and it is worth stating rather than implying:

• Rural districts have the highest off-grid density. Gas networks follow population density, so sparsely populated areas were never economic to connect.
• Urban Norwich is largely connected, as you would expect.
• Several Norfolk districts sit in the Great Britain top twenty-five for the proportion of property off the network.
• The coastal and Broads areas include a high proportion of older housing stock and second homes, both of which change the heating picture.

We hold the full district-level dataset and the LSOA dataset, and the interactive map is built from them. Where an LSOA is suppressed it renders as no data.

Using this data yourself

It is open data under the Open Government Licence and you are free to use it, subject to attribution. If you are doing so:

1. Go to the source rather than taking figures from a secondary site, including this one.
2. Use the LA dataset for district figures and the LSOA dataset for small-area analysis. Do not aggregate LSOAs.
3. Render suppressions as no data, not as zero.
4. Say ‘estimated’.
5. Check the publication date, because the series is revised annually.
6. Attribute it, which is a licence condition.

The landing pages for both datasets are linked in the sources below.