What does a structural roof survey check before a solar array goes on?

A roof under a solar array is carrying more than just the weight of the panels. A structural survey for a commercial system assesses four distinct load cases and then combines them to find the worst case the structure has to survive.

• Dead load. The permanent weight: the modules, the mounting rail or ballast, the fixings and the existing roof build-up itself.
• Imposed load. The variable load from use, principally maintenance access onto the roof.
• Wind load. On a lightweight roof, wind uplift is usually the governing case. The wind tries to lift the array off the deck, not press it down, and that is often the action the structure is least able to resist.
• Snow load. The downward load from accumulated snow, which varies sharply with location, altitude and roof geometry.

These are quantified using the UK Eurocode actions suite and combined to the basis-of-design standard, BS EN 1990. Wind actions come from BS EN 1991-1-4 and its UK National Annex; snow actions from BS EN 1991-1-3 and its National Annex; the member and connection checks run to the relevant material Eurocode, for example BS EN 1993 for steel. The point a finance director should hold onto is that wind and snow loads are site-specific. They vary by geographic location, altitude, building height and roof shape, so a generic figure from a brochure is not valid for any real roof. BRE Digest 489, the PV-specific wind reference revised in 2014, sits on top of the wind Eurocode for rooftop arrays. On a coastal or otherwise exposed frontage the wind case is harder again, and the zoned holding-down pattern that answers it is set out on our coastal and exposed sites page.

On a warehouse or factory roof the structure rarely fails on gross weight alone. The survey traces the load path from the panel clamp, to the rail, to the fixing, to the sheet or purlin, to the rafter or portal frame, and finally to the column and foundation. The array is only as good as the weakest link in that chain, and on industrial roofs the weak links are usually near the top.

The typical limiting elements are the purlins, the roof sheet and the fixings. The assessment checks purlin span and section capacity, often against the manufacturer's span tables; the capacity of the profiled sheet spanning between purlins; and the pull-out and shear capacity of both existing and new fixings under wind uplift. This is also where weather-tightness can be lost: drilling or fixing into purlins or sheets without a proper assessment can compromise the roof's ability to keep water out, which is a separate failure from a structural one.

As an indication only, a rail-mounted array typically adds in the order of 15 to 25 kg per square metre, while a ballasted flat-roof system can add far more, roughly 50 to 80 kg per square metre, because of the ballast needed to resist uplift. Those are typical ranges, not design values. The real number for your roof comes from the project-specific calculation, and a ballasted flat-roof install is one of the cases that routinely needs a structural engineer because of the extra load it imposes.

Structural capacity and roof condition are two different questions, and a thorough survey answers both. A roof can be strong enough to carry the array and still be the wrong roof to install on, because it does not have enough life left in it.

A solar array has a design life generally taken as 20 to 30 years. Industry guidance from the NFRC is that the roof beneath it should have a comparable remaining lifespan, so the roof does not fail before the asset on top of it does. Installing on an end-of-life roof risks premature roof failure, water ingress beneath the array, the cost of removing and reinstating panels to get at the roof, and disputes over who is responsible for the consequences. The recommended sequence is to address the roof first: where remaining life is inadequate, re-roof or remediate before the array goes up, rather than building over a problem.

This is also why an on-site assessment beats a desktop or satellite estimate. Satellite and aerial imagery cannot reliably show the true condition of existing fixings, corrosion, the state of the sheet or membrane, or near-field obstructions. A drone survey delivers far higher geometric accuracy and can carry thermal imaging to flag moisture ingress, while an actual structural assessment establishes the real load path and remaining capacity. A desktop estimate can produce a plausible-looking layout on a roof that physically cannot take it, or is near the end of its life.

Building Regulations Approved Document A (Structure) always applies to a rooftop array: the structure must be shown by calculation to carry the worst-case combination. For systems up to 50 kWp, which sit inside the domestic-scale MCS scheme, a competent assessor can sign off simple standard roofs and a qualified structural engineer is required for the higher-risk cases.

An engineer, typically chartered through the IStructE or with CEng MICE, should be engaged rather than a general assessor where the roof is non-standard or higher-risk. The cases commonly cited include hipped, valley or asymmetric roofs, dormers and parapets, shallow pitches, ballasted flat-roof installations, older buildings, previous structural alterations, thin metal sheeting or ageing purlins, any visible signs of distress, and high wind-exposure zones.

For a commercial system over 50 kWp this is the key honesty point. That MCS competent-assessor route does not apply: the system sits outside MCS scope, so MCS is not the compliance mechanism at all. The project relies on Building Regulations plus a structural engineer's Eurocode-verified report instead, and that sign-off effectively becomes the expected route. The report concludes with a clear determination: a Pass, a Conditional Pass, or a Fail, supported by the load calculations and a written engineering narrative. A Conditional Pass sets out defined next steps, for example a lightweight non-penetrating mounting system, a reduced array density, or localised strengthening, so you know exactly what the roof will and will not take before you commit.

