Methodology: how we check stock, compliance and yield

This page explains where every number on minisolar comes from, what we assume, and where the gaps are. If something here doesn’t match what you see on the site, please tell us (see the editorial policy for how to report errors).

Stock and prices

What we record

For each kit at each retailer we record the price, the stock status, the lead time where the retailer gives one, and the time we last checked. Stock status is always one of:

• In stock
• Low stock (the retailer says stock is limited)
• Pre-order
• Back order (with a lead time, if shown)
• Out of stock
• Discontinued
• Unknown

Prices are the prices the retailer shows on its own site at the time we check. Delivery costs, discounts at checkout and basket-level offers may not be included. Always check the final price at the retailer before you buy.

How we check

We read each retailer’s own public data, in this order of preference:

1. Shopify product data. Many retailers run on Shopify, which publishes product, price and availability data in a standard machine-readable format.
2. Structured data in the product page. Many sites include standard product information (schema.org Offer with price and availability) for search engines. We read that.
3. Official feeds and APIs. Where a retailer offers a product feed or API, directly or through an affiliate network, we use it once we have access.

A few retailers publish none of these. For some we read the stock text shown on the page (for example “In stock” or “Sold out”).

We check politely:

• We read each site’s robots.txt and don’t fetch anything it disallows. If we can’t read a site’s robots.txt because the server returns an error, we treat the whole site as off limits until we can.
• We limit how often we request pages from each site, and follow any slower crawl delay the site asks for. If a site tells us to slow down, we back off.
• Our checker identifies itself as minisolar-bot and includes a link to our data sources page and a contact email address, so a retailer can reach us.
• We never add items to baskets, touch checkouts, log in, or try to get round bot protection.

Retailers we can’t check automatically

Some large retailers block automated requests, or only show stock after you enter a postcode or address. We don’t try to get round that. For these retailers stock shows as Unknown unless we have a feed or another permitted source. We may still list the retailer so you know where to look.

Change-only history

We don’t store every check. We store a new record only when the price or the stock status changes, plus one “latest” record for each kit at each retailer, which holds the current state and the time of the last check. Price and stock history, “back in stock” updates and our availability figures all come from these change records.

“Checked N min ago” and stale data

Every stock and price shows when we last checked it successfully. If the last successful check is more than 6 hours old, the status shows as Unknown, whatever it was before. This stops out-of-date “in stock” labels from misleading you. In lists sorted by availability, Unknown sits between Back order and Out of stock.

Compliance

The register we use

To be legally used in Great Britain, a plug-in solar device must be listed as Compliant on the Energy Networks Association (ENA) type-test register. The live register is ENA Connect Direct, which has a dedicated “Plug-in Solar” device type. The legal basis is The Plugs and Sockets etc. (Safety) Regulations 1994 and Electricity Safety, Quality and Continuity Regulations 2002 (Amendment) Regulations 2026 (SI 2026/848) and the DESNZ Plug-in Solar Device Interim Product Specification.

We sync every “Plug-in Solar” entry from the register’s public data several times a day. We keep a history of changes, so we can see when an entry is added, changes status or is removed. An entry that disappears from the register stops counting as compliant on our site straight away.

The public register only shows two statuses: Compliant and Non-compliant. Finer statuses exist but are only visible to network operators and manufacturers, so we can’t show them.

Five compliance states

Every kit on minisolar has one of these states:

State	What it means
Compliant	Linked to at least one ENA “Plug-in Solar” entry that is currently Compliant.
Not compliant	Linked only to entries that are Non-compliant or have been removed from the register.
Outside GB rules	The kit has an integrated or plug-in battery, or its inverter is rated above 800 VA. The rules exclude these regardless of the register (SI 2026/848, reg 2(3); Interim Product Specification §1 and §4.1).
Not on ENA register (unverified)	We haven’t linked the kit to any register entry yet. It may be registered under a name we haven’t matched, or it may not be registered at all.
Unknown	Our last successful sync of the register is more than 48 hours old, so we can’t vouch for the current status.

Only Compliant kits appear in the default In Stock views and in our ranked lists. Other kits are still listed, with their state clearly shown, so you can see why they’re excluded.

How we link kits to register entries

A register entry covers a complete kit (panels, inverter and mounting), not just the inverter. The register’s own guidance says inverter-only products can’t be registered as plug-in solar. So:

• We link a kit to an entry only where the entry’s model text names that kit. Manufacturers often register a separate entry for each panel and mounting variant, so one product can link to several entries.
• A microinverter sold on its own is never shown as compliant plug-in solar, even if that inverter is type-tested under a different device type on the register.
• Where a match is a judgement call (for example, the model text is ambiguous or the retailer’s name for the kit differs from the register’s), the editor reviews it before the link goes live.

Every compliant kit page shows its ENA reference, so you can check it yourself on the register.

Yield, savings and payback

One engine produces every yield figure on the site: the calculator, kit pages, region and town pages, situation pages and rankings.

Solar data

• Source: PVGIS version 5.3 from the European Commission’s Joint Research Centre, using the PVGIS-SARAH3 solar radiation database.
• System losses: 14% (the PVGIS suggested default), for cabling, inverter and other losses.
• Monthly output: for every town and region centre, we hold PVGIS monthly output per kW of panels for 8 compass directions (N, NE, E, SE, S, SW, W, NW) and 6 tilts (0°, 15°, 30°, 45°, 60° and 90°, where 90° is vertical, as on a balcony railing or wall). For directions and tilts in between, we interpolate between the nearest values.
• Your postcode: we look up the centre of your postcode district (the first half of your postcode, such as BS3) and use the data for the nearest town we hold.
• Hourly shape: for each network operator region, we hold a histogram of hourly output from PVGIS (years 2022–2023) for each direction and tilt. We scale it to your location’s monthly totals. This lets us model the 800W limit and how much you use at home, hour by hour.

