The State of UK Solar 2026

1. Deployment is at a record, and the shape of the series explains British solar policy

Three features, and each is a policy decision rather than a market one.

The 2015/16 peak and the collapse after it. 100,088 installations, then 26,094 — a drop of nearly three quarters in a single year. That is the Feed-in Tariff being cut, and it is the clearest illustration available of what subsidy withdrawal does to a domestic market.

The long flat decade. From 2016/17 to 2020/21 the market ran at between a quarter and a little under half of its 2015/16 level, bottoming at 25,565 in 2017/18. It has since grown more than tenfold from that floor.

The recovery is not a subsidy. Deployment has risen every year since 2020/21 and is now far above the Feed-in Tariff peak, with no equivalent subsidy in place. Energy prices did that, not policy.

2. Solar has not been getting steadily cheaper

This is the finding most likely to surprise, and it is not a modelling result — it is the government’s own series.

On the basis the source uses for its long run, the median installation in the smallest size band cost £1,897 per kW in 2013/14 and £1,780 per kW in 2025/26. That is about 6% lower after thirteen years, in cash terms, before any adjustment for inflation.

The path between those endpoints is the real story: a trough of £1,481 in 2019/20, then a rise of 53% to a peak of £2,263 in 2023/24, and a fall of 21% since.

Why the story everybody tells is wrong. Module prices did collapse — spectacularly, and over a longer period than this series covers. But modules are a minority of an installed price. Scaffolding, labour, electrical work, inverters, certification and the installer’s margin are the rest, and those move with the UK economy rather than with global manufacturing.

The size effect is large and consistent: the median 10–50 kW installation costs about 25% less per kW than the median 0–4 kW one. That is why cost per kW, not total price, is the only sane way to compare two quotes — and why a quote that gives only a total is making comparison harder on purpose or by accident.

New build is cheaper again: in the latest year the mean domestic new build cost £1,506 per kW against £1,817 for retrofit, about 17% less. The scaffolding is already up and the roof is open.

3. Where you live matters far less than almost everyone assumes

We modelled 1,356 UK places — 995 towns and all 361 council areas — one PVGIS run each, at each place’s own postcode centroid.

The headline spread is 696 to 1,074 kWh per kWp, about 54% from worst to best. That figure gets quoted, including by us, and on its own it is misleading.

The middle half of UK places sit between 864 and 936 kWh per kWp — about 8% apart. Seventy-two per cent of the places we model fall between 850 and 975.

The 54% is the distance from the sunniest corner of the Sussex coast to the Western Isles. It is not a choice anybody is making. For the overwhelming majority of British households, the honest answer to “is it sunny enough where I live” is that the question is close to irrelevant.

4. What your roof faces matters enormously

Same model, same locations, varying only the direction the panels point.

Averaged across our 25 reference locations at 35 degrees, a north-facing array produces 54% of what a south-facing one does on the same roof. East or west keeps 79%.

Put the two findings together and the comparison is stark:

Geography is the factor people ask about. Orientation is the factor that decides.

5. Shade is a winter problem wearing an annual disguise

The worst scenario we model — a tall, wide building due south at 45 degrees of elevation — costs 32.8% of annual generation. Bad, and the sort of number someone might weigh against a discount.

Its winter figure is 75.6%.

In five of our fourteen scenarios the winter loss is more than three times the annual one. Winter is when electricity is dearest, when the house is occupied in the dark, and when a battery has least to work with. An annual average spreads the damage across the summer months, where the same obstruction does almost nothing, and produces a number that reads as survivable.

Anyone assessing a shaded roof on an annual figure alone is being shown the flattering half.

6. The export market has quietly bifurcated

This is the finding we did not expect.

Every one of the ten highest-paying export tariffs carries a condition. Seven require you to buy your import electricity from the same supplier. Three go further and require that supplier to have installed your system — a tier that did not exist in Ofgem’s reporting and that is unreachable retrospectively. If your panels are already on the roof and somebody else fitted them, the best advertised rates in the market are not available to you at any price.