A structural roof survey is a physical site visit that gathers the evidence the calculation later depends on. The surveyor first establishes safe access under a method statement, because every measurement taken on a fragile or lightweight roof is governed by the work-at-height duties planned under CDM 2015. Once on the deck they record the build-up: the sheet or membrane type and profile, the purlin section and spacing, the rafter or portal-frame layout, the fixing pattern, and the span dimensions that drive the member checks. Where the structure is concealed, the original structural drawings are requested and reconciled against what is visible, because a roof rebuilt or re-clad since construction rarely matches its as-designed paperwork.

The visit also captures the condition evidence that no desktop study can produce. The surveyor probes for corrosion at fixings and laps, checks for sheet fatigue and ponding, looks for prior structural alteration, and notes any visible distress such as sagging purlins or cracked welds. On an older industrial roof the asbestos position is read alongside the structure, because the two cannot be separated on this fabric, a point developed in our asbestos on older roofs guide. This is also the stage where a measured roof model is built, whether by hand survey or by drone photogrammetry, so the array layout sits on the true geometry rather than an aerial approximation. The findings feed directly into the wider feasibility study and the usable-area question in our how much roof space guide, so the structural visit pays for itself by killing unviable layouts early.

On a commercial solar project the structural survey is the first design-risk control under the Construction (Design and Management) Regulations 2015. Because a system over 50 kWp sits outside the domestic MCS scheme, CDM 2015 is the framework that governs how design risk is identified and discharged, and the structural assessment is where the principal designer first tests whether the array can be built safely at all. The HSE, which enforces CDM, treats the choice of mounting system and the loading it imposes as a design decision, so the engineer's load model and the principal designer's risk register are read together rather than in sequence. Our CDM 2015 guide sets out how those duties are allocated across the project.

The survey also shapes the rest of the design, beyond the go or no-go decision. A Conditional Pass that calls for a lightweight non-penetrating mounting system, a reduced array density, or localised strengthening becomes a design input that the principal designer carries forward into the installation method and the maintenance access plan. A penetrating fixing into a fragile sheet, for instance, is both a structural action and a weather-tightness and safety risk, so the survey result and the CDM risk control are the same decision viewed from two sides. The resulting engineer's report and load calculations also matter commercially: they are the document an insurer and a future buyer rely on, which is why we treat them as part of the asset record alongside the cover described in our commercial solar insurance guide.

On a modern shed the structural check is short because the design intent is recorded. Someone sized the purlins for a stated imposed load, and a framed array either sits inside that allowance or it does not. Drawings and a spot check answer it.

An older trussed building has no design file. The trusses were sized for the covering, wind and snow expected of them at the time, with whatever margin the engineer of the day chose, and nobody wrote the margin down. Worse, it has usually been spent since: over-boarding for insulation, a heavier replacement covering on one bay, a plant platform dropped in during a conversion, a shaft line cut out and its members left weakened.

So the read has to be taken from the building itself. Truss type, spacing and member sizes. The condition of timber at the bearings and at the ends where rot and beetle sit. The state of the ironwork at the connections. Every alteration anyone can point to. Against all of that goes the array load, dead weight plus the wind case calculated to BS EN 1991-1-4, which on an exposed older roof matters more than the modules do.

What comes back from a read of this kind is rarely a yes or a no across the whole building. It is normally a capacity stated bay by bay, and the useful version says where the capacity is rather than merely that some exists. The array is then laid out to the bays that pass, which is exactly why the model and the read are done before the layout rather than after it.

Two consequences follow. The array gets smaller, so the sizing conversation moves off roof area and onto what the load can actually absorb. And a strengthening scheme on a handful of bays is occasionally proportionate, while on other buildings it plainly is not. Both are worth saying out loud at survey stage rather than designing around quietly.

On a flat or shallow roof there are two ways to hold an array down, and an older structure makes both of them uncomfortable. A ballasted system avoids penetrating the waterproofing, and the cost is dead load. Uplift is highest at the perimeter and in the corners, so a ballasted layout is heaviest around the edges of the roof plate, which on an older building is where the trusses are shortest, the bearings oldest and the parapet detailing least forgiving. On a structure with no measured margin, ballast is often the thing you cannot afford.

The alternative transfers load into structural members, purlins or trusses, rather than spreading it across a deck that was never asked to take it. That is frequently the better engineering answer on older stock, and it brings the detailing at every fixing, the remaining life of the covering and any roof warranty into the design at the same time.

The third route is to need less holding down. Low-tilt east-west layouts cut uplift and therefore ballast, lighter framing helps, and placing fewer modules on the bays that passed the read beats spreading them across bays that did not. On a pitched slate roof neither flat-roof answer applies: hooks go to the rafters and slates are dressed back around them, which on stone slate is slow, skilled work that a programme has to carry honestly.