Commercial Solar Carbon Savings: kg CO2 per kWh on 2026 Factors
Updated 23 October 2026 · SEO Dons Editorial
How much CO2 does a kWh of self-consumed solar save?
On the 2026 UK electricity factor, about 0.13 kg of CO2e. The Department for Education's SECR guidance for academy trusts applies 0.13096 kg CO2e per kWh, labelled "2026 UK electricity", to purchased electricity in its Scope 2 worked example, using the Government's 2026 conversion factors. Multiply the kWh of solar used on site by that figure to get the saving: 10,000 kWh used on site saves about 1.3 tonnes.
Two points define what that number is. First, it is a generation-only factor. The GHG Protocol says Scope 2 methods use "generation-only emission factors, which do not include T&D losses or upstream lifecycle emissions", and the 2026 conversion factors keep a separate transmission and distribution tab. The saving here therefore excludes network losses and upstream fuel emissions. Second, only units that replace purchased electricity count. Exported units do not reduce a Scope 2 figure; the on-site solar and Scope 2 page explains how exports and certificates are treated.
How many tonnes do 50, 100 and 250 kWp save a year, by region?
The factor is the same across the UK, so regions differ only through how much a roof makes. The table uses PVGIS yields for a flat roof with east-west rows at 10 degrees, and assumes every kWh generated replaces a purchased kWh, which makes it an upper bound.
| Region (city used) | kWh per kWp a year | Tonnes CO2e, 50 kWp | Tonnes CO2e, 100 kWp | Tonnes CO2e, 250 kWp | 100 kWp, south 35 degrees |
|---|---|---|---|---|---|
| Scotland (Glasgow) | 707 | 4.6 | 9.3 | 23.1 | 11.2 |
| North West (Manchester) | 741 | 4.9 | 9.7 | 24.3 | 11.6 |
| Northern Ireland (Belfast) | 756 | 5.0 | 9.9 | 24.8 | 12.0 |
| North East (Newcastle) | 764 | 5.0 | 10.0 | 25.0 | 12.4 |
| Yorkshire and the Humber (Leeds) | 775 | 5.1 | 10.2 | 25.4 | 12.4 |
| East Midlands (Nottingham) | 796 | 5.2 | 10.4 | 26.1 | 12.7 |
| West Midlands (Birmingham) | 800 | 5.2 | 10.5 | 26.2 | 12.7 |
| London and the South East (London) | 840 | 5.5 | 11.0 | 27.5 | 13.3 |
| East of England (Norwich) | 843 | 5.5 | 11.0 | 27.6 | 13.5 |
| Wales (Cardiff) | 865 | 5.7 | 11.3 | 28.3 | 13.6 |
| South West (Plymouth) | 932 | 6.1 | 12.2 | 30.5 | 14.7 |
A 100 kWp array saves about 12.2 tonnes a year at Plymouth against 9.3 at Glasgow, a gap of about 32%. A south-facing 35 degree layout makes roughly 21% more per kWp in Birmingham (the last column).
The share of output used on site then scales the result. For 100 kWp in Birmingham:
| Share of output used on site | kWh used on site | Tonnes CO2e a year |
|---|---|---|
| 100% | 80,018 | 10.5 |
| 75% | 60,014 | 7.9 |
| 50% | 40,009 | 5.2 |
How much a real site uses depends on its load, which the load profile page illustrates with hourly modelling. The savings calculator is the place to run a site's own figures.
Why do older carbon claims overstate savings today?
- The grid has cleaned up. NESO reports carbon intensity of 125 gCO2/kWh for 2024, "a record low and 70% lower than 2014". NESO measures the grid on its own basis, so this shows the direction rather than replacing the conversion factor. Savings quoted from earlier years rest on dirtier electricity.
- Factors are issued yearly and old ones are not revised. The Department for Energy Security and Net Zero publishes an update each year, and as a general policy previously published factors are not revised. The 2026 edition uses a new methodology that cuts the data lag from two years to one and gives a "significantly lower electricity emissions factor compared with 2025". A calculator on an earlier edition overstates by the ratio of the two factors.
- Lifetime totals built on one year's factor are upper-end. If the grid keeps decarbonising, later years save less than the first.
The Department for Education's example advises using the 2026 factors for the 2025 to 2026 financial year because they cover most of it. State the factor year beside any saving.
How long does a module take to repay its manufacturing carbon?
Fraunhofer ISE's 2021 life cycle analysis of six monocrystalline silicon modules found footprints of 810 kg CO2e per kWp for glass-foil modules made in China, 580 for Germany and 480 for the EU, with 750, 520 and 420 for glass-glass modules. It says the electricity mix of the producing country has the greatest influence and that, at 50 to 63 percent, the share of energy required in manufacturing is the most influential factor. These cover production, transport and end of life for the module only, not inverters, mounting or cabling.
