How Many Solar Panels Does a Commercial Building Need?
27 June 2026 · SEO Dons Editorial
How commercial solar systems are sized — from half-hourly load, not roof area — with worked examples by building type and the self-consumption that drives returns.
A commercial building should be sized from its electricity demand, not its roof. The right system is the one whose output you consume on site for most of the year, because every unit you use yourself displaces grid power at 24–28p/kWh, while every surplus unit exported earns only 12–16p under the Smart Export Guarantee. So the question is not “how many panels fit?” but “how much daytime load is there to feed?” Get that answer first, and the panel count follows.
In practice a mid-sized commercial roof carries somewhere between a 50kWp and a 1MWp system, and a useful rule of thumb is that 1kWp of panels needs roughly 5–6 m² of usable roof and generates around 950 kWh a year in the UK. But the number that matters is set by your half-hourly consumption profile, so that is where any honest sizing exercise begins.
Size from load, not roof area
The data that should drive the design is your half-hourly meter data — the readings your electricity supplier records every 30 minutes. They show two things a panel count alone cannot: how much you use, and crucially when. A building that consumes 250,000 kWh a year but does most of it overnight in cold stores is a very different proposition from a 250,000 kWh office that runs 8am–6pm. The first has little daytime load to soak up solar; the second is close to a perfect match.
What you are trying to do is shape the array so that its midday output lands underneath your daytime demand curve as often as possible. Match generation to consumption and you get high self-consumption — the single biggest driver of return. Overshoot it and you spill the excess to the grid at the lower export rate.
A simple first-pass calculation works like this. Take your annual consumption in kWh, decide what share of that you realistically want solar to cover (the displacement target), then divide the kWh you want to generate by ~950 to get the kWp. A site using 200,000 kWh a year that wants solar to cover a quarter of it needs about 50,000 kWh of generation, which is roughly 53kWp. That is the starting point; the load shape then pulls the number up or down.
Worked examples by building type
The figures below are indicative ranges, not quotes — every site turns on its own meter data, roof and grid connection. They show how building type maps to a sensible system size and the self-consumption you can expect.
| Building type | Typical annual use | Indicative system | Roof area needed | Self-consumption (no battery) | Capex ex-VAT |
|---|---|---|---|---|---|
| Small trade unit / showroom | 30,000–60,000 kWh | 30–50kWp | ~150–300 m² | 50–70% | £35–60k (50kWp) |
| Office or retail (mid) | 80,000–250,000 kWh | 100–250kWp | ~600–1,500 m² | 40–60% | £82–240k |
| Light industrial / workshop | 150,000–400,000 kWh | 150–350kWp | ~900–2,000 m² | 50–70% | £110–300k |
| Big-box logistics / warehouse | 250,000 kWh–2 GWh+ | 250kWp–1MWp+ | ~1,500–6,000 m² | 30–80% | £350k–900k+ |
A small trade unit with a single-shift daytime pattern is the easiest case: a 50kWp array sits comfortably under daytime load and self-consumes most of what it makes. A mid-sized office or retail unit is a good match too, though air-conditioning and lighting concentrate demand in working hours, so 100–250kWp typically lands the array near the demand curve.
Big-box logistics is the most variable. A live distribution centre running chillers, conveyors and charging 24/7 can absorb a 750kWp–1MWp array at very high self-consumption. A lightly occupied storage shed with a vast roof and little load is the opposite — the roof would physically take 1MWp, but the building only consumes a fraction of it during daylight, so a smaller array (or a roof lease to a third party) makes far more sense.
Why bigger is not better
The instinct is to fill the roof. Resist it. Beyond the point where output matches daytime demand, each extra kWp produces power you cannot use on site, so it gets exported. The economics of that are stark: a kWh you self-consume is worth the 24–28p you would otherwise have paid the grid, while the same kWh exported earns roughly 12–16p. You are halving the value of every marginal unit.
Oversizing also drags down the headline metric. A 30% self-consumption system pays back far more slowly than a 70% one at the same capex, because most of its output is being sold cheap rather than displacing expensive imports. The payback range for well-sized commercial solar is typically 4–8 years, and high-load sites that self-consume most of their generation sit at the fast end of that. The way to stay there is to size to load, not to roof.
That said, deliberate oversizing has two legitimate uses: where you expect demand to grow (a new production line, more EV charging), or where you have a route to monetise export through a third-party arrangement. Both are decisions to take with eyes open, not by default.
Battery storage lifts self-consumption
If your roof can take more than your daytime load can use, a battery is the bridge. It captures midday surplus that would otherwise be exported and releases it into evening or overnight demand, lifting self-consumption from a solar-only 30–50% to 60–80%, and to 90–95% on sites with round-the-clock load. That changes the sizing maths: a battery lets you justify a larger, fuller-roof array because the surplus now has somewhere to go at full import value rather than being dumped to export.
Whether storage pays depends on your tariff spread and load shape — it earns its keep where the gap between import and export prices is wide and where demand persists after the sun drops. It is an addition to model alongside the array, not a default. The half-hourly data tells you whether it stacks up.
When the roof is the binding constraint
Load sets the ideal size; the roof sometimes overrides it. The binding constraint is whichever runs out first — your daytime demand or your usable roof.
For most commercial buildings, load is the limit: there is more roof than there is daytime consumption to feed, so you size to demand and leave roof spare (or lease it). But some high-consumption sites — a busy industrial unit on a small footprint — want more solar than the roof can hold. There the roof caps the array, and you take what daylight you can, knowing every unit is self-consumed.
Two practical limits sit alongside roof area. Structural capacity matters: an older roof may not carry the dead load of a full array without strengthening, which an early structural survey will flag. And the grid connection is often the real bottleneck — any system above roughly 50kW needs a G99 application to your DNO, and the connection offer (cost and timescale) can shape the final size more than the roof does. We cover this in the planning and grid guide.
Getting to a real number
The honest answer to “how many panels?” is that nobody can give you one from the roof alone. The credible route runs through your half-hourly data, your load shape, your tariff and your DNO connection — that is what separates a system that pays back in five years from one that exports half its output at a loss. If you want the mechanics of the technology itself, our commercial solar panels page covers panel and inverter specification, and the cost page sets out pricing by system size. Landlords weighing this against the split-incentive problem should also read the office investment property vertical.
Send us your last 12 months of half-hourly data and a roof plan, and we will size the array to your actual load, model self-consumption with and without storage, and give you a system size and payback you can stand behind. Request a quote to start.