How Much of Your Scope 2 Reduction Is the Grid, Not You?

A Scope 2 reduction can appear in a report without one kilowatt hour being saved. This article answers one question. If a building uses the same electricity every year how far does its location based Scope 2 fall from conversion factor updates alone? The answer draws on the four UK government factor sets from 2023 to 2026 and the reports of four UK listed landlords and six peer reviewed papers. The finding is that the same consumption reports 36.8% less on the 2026 set than on the 2023 set. Where a portfolio also cuts its electricity use by 10% about four fifths of the fall it reports still comes from the grid.

Summary of findings

The factor fell by more than a third. The UK electricity emission factor for generation was 0.20707 kgCO2e per kWh in the 2023 set and 0.13096 in the 2026 set [1]. One million kWh reports 207.07 tCO2e on the first and 130.96 tCO2e on the second. That is a fall of 36.8% with no change in consumption. Nearly all of it came in two steps. The 2025 set was 14.5% below the 2024 set and the 2026 set was a further 26.0% lower.

About a quarter of the last step is method and not grid. The Department for Energy Security and Net Zero (DESNZ) cut the data lag behind the factor from two years to one in the 2026 set. It attributes 16 percentage points of the 26% fall to changes in the grid mix between 2023 and 2024 and 3 points to changes between 2024 and 2025. It puts the remaining 6 to 7 points down to methodological changes and improvements and corrections [2].

The effect is already visible in landlord reports. Derwent London used 1% more electricity in landlord controlled areas in 2025 and reported 13% less location based Scope 2 [3]. Regional REIT used 14% less like for like electricity and reported 26% less like for like Scope 2. It notes that the carbon content of UK electricity fell 15% [4].

A reported fall can be split into two parts. In a worked example a portfolio cuts electricity use by 10% between the two sets and reports a fall of 43.1%. Pricing both years at the 2023 factor shows a fall of 10.0%. About 77% of the reported Scope 2 reduction is the grid.

Could you show how much of the Scope 2 reduction you reported last year came from your own buildings? See how Sustainify AI keeps every figure beside the factor set that priced it.

Why the source of a Scope 2 reduction matters

Location based Scope 2 is the figure most frameworks ask a UK landlord for. The templates in the government reporting guidelines mark location based Scope 2 as mandatory for Streamlined Energy and Carbon Reporting (SECR) and market based Scope 2 as optional [5]. UK Sustainability Reporting Standard S2 (UK SRS S2) requires a company to disclose its location based Scope 2 (paragraph 29(a)(v)) [6]. The Financial Conduct Authority (FCA) final rules in PS26/19 apply the standard to listed companies in scope on a comply or explain basis for accounting periods beginning on or after 1 January 2027 [7].

GRESB (the real estate sustainability benchmark) makes the location based figure mandatory wherever Scope 2 is reported. A market based figure is optional and is not factored into the final score. Its indicator GH1 gives 5 points for data coverage and 2 points for like for like performance improvement. On the like for like part assets with increased or unchanged emissions compared to the previous year automatically receive a score of 0 [8].

In every one of those places the grid arrives inside the number. The GHG Protocol notes that the grid average emission factor is largely outside of corporate control [9]. A sustainability lead presenting a 26% Scope 2 reduction needs to know how much of it the buildings earned. A finance director signing the report needs to know what happens to the trend if the factor rises. It rose 7% in the 2023 set [10].

Targets carry the same effect. The Science Based Targets initiative (SBTi) requires a single and consistent Scope 2 approach for calculating base year emissions and tracking performance against a target [11]. A target tracked on the location based method therefore counts a falling grid factor as progress.

Definitions

Scope 2 covers indirect emissions from purchased electricity and heat and steam and cooling. A location based figure applies the average emissions intensity of the grid the building sits on. A market based figure applies the emissions intensity of the electricity contracted for. A REGO (Renewable Energy Guarantee of Origin) is a certificate evidencing that a megawatt hour of electricity came from a renewable source. All four terms are defined in the Sustainify AI glossary.

