Commercial solar by UK region: how location changes payback
Updated 23 October 2026 · SEO Dons Editorial
What does location change in a payback calculation?
This page does not give a payback in years or pounds for each region, because no dated official source pairs installed cost with bill saving by region. What it can show is the one part of payback that location changes directly, which is how many kilowatt-hours each kilowatt-peak generates. Simple payback is net cost divided by annual saving. If cost, tariffs and the share of generation used on site are identical, the saving moves in step with yield, so payback varies inversely with yield. A site that generates 20% less takes 25% longer to pay back. That is arithmetic, not a forecast.
How much does yield differ across the UK?
PVGIS gives about 757 to 987 kWh per kWp a year across the 13 locations below, with Exeter highest and Inverness lowest. Exeter is about 30% above Inverness. The last column divides Exeter's yield by each location's yield. It is the factor by which payback would stretch if only yield changed.
| Location | Region | kWh per kWp per year | Year-to-year variation (kWh per kWp) | Payback multiplier vs Exeter |
|---|---|---|---|---|
| Exeter | South West | 987 | 32 | 1.00 |
| Cardiff | Wales | 976 | 33 | 1.01 |
| Norwich | East of England | 958 | 38 | 1.03 |
| Reading | South East | 955 | 35 | 1.03 |
| London | Greater London | 951 | 37 | 1.04 |
| Birmingham | West Midlands | 904 | 30 | 1.09 |
| Nottingham | East Midlands | 902 | 27 | 1.09 |
| Leeds | Yorkshire and the Humber | 880 | 28 | 1.12 |
| Newcastle | North East | 874 | 28 | 1.13 |
| Belfast | Northern Ireland | 856 | 24 | 1.15 |
| Manchester | North West | 832 | 29 | 1.19 |
| Glasgow | Scotland | 796 | 24 | 1.24 |
| Inverness | Scottish Highlands | 757 | 33 | 1.30 |
Settings for every row: PVGIS 5.3 PVcalc, PVGIS-SARAH3 radiation data for 2005 to 2023 with the calculated horizon, 1 kWp, 14% system loss, fixed mounting facing south at 15 degrees, city-centre coordinates, retrieved in October 2026. The 15 degree tilt sits between the 10 and 35 degree layouts on the yield by region page. The variation column is PVGIS's standard deviation of annual output, 24 to 38 kWh per kWp, which is only about 3% to 4% of the annual figure. Weather-year noise is small next to the gap between regions. Capacity by region is a different question, covered on the capacity by region page.
As an illustration only, take a proposal that shows a payback of 6.0 years at an Exeter-level yield. With identical cost, tariffs and self-consumption, the same system at Glasgow's yield would show 7.4 years, and at Inverness's 7.8 years (6.0 multiplied by the column value). Those two figures are hypothetical, not regional paybacks.
Why can northern sites still pay?
Because yield is one input among several, and others can move payback by as much. A major one is the share of generation the building uses itself. Value per generated kWh is the on-site share times the avoided import price plus the export share times the export rate. Ofgem says SEG licensees determine the rate, contract length and other terms for exported power, and that the rate must always be above zero. So the export rate varies by licensee, not by latitude.
The table below uses a hypothetical ratio of export value to avoided import price, shown as 0.2 and 0.4 (no prices are assumed). It shows the factor by which payback changes against a 60% on-site share, all else equal.
| Share of generation used on site | Payback factor if export is worth 0.2 of import | Payback factor if export is worth 0.4 of import |
|---|---|---|
| 40% | 1.31 | 1.19 |
| 60% | 1.00 | 1.00 |
| 80% | 0.81 | 0.86 |
| 90% | 0.74 | 0.81 |
On the 0.2 assumption, lifting the on-site share from 60% to 80% shortens payback by a factor of 0.81, which is about the reverse of the 1.24 factor separating Glasgow from Exeter. On the 0.4 assumption the effect is smaller (0.86). A lower yield per kWp also means a given building uses a larger share of what each kWp makes, which narrows the gap. On those assumptions a building with a steady daytime load can gain more from a higher on-site share than it loses to latitude, and the load profiles page shows how to read a load profile.
Which inputs matter more than latitude?
| Input | Why it matters for payback | Where to read more |
|---|---|---|
| On-site share of generation | Sets the value per kWh, as the table above shows | Load profiles |
| Export arrangement | Licensees set their own rates and terms, and the scheme covers installations in Great Britain only, so Belfast export terms differ | SEG rates for businesses |
| Layout and shading | Tilt, orientation and obstructions change yield at the same site | Shading |
| Capital cost | The denominator of payback; no regional commercial price list is sourced here | Cost guide |
| Grid connection | Capacity and connection terms differ by network area | Network capacity maps |
| Rates and funding | Business rates treatment and funding schemes differ by nation | Business rates, funding |
Why are there no regional payback figures?
DESNZ's solar PV cost data page gives the cost per kW of solar PV installed by month, with the 2025 to 2026 data published on 28 May 2026. DESNZ says the data are taken from the MCS Installation Database, and MCS is the UK's quality mark for small-scale systems. The figures sit in spreadsheets, which this page does not use, and the page is not a regional price list for commercial roofs. Without a dated cost and saving pair for each region, a payback figure per region would have to be invented, so none is given. The payback, ROI and IRR page explains how to read a proposal's own figures.
How should you use this with a proposal?
- Divide the proposal's annual kWh by its kWp and compare it with the nearest row.
- Ask for the on-site share from half-hourly meter data, and the export rate used.
- Check the capital cost per kWp against the cost guide's method.
- Run your own figures through the savings calculator, and send an enquiry through this site's form if you want a roof looked at.
Sources
- European Commission Joint Research Centre: PVGIS 5 user manual, and 13 PVcalc runs (PVGIS 5.3, south 15 degrees, 1 kWp, 14% losses) for Exeter, Cardiff, Norwich, Reading, London, Birmingham, Nottingham, Leeds, Newcastle, Belfast, Manchester, Glasgow and Inverness, retrieved October 2026, checked 22 October 2026.
- Department for Energy Security and Net Zero (GOV.UK): Solar photovoltaic (PV) cost data, last updated 28 May 2026, checked 22 October 2026.
- MCS: MCS role and scope page, checked 22 October 2026.
- Ofgem: SEG scheme page, checked 22 October 2026.
Common questions
Is commercial solar worth it in Scotland?
Yield is lower: PVGIS gives about 796 kWh per kWp for Glasgow against 987 for Exeter on the same settings, which stretches payback by about 1.24 times if nothing else differs. Whether it is worth it then turns on the on-site share, the export terms and the capital cost, none of which this page can source by region.
Why does the table use a 15 degree tilt?
It sits between the 10 degree and 35 degree layouts on the yield by region page, so these figures are a different cut of the same tool, not a repeat of that table. Check the tilt and orientation in your own proposal against the nearest row.
Does a lower yield always mean a proportionally longer payback?
Only if cost, tariffs and the on-site share are the same. A lower yield per kWp means a given building uses a larger share of each kWp's output, which narrows the gap, so treat the multiplier as a ceiling on the effect for the same building.
Where do I find a payback figure for my own site?
From a proposal built on your half-hourly meter data, with the cost, on-site share and export rate stated. The savings calculator on this site gives a first-pass projection from your own inputs.