Solar payback: the honest arithmetic

The arithmetic, in full

Cost from Solar photovoltaic (PV) cost data, Department for Energy Security and Net Zero, 2025/26 median for the 0-4 kW band; at the mean instead (£7,234) every payback figure is about 13% longer. Generation is our own PVGIS modelling, south-facing at 35°, median of 25 UK locations. Self-consumption rates 14/19/23% are MGD 003 archetypes without storage; 40% and 60% are shown because storage or an EV can reach them. Export rates are the best and median open untied rates in our SEG table. Simple payback only — no degradation, no inverter replacement, no maintenance, and no change in electricity prices.

That is the whole calculation. No proprietary model, no assumptions we have not named.

Why the spread is so wide

Because a unit used and a unit exported are worth completely different amounts. At the price cap rate, a unit you use is worth 26.11p. At the median open untied export rate it is worth 4.1p if you sell it — six times less.

So the question “what does solar save?” is really “what proportion of the generation lands on the expensive side of that gap?” And the answer to that is self-consumption, which MCS’s own guidance puts at 14% to 23% for a typical household without storage.

What simple payback leaves out

Everything here is simple payback — cost divided by first-year value. It is the number everyone quotes, and it is incomplete in four specific ways, all of which we would rather state than hide:

It ignores inverter replacement. The component most likely to fail first, at a cost that lands squarely in the payback window.

It ignores degradation. Panels produce slightly less each year.

It ignores maintenance, and the cost of removing and refitting an array if the roof needs re-covering — see the costs outside the headline.

It assumes electricity prices stand still. They will not. If import prices rise, payback shortens; if they fall, it lengthens.

The first three all make the true figure worse than the table. The fourth could go either way. On balance, treat the table as the optimistic case, not the expected one.

Why we do not print a single payback number

Our rooftop calculator deliberately reports generation and value split between what you use and what you export, and stops there.

That is a considered position, not an omission. A single payback figure requires us to choose your self-consumption rate, your export tariff and a twenty-year electricity price path — three assumptions that between them swing the answer by well over a decade. Choosing them for you and printing one number would look more helpful and be less true.

What we will do is show you the grid, name every input, and let you find your own row.

Where solar pays back fastest

From the arithmetic rather than from enthusiasm:

• Someone home during the day, using power as it is generated
• A battery, which MGD 003’s tables show lifting self-consumption substantially
• An EV charged at home in daylight
• A heat pump — MGD 003 excludes such loads from its method and says so, meaning its self-consumption figures are likely conservative for those households
• A good export tariff, which is free to switch to
• A quote at or below the median cost per kW — see the benchmark

And where it does not

• An empty house on weekdays with no storage, on a poor export rate. That is the 25-year row, and it is a real row.
• A small array at a high price per kW — see 2 kW systems.
• A shaded or north-facing roof, where generation itself is reduced before any of this arithmetic starts — north-facing roofs.

None of that makes solar a bad idea for those households. It makes payback the wrong question for them, and the honest thing is to say so rather than to adjust the assumptions until the answer improves.