A conversion factor turns a unit of activity such as a kilowatt hour (kWh) into a quantity of greenhouse gas. A factor set is a complete dated collection of conversion factors published together. This article uses the UK Government GHG Conversion Factors for Company Reporting and calls each annual edition a set. The generation factor is the factor for electricity generated and it is reported in Scope 2. Transmission and distribution losses are the electricity lost in the grid between generation and the meter. They carry a separate factor and are reported in Scope 3 [1].

A baseline is the reference year a reduction is measured against. It is recalculated when the organisational boundary changes materially or when a change in calculation method has a significant impact. A restatement is a formal correction of a previously published figure disclosed with the reason given. Factor versioning means holding every published factor set separately and recording which one produced each figure. Period locking closes a reporting period so that its figures cannot change without a recorded and authorised reopening.

Carbon intensity is emissions per unit of something else such as floor area or a kilowatt hour. Landlord controlled consumption is energy the landlord procures and operates such as common parts and central plant. tCO2e is tonnes of carbon dioxide equivalent.

Method

The generation factor and the transmission and distribution factor were taken from Table 9 of the 2026 DESNZ methodology paper which lists the factors by grid data year [1]. Each data year was matched to the factor set that used it. Publication dates came from the GOV.UK page for each set [12] [13] [14] [15]. The reasons for each movement came from the DESNZ major changes documents [10] [16] [2].

Each factor was applied to a fixed 1 million kWh. The worked example in Finding 3 uses an invented portfolio and is labelled as such. It splits the reported Scope 2 reduction in two steps. The consumption effect is the change in kWh priced at the baseline factor. The grid effect is what remains. A log mean Divisia index (LMDI) decomposition was run as a check [17].

Company figures were read from the latest reports of four UK listed landlords. Peer reviewed papers were found through Google Scholar. Each was checked at its publisher page or its Digital Object Identifier (DOI) record or an open repository copy. All sources were checked on 7 October 2026. Factor values were read from the methodology paper and not from the factor spreadsheets.

Findings

Finding 1. The same electricity reports 36.8% less on the 2026 set than on the 2023 set

Table 1 applies each generation factor to 1 million kWh. Nothing about the building changes between the rows.

Table 1. UK electricity generation factor by set and the Scope 2 it gives for 1 million kWh

Source. DESNZ 2026 methodology paper Table 9 [1]. Tonnes and percentages calculated by this article.

The 2024 set barely moved. The 2025 set was 14.5% below it and the 2026 set was a further 26.0% lower. Across the three updates the fall is 36.8%. A building would have needed to cut its electricity use by 36.8% to earn the same fall on a fixed factor. Read the other way a building could use 58% more electricity on the 2026 set than on the 2023 set and report the same Scope 2 (0.20707 ÷ 0.13096 = 1.58).

Transmission and distribution losses move the same way. The factor for electricity consumed adds those losses to the generation factor. It was 0.22500 in the 2023 set and 0.14396 in the 2026 set which is a fall of 36.0% [1].

Finding 2. The factor describes an earlier grid and the 2026 set covers two years of change

Each set is built from grid data that is already old when it is published. Up to the 2025 set the lag was two years so the 2025 set described the grid of 2023. For the 2026 set DESNZ reduced the lag to one year. The 2026 set uses 2025 data and the series has no 2024 data year [1].

Table 2. Publication date and grid data year and stated reason for each set

Source. GOV.UK publication pages [12] [13] [14] [15]. Data years from Table 9 [1]. Percentages as stated by DESNZ.

DESNZ splits the 26% fall in the 2026 set into three parts.

Table 3. What makes up the 26% fall in the 2026 generation factor

Source. DESNZ 2026 major changes report [2].

About three quarters of the 2026 step is the grid and about one quarter is method. DESNZ names the method changes. Power sector emissions are now built from fuel use data for the year before publication. The calculation also improves how imports and exports of electricity are counted and corrects a small double counting issue that affected autogenerators [1] [2].

The grid part is real. The system operator reports that the last coal power station in Great Britain came off the system on 30 September 2024 and that 2024 was the lowest carbon intensity year at an average of 125 gCO2 per kWh [18]. That is a Great Britain figure and the DESNZ factor is a UK one.

