Dataset Overview
This dataset estimates the suitability and cost of replacing existing gas boilers with hydrogen boilers for properties connected to the gas network. It covers both domestic and non-domestic buildings.
A property is considered suitable if it falls within a gas-supplied area. For each suitable property, the dataset provides an estimate of annual gas demand, peak demand, recommended hydrogen boiler size, and the associated installation cost (broken down into material, labour, and ancillary costs).
Hydrogen boilers operate in the same way as conventional gas boilers but are designed to run on hydrogen gas. They are considered a potential low-carbon heating option for properties already connected to the gas network, as they require minimal changes to the existing heating system and pipework.
Figure 1: Hydrogen heating suitability assessment. Blue buildings are suitable for hydrogen adoption; grey buildings are unsuitable due to the absence of gas network infrastructure.
Methodology
Gas Supply Areas:
Only properties within gas-supplied areas are assessed. A national off-gas postcode dataset is used to identify which areas have mains gas supply. Properties outside these areas are marked as unsuitable and no demand or cost estimates are provided.
Suitability:
A property is suitable if it falls within a gas-supplied area. For buildings with multiple address records (Shell, Parent, or Child), the suitability rules are:
Child records are always assessed individually.
Parent and Shell records are only assessed where they are the sole address within their building footprint (i.e. they are standalone buildings, not wrappers around multiple units).
Gas Demand:
Annual gas demand is estimated by multiplying the property's floor area by its Energy Use Intensity (EUI), which represents the typical gas consumption per square metre for its building type.
Annual gas demand (kWh) = EUI (kWh/m²) x Floor area (m²)
Peak demand is derived from the annual gas demand, adjusted for the proportion of time the heating system is in use.
Boiler Sizing:
The recommended boiler size is determined from the estimated peak demand. The smallest available boiler size that meets or exceeds the peak demand is selected.
Installation Costs:
Installation costs are estimated using a market cost model that provides material, labour, and ancillary costs for each boiler size. The total cost is the sum of these three components.
Non-Domestic Differences
The non-domestic pipeline follows the same methodology as above, with the following differences:
Gas demand calculation: Rather than a flat EUI lookup, non-domestic buildings use a load factor table that provides a median gas intensity (kWh/m²), reference operating hours, and a load factor for each building classification code. Peak demand is calculated as:
Peak demand (kW) = Annual gas demand / (Reference operating hours x Load factor)
Boiler sizing and cascading: Non-domestic buildings with a peak demand exceeding the largest available single boiler size are served by multiple boiler units installed in cascades. Each cascade unit is 280 kW. The recommended boiler size field reflects the total installed capacity.
Excluded classification codes: The following classification codes are excluded from the non-domestic assessment:
CR05, CR11, PS: marked as unsuitable with no demand or cost estimates. These cover road infrastructure and parking, which are not heated buildings.
P, PP: parent shell records covering mixed-use or unclassified buildings. Suitability is retained but demand and cost fields are not populated.
Assumptions
A property is assumed to be a gas heating candidate if it falls within a gas-supplied postcode area. The dataset does not confirm whether an individual property is currently connected to the gas network.
EUI values are based on typical consumption for each building classification. Actual consumption will vary depending on occupancy, condition, and heating behaviour.
Indoor design temperature is not used in this pipeline. Peak demand is derived purely from annual demand and operating patterns.
Boiler sizing is based on peak demand only. It does not account for existing system constraints, pipework capacity, or hot water demand.
Installation costs are indicative estimates based on an internal market cost model. Actual costs will vary by installer, location, and site-specific conditions.
Sources
Off-gas postcode dataset (Source)
Domestic EUI lookup (Source)
Non-Domestic median gas intensity by building use (Source)
Non-Domestic reference operating hours (Source)
Non-Domestic load factor (Source)
EPC - England and Wales (Source)
EPC - Scotland (Source)
OS NGD Building (Source)
OS NGD Address (Source)
Fields: Domestic & Non-Domestic
Field Name | Field Description |
UPRN | Unique Property Reference Number |
Type | Classification of the building according to building topology (i.e. Shell, Parent, Child) |
Overall Suitability | The overall suitability of the property for a hydrogen boiler |
Estimated Gas Demand (kWh) | The estimated annual gas demand of this property |
Estimated Peak Demand (kW) | The estimated peak gas demand of this property |
Recommended Boiler Size (kW) | The recommended boiler size for this property, considering peak demand |
Estimated Boiler Ancillaries Cost (£) | The estimated ancillaries cost of a hydrogen boiler at the recommended size |
Estimated Boiler Labour Cost (£) | The estimated labour cost of a hydrogen boiler at the recommended size |
Estimated Boiler Material Cost (£) | The estimated material cost of a hydrogen boiler at the recommended size |
Estimated Boiler Total Cost (£) | The estimated total cost of a hydrogen boiler at the recommended size |
Fuel Type | The current main fuel type serving this property |
Gas Demand Percentile Band | Shows how a property's gas demand compares to others in the same local authority, grouped by percentile (10th to 100th) to reflect relative consumption levels. |
Fields: Non-Domestic only
Field Name | Field Description |
Building Classification | Plain-language description of the building type based on its OS AddressBase classification code. |