The 800W limit

Plug-in solar inverters are limited to 800W of AC output, while the panels can be larger (up to 2,000W). When the panels produce more than the inverter can deliver, the extra is lost (“clipping”). We apply the kit’s inverter rating, never more than 800W, to each hour of output.

Hourly averages hide short peaks, so our clipping estimate is a lower bound. As an example, for Bristol, south-facing at 35°, using PVGIS hourly data for 2020:

Panel size	Output before the limit	Lost to the 800W limit
880W	943 kWh	0 kWh (0%)
1,200W	1,286 kWh	35 kWh (2.7%)
1,600W	1,714 kWh	205 kWh (12.0%)
2,000W	2,143 kWh	452 kWh (21.1%)

Shading

You choose one of four shading presets. Each reduces output by a fixed share:

Preset	Output lost
No shading	0%
Light (a few hours a day)	10%
Moderate (half the day)	25%
Heavy (most of the day)	45%

These percentages are our own assumptions, not measured values. Real shading depends on what’s in the way and when. Treat shaded results as a rough guide.

How much you use yourself

Plug-in solar saves you money mainly through the electricity you use at the moment it’s generated. We model your daytime use as a steady load, based on the usage band you choose:

Usage band	Steady daytime load	Share used at home (880W south-facing kit, Bristol)
Low: out most of the day	100 W	about 33%
Medium: home some of the day	170 W	about 50%
High: home all day	240 W	about 63%

In each hour, the energy you use is the smaller of the solar output and the load. We set the bands so that “High” lands close to the Department for Energy Security and Net Zero’s figure of 62% self-use for a GB-average home (3,323 kWh a year) occupied all day with an 800W south-facing system at 30° (DESNZ final-stage impact assessment, Table 7).

A steady load is a simplification. Real homes use power in bursts, and hourly figures tend to overstate self-use slightly compared with minute-by-minute data. Where we don’t have an hourly profile for a location, we assume 50% self-use and label the result as approximate.

Money

• Unit rate: we value the electricity you use at the Ofgem energy price cap unit rate for your region (single-rate electricity, paid by Direct Debit), from Ofgem’s regional price cap tables. We update the rates when Ofgem publishes a new cap period. If your tariff differs, your real savings will differ too.
• Your region: we find your network operator region by placing the centre of your postcode district on the official boundary map of the 14 GB network operator areas, published by NESO. Postcode districts that straddle a boundary get the region their centre falls in. Northern Ireland isn’t covered by the Ofgem price cap, and plug-in solar isn’t yet permitted there, so we don’t look up Northern Ireland (BT) postcodes.
• Export: electricity you don’t use goes to the grid. We value it at zero unless you enter an export rate, because most plug-in kits don’t currently qualify for export payments.
• Degradation: we assume panel output falls by 0.5% a year.
• Payback: the kit price divided by your yearly savings, allowing for degradation. We hold the unit rate constant and don’t include installation, electrician or socket costs. If you need a new socket or other electrical work, add that to the price.

Checking our numbers

These are PVGIS annual figures for 1 kW of panels facing south, at 35° and vertical, with 14% losses. We keep these as reference checks: if our engine gives a different answer for the same inputs, something is wrong.

City	South, 35° (kWh per kW per year)	South, vertical (kWh per kW per year)	Vertical as a share of 35°
London	1,018.84	742.87	72.9%
Bristol	1,022.89	738.67	72.2%
Manchester	886.10	646.43	73.0%
Edinburgh	902.72	682.43	75.6%
Belfast	918.76	682.80	74.3%

Source: PVGIS 5.3, PVGIS-SARAH3 radiation data for 2005–2023, free-standing mounting, terrain horizon on. Belfast is included for comparison only: plug-in solar isn’t yet permitted in Northern Ireland.

As a cross-check, DESNZ puts an 800W system at 690 kWh a year for 30° south at average UK sunshine (impact assessment, Table 7). PVGIS gives about 710 to 820 kWh for 800W at 35° south across the five cities above: the same order, somewhat higher.

How we rank kits

Our main ranking is cost per kWh: the kit’s price divided by the electricity it’s expected to produce over 10 years for your setup.

• Price is the best current offer we’ve found: available offers first, then the cheapest.
• 10-year output is the yearly output for your location, direction, tilt and shading, after the 800W limit, over 10 years with 0.5% yearly degradation. It counts all the electricity produced, so your usage band doesn’t change the order.
• Only Compliant kits with a known price and panel rating are ranked. By default we only rank kits you can get now. Some views also include pre-order and back order.
• If two kits have the same cost per kWh, the one with better availability comes first.

Commissions never affect rankings. Whether a retailer pays us, and how much, isn’t an input to the ranking and doesn’t change which kits we show. See our affiliate disclosure.

What we don’t do (yet)

• We haven’t tested any kits ourselves. Specifications come from retailers, manufacturers and the ENA register.
• We don’t model your exact roof, balcony or garden. Shading is a preset, not a survey.
• We don’t give electrical advice. Follow the manufacturer’s instructions and official guidance, and use a qualified electrician if you’re unsure about your wiring.

All data sources and their licences are listed on our data sources page.