The series does not only fall. The generation factor was 0.19338 in the 2022 set and 0.20707 in the 2023 set [1]. DESNZ put that 7% rise down to more natural gas use and less renewables use [10]. A landlord with flat consumption reported a rise that year.

DESNZ leaves one decision with the reporter. It says users will need to decide whether to adjust their estimates for previous years to use the new factors so that they have a consistent time series. If they do not the 2024 grid data is skipped in their series [1]. The article on UK GHG conversion factors 2026 covers the other changes in that set.

Finding 3. A 10% cut in consumption reports as a 43% fall and about four fifths of it is the grid

Table 4 takes an invented portfolio. It used 1 million kWh in a year priced on the 2023 set and 900000 kWh in a year priced on the 2026 set.

Table 4. Worked split of one reported Scope 2 reduction

Source. Worked example by this article using the factors in Table 1. The portfolio is invented and is not an observed estate.

The consumption effect is the part the buildings earned. The grid effect is the rest. The split depends a little on the method. An LMDI decomposition of the same example gives the grid 81% [17]. Pricing the grid effect first at baseline consumption gives it 85%. Each route gives the grid about four fifths.

The grid share falls as real savings grow. With a 20% cut in consumption the reported fall is 49.4% and the grid share on the two step method is 59.5%. With no cut at all the reported fall is 36.8% and the grid share is 100%.

Finding 4. Landlord reports already show the effect

The 2025 set was the first large step and it priced reporting years that fell mostly in 2025. Table 5 sets the change in electricity beside the change in location based Scope 2 in three reports. Regional REIT and Shaftesbury Capital publish theirs as EPRA sBPR measures.

Table 5. Electricity use and location based Scope 2 in three landlord reports from 2024 to 2025

Source. Company reports as cited. Percentages calculated by this article from the printed figures.

Derwent London is the clearest case. Electricity use in landlord controlled areas rose and location based Scope 2 fell by 365 tCO2e. Dividing Scope 2 by electricity gives 0.206 kgCO2e per kWh for 2024 and 0.176 for 2025. Those sit close to the generation factors in the 2024 and 2025 sets.

At Regional REIT the same division returns 0.20705 and 0.17699 which match the two sets. Pricing its 2025 like for like electricity at the 2024 factor gives 5624 tCO2e. On that basis about 52% of its reported Scope 2 reduction of 1699 tCO2e came from lower consumption and about 48% from the factor.

British Land reports that operational carbon intensity across its managed portfolio has reduced 47% since FY19 as a result of its energy efficiency programme and grid decarbonisation [20]. That sentence names both causes and gives one figure for the two together.

The larger step has not yet reached annual reports. The 2026 set is for activity data that falls entirely or mostly within 2026 [1]. A landlord with flat consumption will report a location based Scope 2 fall of 26% for that year.

Finding 5. Peer reviewed studies find the same pattern in other estates and companies

Table 6. What peer reviewed papers found

Source. Abstracts and open text of the papers cited.

The university study is the closest match to a property portfolio. Its decomposition attributes the fall mainly to the emission coefficient effect and the energy intensity effect [21]. The corporate study shows that a reported Scope 2 trend can run well ahead of real mitigation [22]. The grid study shows how fast the British system has moved before [23].

The last two papers are cautions. An annual average factor is itself an approximation [24]. The market based figure has its own critics which is why this article treats the location based figure as the physical series [25].

A fall earned by the grid and a fall earned by the building look the same in one total. Sustainify AI pins each period to its factor set so every Scope 2 reduction you report can be traced to the factor behind it.

What this means in practice

The base year rule keeps the grid effect in the trend

The GHG Protocol requires base year emissions to be recalculated for changes in calculation methodology or improvements in the accuracy of emission factors that have a significant impact. It also states that changes in emission factor that reflect real changes in emissions do not trigger a recalculation [26]. Grid decarbonisation is a real change. The baseline therefore keeps the factor of its own year and the grid effect stays in the reported trend. That is correct accounting. It is also why the split in Finding 3 has to be shown separately.

