Cost per kWp: the only figure that compares two quotes

Three quotes, three totals, and no obvious way to compare them. One division fixes that: divide each total by the system size in kilowatts peak.

Cost per kWp is the only figure that makes differently sized systems comparable, it takes about thirty seconds to work out, and, unusually for this market, there is official data to hold your results against once you have them.

The calculation

1. Find the system size on each quote, in kWp. It is usually near the top, sometimes expressed as panel count multiplied by panel wattage. Twelve panels at 440W is 5.28 kWp.
2. Take the total price, including VAT, which for a qualifying domestic job should currently be zero, so the total and the ex-VAT figure should match.
3. Divide.

A worked set: £8,900 for 5.28kWp is £1,686; £7,200 for 4.0kWp is £1,800; £6,400 for 3.52kWp is £1,818. The £2,500 spread in totals was mostly capacity: on a per kWp basis the “dearest” quote is the cheapest, and the real conversation is about why the smallest system carries the highest rate, which the fixed-cost section below answers.

Then anchor against the official medians

The Department for Energy Security and Net Zero publishes per kilowatt costs from real certified installations. For 2025/26 the medians are £1,780 per kilowatt for systems up to 4kW and £1,697 for 4 to 10kW, with regional means showing Scotland, Wales and the North East cheapest and London and the South West at the top.

That gives your normalised numbers an external reference no installer controls. A quote well above the medians needs a nameable reason; one far below them deserves suspicion, because the medians describe what real, certified, warranted work costs. All three quotes in the worked set sit inside the plausible band, which is itself information: nobody is trying it on, and the decision moves to specification.

Why it works

Almost everything that varies between quotes scales with capacity or does not. Dividing by capacity normalises the first group and exposes the second.

What remains is a genuine difference, and it will be one of four things:

Equipment tier. Better panels, a better inverter, longer warranties.

Per-panel electronics. Optimisers or microinverters, specified because the roof has shading or splits across orientations. Materially more expensive and often correct, which is why normalising quotes is step one of comparing them, not the whole job.

Site difficulty. A slate roof, restricted access, a long cable run, an old consumer unit.

Margin. Sometimes simply higher.

Only the last is something to push back on, and you cannot see it until the other three are accounted for.

Why small systems look expensive

The fixed costs are the reason, and the full breakdown itemises them: scaffolding priced by the property, the survey, certification and network notification unchanged by size, one inverter covering a range of capacities.

Spread across a small system, those fixed costs dominate, and cost per kWp is poor, which is exactly what the official data shows in the gap between its two size bands. Spread across a larger one, they dilute.

This is the most useful thing the metric tells you: adding capacity to a job that is happening anyway is cheap, and the marginal panel costs far less than the average. If your roof and your consumption both support more, the extra capacity is the best-value part of the whole purchase, and the sizing question is worth revisiting before you accept the quoted size.

The trap

Cost per kWp is a comparison tool, not an objective to minimise.

The cheapest cost per kWp is achieved by fitting the largest possible array with the cheapest possible equipment. That is not the best system if:

• The extra generation exceeds what you can use, so it exports at a rate less than half your import rate.
• The roof has shading that cheap equipment handles badly.
• The panels chosen are physically larger and you are area-constrained.

Use it to compare quotes for similar specifications. Do not use it to choose a specification.

What to record for each quote

• Total price including VAT.
• System size in kWp.
• Cost per kWp.
• Inverter type, and whether it is battery-ready.
• Estimated annual generation, and the tool that produced it.
• What is excluded.

Six lines per quote. With those in front of you, the conversation with an installer changes completely, because you are asking about specific differences rather than asking them to justify a total.

The follow-up question

Once you have the figures, ask the dearest installer:

“Your cost per kWp is higher than the others for a similar specification. What does that buy me?”

A real answer names something concrete: a harder roof, better equipment, optimisers for shading you had not noticed. A vague answer about quality is telling you the difference is margin.

Then check what it returns

Cost per kWp compares purchases. It says nothing about whether the purchase is worthwhile, which depends on your consumption and your tariff.

Take the normalised cost and a conservative generation estimate, apply your own import and export rates separately, and run the projection. The worked 4kW example shows the shape: a median-priced system, honestly modelled, pays back in roughly a decade for a typical household, and every hundred pounds per kilowatt you save against the median arrives straight off that payback. That is the number the decision actually rests on.