The best rate open to anyone is 6p. We checked the four largest suppliers against their own published terms in October 2026. Scottish Power’s SmartGen — the 12p untied rate in Ofgem’s report, and the best open rate in the market when it was published — now pays 6p, with 12p moved behind being a Scottish Power electricity customer and 15p behind having them install it.

The open market halved while the advertised headline rates went up. The median open rate is 4.1p; the best tied rate is 27p.

What this means for a household with panels and no intention of switching supplier is blunt: the advertised top of the market is not addressable, and the realistic range is 1p to 6p with a median of 4.1p.

It also means the cheapest lever we have been recommending has largely closed. Switching export tariff used to take a typical household’s payback from about twenty years to about twelve. On today’s open rates it takes it from about twenty to about eighteen.

7. The number that decides everything is the one nobody measures

Solar earns in two ways. A unit you use avoids buying at your import rate — 26.32p under the current price cap. A unit you export earns your SEG rate — 4.1p at the median untied tariff.

The same kilowatt hour is worth 6.4 times more if somebody is in to use it.

So the question “is solar worth it” is mostly the question “what share of your generation lands on the expensive side of that gap”. MCS’s own guidance note MGD 003 — the authoritative UK method — puts that share, for a typical household with no battery, at 14% to 23% depending on whether anyone is home during the day.

Those figures are substantially lower than the assumptions we have seen on quotes, and lower than the defaults our own calculators used until we read the guidance properly and corrected them.

Nobody measures this at national scale. There is no published UK dataset of realised self-consumption. The single most important variable in domestic solar economics is estimated from a modelled archetype, by everyone, including us.

8. Two tax changes land on the same day

A scheduling fact with real consequences, which we have not seen set out anywhere.

There is no VAT on domestic electricity in Great Britain from 1 October 2026 to 31 March 2027. It lowers the value of every unit solar displaces, so it makes solar marginally less attractive — and it expires on the same day the zero rate on installing solar expires.

One further wrinkle, because the headline hid it: when VAT came off on 1 October, the cap’s electricity unit rate went from 26.11p to 26.32p — up 0.8%. Before tax, it went from 24.87p to 26.32p — up 5.8%. Removing VAT absorbed most of an underlying increase rather than delivering a cut.

Full detail: VAT on electricity bills.

9. What nobody publishes

The most useful section of a report like this is the list of things it cannot tell you. Each of these is a gap we went looking for and failed to fill.

What a home battery costs. Our installed solar costs come from an official statistics series built on MCS installation records. No equivalent series exists for domestic storage that we have found. Every battery payback figure in the UK market is therefore built on a price nobody publishes.

How fast batteries degrade in real use. Manufacturers state warranty limits. Nobody measures realised capacity across a fleet of British homes and publishes it.

Which installers do good work. MCS holds the certification data and its terms prohibit copying, distributing or storing the content. There is no other complete source, so no honest national directory can be built by anyone who is not MCS.

Whether the models are right. Every generation figure in this report, ours included, is modelled. We have found no UK dataset of metered output by location, orientation and pitch that would let anyone check modelled against actual. This is the gap we would most like closed, and it would take one willing dataset holder to close it.

What solar does to a house price. The data exists — EPC records carry solar, and sold prices are published — but joining them at scale needs a bulk extract we have not yet obtained. It is on our list rather than in this report.

Real self-consumption. As in section 7. The most important number in the subject is estimated, universally, from a modelled archetype.

10. What we would tell a household, on this evidence

Compare on cost per kW, not total price. The median is £1,595 for 0–4 kW and £1,546 for 4–10 kW on the latest figures. A quote well above that needs a reason you find convincing.

Stop worrying about your postcode. Unless you are in the far north or the islands, your location is an 8% question.

Start worrying about your roof and your shade. Orientation can cost 46%, and a shaded roof can lose three quarters of its winter output.

Work out your self-consumption before anything else. It moves the answer more than everything else on this page combined, and it is the number a quote is most likely to flatter.

Check the export tariff you can actually have. Not the one at the top of the table.

Do not buy a battery on a payback figure. Nobody can produce an honest one, because nobody publishes the prices.