The 2026 set needs a decision because 6 to 7 points of its fall are method [2]. The government reporting guidelines list changes in calculation methods or improvements in the accuracy of emission factors among the cases where a base year might need to be recalculated if the impact is significant [5].

Five actions for reporting a Scope 2 reduction

1. Report the electricity change beside the emissions change

SECR already requires a figure in kWh for the annual quantity of energy consumed and the disclosures made in the previous year [5]. Put the percentage change in kWh on the same line as the percentage change in tCO2e. A reader then sees at once how much of the fall is consumption.

2. Add a line that holds the factor at the baseline

Price the current year electricity at the baseline factor as in Table 4. Publish the result as an analysis line beside the reported figure. It is not a baseline restatement and the reported figures do not change.

3. Record a decision on the 2026 method change

DESNZ says users need to decide whether to adjust earlier years [1]. Apply the base year policy and record the decision and the reason for it. UK SRS S2 asks a company to disclose any changes it made to its measurement approach and inputs and assumptions during the reporting period (paragraph 29(a)(iii)) [6].

4. Pin each period to its factor set

Keep every factor set separately and record which set priced each figure. That is factor versioning. Close each finished period with period locking. The record then forms part of the calculation lineage of each figure. The comparison in action 2 can then be reproduced by someone who was not there when it was made.

5. Test the target with consumption held flat

State which Scope 2 method the target uses [11]. Then work out what the target year figure would be if kWh did not change. UK SRS S2 asks for an analysis of trends or changes in performance against each climate related target (paragraph 35) [6]. The article on setting a science based net zero target covers why a target has to survive acquisitions and disposals.

How Sustainify AI supports Scope 2 reporting

Factor sets pinned to the period. Under carbon data governance every factor carries a version reference and an effective date and a source. When the UK factors update closed periods stay locked.

Restatements on the record. A figure in a closed period changes only through a deliberate recorded restatement. A decision on the 2026 method change therefore leaves a trace.

Every figure traces to its factor. Carbon data lineage shows each factor applied to an entry with its name and date and publisher.

Variance surfaced for review. Living AI gives a written read on the current period including which scope moved. It answers questions from the governed data and surfaces a variance for a person to act on.

A governed baseline for the target. The net zero pathway tools model reduction pathways against a governed baseline and report variance by asset.

The platform supports a disclosure and does not replace an assurance provider.

Disclosure of interest. This article is published by Sustainify AI. The analysis used the public sources and the method set out in the Method section and was carried out independently of the platform.

Limitations

The factors were read from the methodology paper. Table 9 of the 2026 paper lists the factor for each data year. The factor spreadsheets were not opened. The 2023 methodology paper printed 0.20730 for the 2021 data year [27] and the 2026 paper prints 0.20707 [1]. This article uses the later figure. The difference is 0.1%. The carbon dioxide part is 0.20496 in both papers. The gap sits in methane and nitrous oxide which the 2026 paper aligns with AR5 global warming potential values.

The split is a method choice. The two step method gives the grid 77% of the worked example. An LMDI decomposition gives 81% and the reverse order gives 85%. The share also changes with the size of the real saving.

The worked example is invented. It shows the size of the effect for one assumed saving and not a typical result.

The company comparison is small. Three reports are not a sample of UK landlords. The Shaftesbury Capital figures are absolute and include the effect of asset sales. The factors implied by dividing Scope 2 by electricity are calculations by this article. The Scope 2 line may cover more than the electricity line beside it. British Land reports an intensity for its own financial year and not a Scope 2 total.

A national annual average hides time and place. The UK factor is one average for the national grid in each data year and it includes net imports of electricity through interconnectors [1]. It does not show when or where a building used its electricity.

The studies are not about UK commercial property. They cover universities and a global set of companies and facilities in the United States. One is a perspective paper and not an empirical study. They show the pattern and not its size in any portfolio.

Market based figures are outside the scope. Derwent London bought 100% of its electricity on REGO backed tariffs and reports market based Scope 2 of 4 tCO2e for 2025 [3]. A figure like that barely moves with the grid factor.

