Energy consumption in new domestic buildings 2015 2017

Pdf File 515.82 KByte, 16 Pages

Energy consumption in new domestic buildings 2015 ? 2017 (England and Wales)

Energy consumption in new domestic buildings 2015 ? 2017 (England and Wales)

19th December 2019

Experimental Statistics

This document gives an overview of energy consumption per square meter of new builds in England and Wales. Energy consumption is counted as the first year of metered gas and electricity consumption. Consumption of other fuels isn't included due to a lack of data. The appendix at the end of this document sets out the method used. Summary statistics are available here.

In this document "consumption" is shorthand for "annual energy consumption per square meter of floor area". All mention of "properties" refer to new builds only. Consumption data tends to lag build year by one year, so consumption in 2015 ? 2017 is used for new builds with EPCs in 2014 ? 2016.

Key points

? The first full year of metered energy consumption per square meter of floor area can be used as a proxy for the efficiency of new buildings.

? Data from Energy Performance Certificates are linked to the National Energy Efficiency Data-Framework (NEED) to find metered annual gas and electricity consumption for new build properties. Properties which couldn't be linked or had communal heating were excluded. Gas and electricity are summed to get annual energy consumption.

? The average consumption of properties built between 2014 and 2016 fell by 4.3% (6 kWh/m2) (weighted by property type, floor area band and whether the property is using gas).

? Average consumption is different for each property type. For properties built in 2017, flats use the least (120 kWh/m2), and bungalows used the most (143 kWh/m2). Energy consumption per m2 tends to decrease with property size.

? On average, properties not using gas used less energy overall than properties using gas (88 kW/m2 and 125 kWh/m2 respectively in 2017). This in part is due to properties not using gas being more likely to use other fuels (e.g.: solid-fuel burners) which are not captured in the data. At end use gas is less efficient for heating than electricity, as the efficiency losses for electric heating come at the point of generation rather than in the home.

? Properties with EPC ratings A and D had the lowest consumption in 2017 (lower than B and C). This is likely to be because a much higher proportion of D rated properties don't use gas (94% in 2017, compared to 5% for B rated properties).


Energy consumption in new domestic buildings 2015 ? 2017 (England and Wales)


The Energy Performance Certificate (EPC) data published online gives energy efficiency information for over half the domestic properties in England and Wales1. This includes all properties built since 2008. The EPC energy efficiency rating is based on how much energy a dwelling will consume, when delivering a defined level of comfort and service provision, using standardised assumptions for occupancy and behaviour. This enables a like-for-like comparison of dwelling performance.

However, the EPC rating doesn't account for individual resident's behaviour (e.g.: whether the resident spends time away from home, is more or less strict with their energy use, has a preference for a higher temperature, etc). This document presents new figures for understanding the efficiency of new builds, which account for such effects.

Domestic energy consumption is affected by many factors, including property type, length of residency and household income2. Metered energy consumption captures the impact of all of these through a figure based on actual consumption rather than modelled values.

The metered gas and electricity consumption held in the National Energy Efficiency Data-Framework (NEED) is linked with EPC data. This is used to find the gas and electricity consumption per square meter. The inspection date of the EPC is used to determine the date the property was built, with the first year of electricity consumption being looked for after this "build year". Table 1 below shows the data sources used.

Table 1: Data sources

Data Metered energy consumption Build year EPC rating Property type Identify converted properties Floor area band

Source NEED EPC EPC EPC Valuation Office Agency (VOA) EPC

Trends in annual energy consumption fluctuate greatly at the household level. The large sample used sees these effects average out, giving a reliable metric.

1 Calculated comparing the EPC dataset released in July 2019, which has certificates for 14.8 million domestic properties, against the 25.9 million properties in England and Wales as published by the Valuation Office Agency in September 2019. 2 An in-depth analysis of the determinants of household gas use is published here.


Energy consumption in new domestic buildings 2015 ? 2017 (England and Wales)

The gas and electricity consumption data are adjusted4 to remove the impact of weather from on year-on-year trends.

The method used is described in detail in Appendix 1 of this document.

