Ground-mount vs rooftop solar: how to decide for a commercial site

Most commercial solar conversations start with the roof, because that’s usually where the panels end up. But “put it on the roof” is a default, not a decision. On a meaningful share of the sites we assess, it turns out to be the wrong one.

Roof condition, available land and planning policy can all point somewhere else. So can straightforward cost. Getting the choice wrong is expensive to unwind once a specification has gone to tender.

This is the framework we use to decide between rooftop, ground-mount and a car-park canopy. It’s the middle option most people forget.

When roof condition rules rooftop out

A modern solar array adds roughly 18–25 kg per square metre of permanent dead load to a roof. That load is modest next to the wind and snow loads the original design allowed for. But it’s not negligible on a building carrying decades of wear. The questions that actually decide viability are:

• Structural capacity. Look at purlin and truss condition, corrosion, and how much of the original design’s load margin is still available. That’s not the capacity shown on the drawings from when the building went up.
• Roof age and remaining life. Suppose a membrane or sheet covering is due for replacement within five to ten years. That timeline changes the economics of fitting an array now. A combined re-roof-and-solar project is often the better call than solar now, re-roof later.
• Covering type. Older industrial roofs sometimes used asbestos cement sheeting, which restricts direct fixing under the Control of Asbestos Regulations 2012. That usually forces a re-roof or overlay before panels can go on at all.
• Usable, unshaded area. Rooflights, plant, HVAC ductwork and party-wall shading can all eat into a roof’s real capacity well before its edges do.

Any one of these can turn a roof that looks promising from the ground into a poor candidate. Or it can mean solar only makes sense bundled with re-roofing works. That’s exactly why a structural survey is the first thing we do on a feasibility study. It’s not an afterthought bolted on once someone has already assumed a system size.

When land availability makes ground-mount the better fit

Ground-mount earns its place in two situations. In the first, the roof genuinely can’t take the load a business wants. In the second, a site simply has spare land that would otherwise sit idle. That could be a former yard, an unused paddock, or a buffer strip around an industrial estate.

Ground-mount also has a real technical advantage over most existing roofs. You choose the tilt and orientation from scratch. You’re not stuck with whatever pitch and aspect the building happens to have. That freedom can lift yield per kWp installed.

The catch is planning. Ground-mounted arrays above a fairly modest scale need full planning permission from the local authority. One factor surprises people who assume this is only relevant to farmland.

Agricultural Land Classification (ALC) grade matters even when the site sits next to an industrial estate or business park. ALC grades undeveloped land 1 to 5 by its soil and climate quality. That grading applies regardless of who owns the land or what’s built next door.

ALC grades a paddock or field beside a warehouse on the same basis as one in open countryside. National planning guidance treats Grades 1, 2 and 3a as “best and most versatile” (BMV) land. Planning policy then steers ground-mount solar away from BMV land and towards Grade 3b and below.

If nobody has ALC-surveyed the spare land attached to your site, get that done early. Do it before a lease, a specification or a planning application, not after.

Two other checks belong in the same early pass:

1. Grid connection. Ground-mount schemes above the smallest domestic scale connect under Engineering Recommendation G99. That’s the same DNO application process that applies to rooftop systems of similar size. Land doesn’t exempt you from it.
2. Whether developing it yourself is even the best use of the asset. The land might be BMV-graded, awkwardly shaped, or better suited to a third party’s scale of development. In that case, leasing it to a solar developer can outperform building your own array on it. That lease often comes with an index-linked rent. See our land lease vs on-site solar guide for how that comparison works in practice.

The car-park canopy: the option most people skip

Between “roof” and “field” sits an option that fits a specific gap well. It’s a solar canopy over an existing staff or HGV car park. It doesn’t depend on roof condition at all — it’s a new-build steel structure.

And it doesn’t need spare land beyond what’s already tarmacked over for parking. It also pairs naturally with EV charging, turning generation and weather-protected parking into one project rather than two.

Factor	Rooftop	Ground-mount	Car-park canopy
Typical installed cost	Baseline (roughly £700–£1,100/kWp at commercial scale)	Roughly 10–20% above rooftop — commonly £1,000–£2,500/kWp more in absolute terms, for foundations and land works	Roughly £950–£1,300/kWp, reflecting the canopy structure
Main constraint	Roof structure, age and covering	Planning, ALC grade, grid connection	Existing car park area and layout
Planning position	Usually permitted development or a straightforward consent	Full planning permission almost always required; ALC/BMV policy applies	Usually no full planning permission needed — since December 2023, Class OA permits car-park canopies via prior approval instead, subject to a 4m height limit and siting/design/glare checks
Best used when	Roof structure and area both check out	Roof is constrained and suitable land is available and lower-grade	Roof is constrained, land is limited, or EV charging is also wanted

That £1,000–£2,500/kWp premium for ground-mount over an equivalent rooftop system reflects real costs a roof install doesn’t have. These are foundations, cabling runs and land preparation. That premium isn’t a reason to rule ground-mount out where the land is right.

It is, however, a number worth modelling properly rather than assuming it. Pair it with the capex and payback figures for your system size. A canopy’s higher per-kWp cost buys something a roof array can’t offer. You get a structure independent of an ageing roof, and often a mounting point for EV charge points.

Getting the sequence right

None of these three options is inherently “better.” The right one depends on your specific site. That means roof condition, spare land and its ALC grade, and the grid position each option would connect through. The mistake that costs money is picking one before those facts are in, rather than after.

A proper feasibility study runs the structural survey, the land and planning check, and the grid enquiry in parallel. It then produces an honest recommendation. That recommendation covers which of rooftop, ground-mount or canopy — or some combination — actually fits your site. That happens before you commit any capital or specification work.

Sources: GOV.UK / Natural England — Guide to assessing development proposals on agricultural land (ALC grades & BMV); Energy Networks Association — G99 connection guide; HSE — Control of Asbestos Regulations 2012; The Town and Country Planning (General Permitted Development etc.) (Amendment) (England) Order 2023 (SI 2023/1279), Article 8 — insertion of Class OA.