What not to conclude. A Scope 2 reduction driven by the grid is not a false number. The emissions did fall. The point is to say who earned the fall.

Ready to report a Scope 2 reduction that an assurance provider can trace back to its factor set? Talk to Sustainify AI about a governed baseline for your portfolio.

Frequently asked questions

How much did the UK electricity factor fall between the 2023 and 2026 sets?

The generation factor fell from 0.20707 to 0.13096 kgCO2e per kWh which is 36.8% [1]. The factor for electricity consumed includes transmission and distribution losses and fell 36.0%.

Why did my Scope 2 emissions fall when my electricity use did not?

Location based Scope 2 is electricity use multiplied by the grid factor for the year. When the factor falls the result falls even if the kWh figure is unchanged. Derwent London used 1% more landlord electricity in 2025 and reported 13% less location based Scope 2 [3].

Which factor set applies to my reporting year?

DESNZ says the 2026 set is for use with activity data that falls entirely or mostly within 2026 [1]. The 2023 paper used the same wording for its own year [27]. A reporting year from April 2026 to March 2027 has nine months in 2026 so it takes the 2026 set.

Why did the 2026 electricity factor fall by 26%?

DESNZ reduced the data lag from two years to one so the set reflects two years of grid change. It attributes 16 points to the grid mix between 2023 and 2024 and 3 points to 2024 to 2025 and 6 to 7 points to method [2].

Do earlier years need restating when a new factor set is published?

Not for a change in the grid. The GHG Protocol states that changes in emission factor that reflect real changes in emissions do not trigger a base year recalculation [26]. A grid driven fall therefore needs no baseline restatement. For the 2026 method change DESNZ says users need to decide whether to adjust earlier years [1]. Carbon assurance readiness depends on recording that decision.

How do I show how much of a Scope 2 reduction is my own?

Price the current year electricity at the baseline factor. The gap between that figure and the baseline is the consumption effect. The rest of the reported fall is the grid effect. In the worked example the grid accounts for about 77% of a 43.1% fall.

Can the electricity factor go up?

Yes. The UK electricity emission factor in the 2023 set was 7% higher than in the 2022 set because of more natural gas use and less renewables use [10]. A building with flat consumption reported a rise that year.

Does the grid effect change a market based Scope 2 figure?

Much less. A market based figure uses the emissions intensity of the electricity contracted for. Derwent London reports market based Scope 2 of 4 tCO2e against 2340 tCO2e location based [3]. The location based figure is still required [6].

Does SECR ask for electricity use as well as emissions?

Yes. The report must disclose a figure in kWh for the annual quantity of energy consumed and the disclosures made in the previous year [5]. Showing the kWh change beside the tCO2e change is the simplest way to separate the two effects in SECR reporting.

Does GRESB score a fall in location based emissions?

Yes. GRESB makes the location based figure mandatory where Scope 2 is reported and does not factor a market based figure into the final score. It gives 2 of the 7 points in GH1 for like for like performance improvement [8]. A like for like fall produced by the factor is still a fall in the reported figure. Prepare the split before a GRESB submission so the commentary can explain it.

References

[1] Department for Energy Security and Net Zero. 2026 Government Greenhouse Gas Conversion Factors for Company Reporting Methodology Paper. June 2026. Table 9 and its notes and paragraphs 1.10 and 3.5 to 3.12. Accessed 7 October 2026.

[2] Department for Energy Security and Net Zero. 2026 Government Greenhouse Gas Conversion Factors for Company Reporting Major Changes to the Conversion Factors. June 2026. Reference 17 on page 6. Accessed 7 October 2026.

[3] Derwent London plc. Responsibility Report 2025 carbon footprint pages. Pages 10 and 11. Accessed 7 October 2026.

[4] Regional REIT Limited. EPRA Sustainability Report 2025. Year ended 31 December 2025. Pages 4 and 5. Accessed 7 October 2026.

[5] HM Government. Environmental Reporting Guidelines Including streamlined energy and carbon reporting guidance. March 2019. Pages 17 to 18 and 34 and 39 to 40 and 51. Accessed 7 October 2026.