Limitations of the data

The energy consumption data only includes metered gas and electricity. Energy used from other sources isn't accounted for, including solid fuel burning, gas from cannisters and electricity generated from solar panels on site.

If it is assumed that the uses of energy remain the same for properties over time (e.g.: heating, cooking, lighting, etc) the changes between years can be viewed as a metric for changes in building efficiency. If energy uses changed (e.g.: if electric cars become more prevalent and are charged at home) then this assumption would no longer be valid under the current methodology.

Some new builds are excluded from the analysis where the EPC data can't be linked to NEED or electricity consumption isn't recorded. Data is linked more successfully for some property types than others, in large part because of how the addresses are formatted. Flats have the lowest match rate. A weighting process is used to reduce the bias this introduces, with the property type accounted for. While the weighting reduces the bias, properties in Inner London are underrepresented in this analysis (see Appendix 1 for further detail).

Shared meters A total of 168,000 new builds in January 2014 ? July 2019 are heated by gas logged with a shared meter (also known as communal heating). This means that a single meter logs the gas consumption of multiple properties. Gas consumption from these shared meters can't be reliably linked to the properties served by the meters. For this reason, new builds with shared meters are excluded from this analysis. A weighting process is used to reduce the bias this introduces to the sample.


Once the EPCs are linked to NEED and the filters applied, the numbers of properties are 90,000 (2015), 107,000 (2016) and 113,000 (2017) (this is the first year of electricity consumption). The weighted and unweighted headline results are shown in Table 2 below. For year on year comparisons the weighted results should be used.

4 Gas consumption is weather corrected by Xoserve, who provide BEIS with gas consumption data. Electricity consumption is weather corrected using correction factors calculated from BEIS' UK Supply and use of fuels (ET 1.3) table.


Energy consumption in new domestic buildings 2015 ? 2017 (England and Wales)

Table 2: Weighted and unweighted results

Unweighted Unweighted Unweighted Weighted Weighted Weighted

First year of Average energy consumption per


square meter (kWh/m2)













Number of properties

89,547 107,278 113,209

89,547 107,278 113,209

Weighting: making sure results represent all new builds Not all newly built domestic properties are included in the analysis: properties not matched to a meter or sharing a gas meter are not used. To reduce bias that may be introduced by this, and so that the results can be compared year on year, weights are calculated for various factors: (i) property type (ii) floor area band (other than bungalows) (iii) whether gas is being used (iv) year

Weights are calculated using all new builds in 2015 (with EPC data) compared to the number of properties after processing in the year in question. Using 2015 as a baseline makes year on year comparisons possible, with changes due to the make-up of the population in a given year removed.

While all other figures in this report are unweighted, the weighted statistics in Table 2 can be used for overall year on year trends.


Energy consumption in new domestic buildings 2015 ? 2017 (England and Wales)

Consumption by property type Figure 1 below shows trends in average energy consumption by property type. Bungalows consistently have the highest consumption, and flats the lowest. Figure 1: Mean energy consumption, by property type and year

The change in energy consumption between 2015 and 2017 varied from a 3% decrease (bungalows) to a 5% decrease (houses). Figure 2 below shows trends in average gas consumption for properties using gas. As with energy consumption, bungalows consistently have the highest gas consumption. Figure 2: Mean gas consumption by property type

The change in gas consumption between 2015 and 2017 varied from a 1% increase (flats) to a 4% decrease (houses). Figures 3 and 4 below present the energy and electricity consumption of properties built in 2017, both using and not using gas, broken down by property type. In all four charts houses have the lowest consumption (this is in part due to houses being larger and the residents being more likely to use alternative fuels, such as gas from cannisters). 5

Energy consumption in new domestic buildings 2015 ? 2017 (England and Wales) Figure 3: Mean energy consumption in 2017, by gas use status and property type

Figure 4: Mean electricity consumption in 2017, by gas use status and property type

How does consumption vary by property size? Figures 5 and 6 below show that larger properties tend to consume less energy per m2. Bungalows tend to use the most for a given floor area band and flats use the least. The longer trend shown in figure 7 shows that consumption per m2 decreases as total floor area increases.


Download Pdf File