Plug-in solar or a proper roof system? An honest comparison

Generation: the ratio is about five to one

Generation scales with panel capacity, so this part is arithmetic rather than argument. Using our own figures — generated from PVGIS at 20% system losses, the same dataset behind our calculator — for panels angled 35° facing south:

Across the 25 reference locations in our dataset, an 800 W kit at that angle yields between 554 and 834 kWh a year, and a 4 kWp array between 2,768 and 4,168 kWh.

Two things that table understates.

A plug-in kit is often not at 35° facing south. It is frequently vertical on a balcony railing, where the same 800 W of panel yields materially less — about 539 kWh in London rather than 735. A roof array is usually pitched at something closer to optimal, because that is what the roof is.

A roof system is not capped at 800 W. The specification caps a compliant plug-in device at 800 VA of output regardless of how much panel you attach. Roof systems are sized to the roof and the demand.

Export payments: only one of them can be paid

This is the difference people discover after buying, and it is structural rather than incidental.

A rooftop system installed by an MCS certified installer can meet the Smart Export Guarantee’s conditions — Ofgem requires applicants to “demonstrate that their installation and installer are suitably certified”, plus an export meter capable of half-hourly readings with an export MPAN. A self-installed plug-in kit meets neither, and Ofgem’s position is then that a licensee “is not obliged to offer payments under the SEG”.

The practical consequence: with a roof system, generation you do not use still earns something. With a plug-in kit, generation you do not use is simply lost. That is why our calculator makes self-consumption the central input, and why the gap between the two options is wider than the generation ratio alone suggests.

Full detail: why a plug-in kit gets no SEG payment.

Where plug-in genuinely wins

Not a consolation list. These are the reasons the government created the route.

You do not own the roof. Renters and leaseholders generally cannot install a rooftop system at all, and the consultation response records respondents arguing that “residents often do not control access to their roofs (e.g. in rented accommodation) so they should be given the right to install plug-in solar devices”.

It is reversible. A compliant device must have an attachment method that is “reversible and non-permanent”. You can take it with you. A roof array stays with the building.

No installer, no scaffolding, no roof survey. The whole premise is consumer installation, which is why the plug amendment was needed at all.

It is an order of magnitude cheaper. We do not publish kit or installation prices — we do not track them, and we are not going to estimate — but the difference in scale between a panel or two and a full roof installation is not in dispute.

Your roof is unsuitable. Shading, orientation, condition, or a roof that is not yours to drill into. A north-facing roof and a south-facing balcony is a real combination, and in that case the balcony wins on the figures.

Where plug-in loses, beyond generation

• No export income, as above.
• One device per household for now, pending the G98 amendment — so you cannot simply add kits until the numbers work. See the 800 W limit.
• No batteries. Battery-integrated products are outside the specification entirely, and a compliant device must carry a warning that it is not to be used with battery storage.
• Buildings restrictions. Prohibited on several cladding types, timber balconies, and buildings under external wall remediation — see fire safety and cladding.
• Permissions. Renters need landlord consent, leaseholders need the freeholder’s; a homeowner with a roof usually needs neither.

The comparison nobody frames properly

The honest framing is not “which is better” but which question you are actually asking.

If the question is what is the best way to generate electricity at my house, and you own a house with a decent roof and can raise the capital, the answer is a roof system. Five times the output and an export tariff is not a marginal difference, and any page that presents plug-in solar as a serious alternative for that person is either not doing the arithmetic or is selling something.

If the question is what can I do given that a roof system is not available to me, plug-in solar is a genuine answer where until 27 August 2026 there was none. That is the whole point of the route, and it is a good deal more useful than pretending it competes with a roof.

What we could not confirm

Costs. We do not publish kit prices, installation costs or payback periods anywhere on this site yet, because we do not have a maintained source for them. Until we do, the cost side of this comparison is qualitative, and you should treat any page giving you a confident payback figure for either option with suspicion about where the numbers came from.

Whether you can run both. We found no source addressing a plug-in device alongside an existing rooftop system — the G98 notification position, and whether the one-device limit interacts with an existing installation, are not things we can answer from the documents. Ask your network operator.

Whether a plug-in kit complicates a later roof installation. Also unsourced. If you are buying a kit as a stopgap, worth asking an MCS installer before rather than after.