Dividing the footprint by the yearly saving per kWp (yield multiplied by 0.13096, all output displacing purchased power) gives a payback in years for Birmingham, glass-foil first and glass-glass in brackets:
| Where the module is made | kg CO2e per kWp | Years to repay, east-west 10 degrees | Years to repay, south 35 degrees |
|---|---|---|---|
| China | 810 (750) | 7.7 (7.2) | 6.4 (5.9) |
| Germany | 580 (520) | 5.5 (5.0) | 4.6 (4.1) |
| European Union | 480 (420) | 4.6 (4.0) | 3.8 (3.3) |
Across the cases on this page the range runs from about 2.9 years (EU glass-glass, Plymouth, south-facing) to about 8.8 (China glass-foil, Glasgow, east-west). Energy payback is shorter: Fraunhofer ISE's Photovoltaics Report, updated 14 July 2026, says silicon use per watt has fallen enough to shorten energy payback "to approximately 1 year, depending on the location". The two measures answer different questions, and the carbon figure moves with the grid factor, so a cleaner grid stretches carbon payback. The IEA PVPS fact sheet adds that emissions are primarily from manufacturing and that impacts have improved markedly compared with 2015 values. What happens to modules afterwards is covered in the WEEE and recycling guide.
Which assumptions sit behind the figures?
- Factor: 0.13096 kg CO2e per kWh, 2026 UK electricity, generation only.
- Yields: PVGIS 5.3 PVcalc, 1 kWp, 14% system losses, PVGIS-SARAH3 radiation data for 2005 to 2023, city-centre coordinates, first-year output with no degradation. East-west is the mean of east-facing and west-facing 10 degree planes.
- Displacement: all output replaces purchased electricity unless the table says otherwise.
- Module footprints: Fraunhofer ISE's 2021 study, for modules and not whole systems. Ask for the declared footprint of the module actually specified.
All arithmetic is illustrative. You can send an enquiry through the site's form if you want a designer to produce a site-specific estimate. The wider specification is in the commercial PV systems guide.
Sources
- Streamlined Energy and Carbon Reporting (SECR) for academy trusts, Department for Education, gov.uk, updated 12 August 2026, checked 22 October 2026.
- Greenhouse gas reporting: conversion factors 2026, Department for Energy Security and Net Zero, gov.uk, published 11 June 2026, checked 22 October 2026; summary of the 2026 changes, Energy Managers Association, EMA, checked 22 October 2026.
- Scope 2 FAQ on T&D losses, GHG Protocol technical assistance, FAQ article, checked 22 October 2026.
- Road to Zero Carbon in numbers, NESO, NESO, checked 22 October 2026.
- European Glass-Glass Photovoltaic Modules Are Particularly Climate-Friendly, Fraunhofer ISE, 23 September 2021, press release, checked 22 October 2026; Photovoltaics Report, Fraunhofer ISE, report page, updated 14 July 2026, checked 22 October 2026.
- Environmental Life Cycle Assessment of Electricity from PV Systems, IEA PVPS Task 12, May 2024, fact sheet, checked 22 October 2026.
- PVGIS user manual, European Commission Joint Research Centre, user manual, checked 22 October 2026. PVGIS 5.3 PVcalc runs, 1 kWp, 14% losses, European Commission Joint Research Centre, checked 22 October 2026: Glasgow s35 e10 w10; Manchester s35 e10 w10; Belfast s35 e10 w10; Newcastle s35 e10 w10; Leeds s35 e10 w10; Nottingham s35 e10 w10; Birmingham s35 e10 w10; London s35 e10 w10; Norwich s35 e10 w10; Cardiff s35 e10 w10; Plymouth s35 e10 w10 (s35 south 35 degrees, e10 east 10 degrees, w10 west 10 degrees).
Common questions
Does the 0.13096 factor include transmission and distribution losses?
No. The GHG Protocol says Scope 2 methods use generation-only factors that exclude T&D losses and upstream emissions. Network losses are a separate Scope 3 calculation, and the conversion factors list a separate transmission and distribution tab.
Do exported kWh count towards the saving?
Not in a Scope 2 figure, which is built from purchased electricity. Any argument about avoided emissions from exports is a separate kind of claim and should be labelled as one.
Why is my earlier calculation higher than these tonnes?
It probably used an earlier edition of the conversion factors. The 2026 edition gives a significantly lower electricity factor than 2025, so the same kWh saves less carbon than it did.
Is the saving the same in every region?
The factor is UK-wide, but the kWh each kWp makes is not. Flat-roof east-west yields in these runs range from about 707 kWh per kWp in Glasgow to about 932 in Plymouth.