[6] Department for Business and Trade. UK Sustainability Reporting Standard S2 Climate related Disclosures. 25 February 2026. Paragraphs 29(a)(iii) and 29(a)(v) and 35. Accessed 7 October 2026.

[7] Financial Conduct Authority. PS26/19 Aligning listed issuers sustainability disclosures with international standards. 30 September 2026. Paragraphs 1.2 and 3.12. Accessed 7 October 2026.

[8] GRESB. 2026 Real Estate Assessment GH1 Scope 1 to 3 Emissions. GRESB Guides. 2026. Accessed 7 October 2026.

[9] GHG Protocol. Top Ten Questions about the Scope 2 Guidance. 17 January 2019. Question 9. Accessed 7 October 2026.

[10] Department for Energy Security and Net Zero. 2023 Government Greenhouse Gas Conversion Factors for Company Reporting Major Changes to the Conversion Factors. June 2023. Reference 17 on page 5. Accessed 7 October 2026.

[11] Science Based Targets initiative. SBTi Corporate Near Term Criteria Version 5.3.1. April 2026. Criterion C8 on page 10. Accessed 7 October 2026.

[12] Department for Energy Security and Net Zero. Greenhouse gas reporting conversion factors 2023. Published 7 June 2023. Last updated 28 June 2023. Accessed 7 October 2026.

[13] Department for Energy Security and Net Zero. Greenhouse gas reporting conversion factors 2024. Published 8 July 2024. Last updated 30 October 2024. Accessed 7 October 2026.

[14] Department for Energy Security and Net Zero. Greenhouse gas reporting conversion factors 2025. Published 10 June 2025. Accessed 7 October 2026.

[15] Department for Energy Security and Net Zero. Greenhouse gas reporting conversion factors 2026. Published 11 June 2026. Last updated 31 July 2026. Accessed 7 October 2026.

[16] Department for Energy Security and Net Zero. 2025 Government Greenhouse Gas Conversion Factors for Company Reporting Major Changes to the Conversion Factors. June 2025. Reference 7 on page 4. Accessed 7 October 2026.

[17] Ang BW. LMDI decomposition approach A guide for implementation. Energy Policy. Volume 86 pages 233 to 238. 2015. Accessed 7 October 2026.

[18] National Energy System Operator. Britains Electricity Explained 2024 Review. 7 January 2025. Accessed 7 October 2026.

[19] Shaftesbury Capital PLC. 2025 EPRA Sustainability Data Report. Pages 12 and 15 and 16 and 32. Accessed 7 October 2026.

[20] The British Land Company PLC. Sustainability Progress Report 2026. Page 12. Accessed 7 October 2026.

[21] Eskander SMSU and Nitschke J. Energy use and CO2 emissions in the UK universities An extended Kaya identity analysis. Journal of Cleaner Production. Volume 309 article 127199. 2021. Accessed 7 October 2026.

[22] Bjørn A et al. Renewable energy certificates threaten the integrity of corporate science based targets. Nature Climate Change. Volume 12 issue 6 pages 539 to 546. 2022. Accessed 7 October 2026.

[23] Staffell I. Measuring the progress and impacts of decarbonising British electricity. Energy Policy. Volume 102 pages 463 to 475. 2017. Accessed 7 October 2026.

[24] Miller GJ et al. Hourly accounting of carbon emissions from electricity consumption. Environmental Research Letters. Volume 17 issue 4 article 044073. 2022. Accessed 7 October 2026.

[25] Brander M et al. Creative accounting A critical perspective on the market based method for reporting purchased electricity (scope 2) emissions. Energy Policy. Volume 112 pages 29 to 33. 2018. Accessed 7 October 2026.

[26] World Resources Institute and World Business Council for Sustainable Development. The Greenhouse Gas Protocol A Corporate Accounting and Reporting Standard. Revised Edition. 2004. Chapter 5 pages 35 to 38. Accessed 7 October 2026.

[27] Department for Energy Security and Net Zero. 2023 Government Greenhouse Gas Conversion Factors for Company Reporting Methodology Paper. June 2023. Section 1 and Table 9 on pages 35 to 36. Accessed 7 October 2026.