Home / 2026 Structural Risk Report / Dataset & Methodology
Dataset & Methodology
The full dataset behind the headline: 575 UK commercial rooftops.
This page publishes the numbers behind the 2026 Structural Risk Report in full: how the sample was built, the outcome distribution, the breakdown by roof type and by building sector, the ranked failure modes and the characteristic load figures. It is free to cite with attribution. If you are a journalist or researcher and want something that is not here, ask and we will get it for you.
575
UK commercial rooftops assessed
35%
Required engineering intervention before PV
78%
Flat-roof ballasted systems needing ballast reconfiguration
62%
Asbestos cement roofs failing wind uplift on first review
How the sample was built
The dataset is every UK commercial rooftop our engineering team assessed for solar PV structural feasibility between 2024 and 2026. It is not a survey, a poll or a modelled estimate. Each entry is a real building that a client asked us to assess before committing to an array, and each was independently assessed by a qualified structural engineer.
| Sample composition | Count | Note |
|---|---|---|
| Commercial projects above 50 kWp | 400 | Outside MIS 3002 scope; assessed directly to the Eurocodes and Building Regulations. |
| Commercial projects at or below 50 kWp | 175 | Within MIS 3002 scope. |
| Total dataset | 575 | UK commercial and light-industrial rooftops. |
| Residential and sub-50 kWp domestic, excluded | 250 | Excluded deliberately. Domestic roofs behave differently and would flatter the intervention rate. |
Building types in the sample span warehouses and distribution or logistics buildings, industrial and manufacturing units, retail units and retail parks including supermarkets, offices, and public-sector buildings such as schools and community centres.
Outcome distribution: where the 35% comes from
The headline figure is not a single measurement. It is the sum of the three outcomes that are not a clean pass. That matters, because it means the number does not rest on one fragile definition.
Intervention rate = 15% conditional + 15% remediation + 5% rejected = 35%. Reported as "one in three" for headline use.
Breakdown by roof type
Share of the 575 by roof construction. Metal sheet of one form or another accounts for 60% of the dataset, which is what you would expect across UK commercial and light-industrial stock.
Bars are scaled relative to the largest category (45%), not to 100%.
Breakdown by sector
Share of the 575 by sector. The dataset is concentrated: logistics, manufacturing and retail account for 75% of all assessed projects between them, which reflects where UK rooftop solar has actually been deployed at scale rather than where it is discussed.
Bars are scaled relative to the largest category (30%), not to 100%. Sector shares sum to 100%.
The five failure modes, ranked by frequency
These are the reasons roofs did not clear on first pass, in order of how often they were the governing issue.
Snow load combined with the array creates incompatible loadingDriven by altitude. The higher the site, the higher the characteristic snow load, and the UK has more high-altitude commercial stock than is generally assumed: upland logistics, hilltop public buildings and rural commercial in elevated areas.
Roof condition precludes fixingsCorrosion, perforation, moisture ingress or asbestos content meaning the covering cannot reliably receive a fixing, regardless of what the frame beneath could carry.
A trigger feature under MIS 3002 V6.0 Section 5.9.6Hipped or valley configurations, low pitch, dormers or parapets, each of which takes the roof outside the standard's own tables.
No documented evidence of the original designNo calculations and no as-built information, so the existing capacity cannot be established from the record and has to be derived from survey.
Wind uplift exceeds the available ballastMost often on flat roofs where the ballast quantity or its layout was set by convenience rather than by the uplift calculation.
Characteristic load figures
For anyone who wants the engineering rather than the percentages. These are dataset averages, not design values for any specific roof.
| Measure | Value | What it means |
|---|---|---|
| Average dead load added by a typical commercial array | 0.75 kN/m² | Across a typical 5.5 kWh/kWp commercial installation, including mounting. |
| Average reserve capacity available on rejected roofs | 0.10 kN/m² | What the structure actually had left to give. |
| Average load deficit on rejected roofs | 0.25 kN/m² | The gap between what was needed and what was there. Rejected roofs were not marginal; they were short by roughly a third of the load being added. |
Three supporting findings
- 62% of asbestos cement roofs in the dataset failed wind uplift adequacy on first review.
- 78% of flat-roof ballasted projects required ballast reconfiguration to satisfy MIS 3002 V6.0 Clause 5.9.13(h), mandatory from 18 June 2026.
- 82% of roof surveys identified defects or maintenance issues capable of affecting long-term waterproofing performance.
Methodology, and its limits
How each assessment was made
- Every project independently assessed by a qualified structural engineer. No software-only desktop reviews are included in the dataset.
- Assessed to the Eurocodes. Wind to BS EN 1991-1-4 and its UK National Annex, snow to BS EN 1991-1-3, PV-specific uplift checked against BRE Digest 489.
- No double counting. Some projects were redesigned mid-project, usually because the client moved the array. Each building is counted once.
- Project-count weighted, not capacity-weighted. A 2 MW site and a 60 kWp site count equally.
Limitations we would want a journalist to know
- This is a client-instructed sample, not a random one. These are buildings where somebody was already considering solar. It is representative of roofs being assessed for PV, not of UK commercial building stock as a whole.
- There is a geographic and altitude bias in the sample, and it is the reason snow load ranks first among failure modes. A dataset weighted towards lowland southern sites would rank differently.
- "Intervention" covers a wide range. It runs from a layout change costing nothing to structural strengthening. The 35% should not be read as 35% needing major works.
- Residential was excluded deliberately. Including the 250 domestic roofs would have lowered the intervention rate materially and made the figure less useful to anyone working on commercial buildings.
How to cite this
Free to use in reporting, research and industry material, with attribution to Solar Surveys Ltd and a link to this page.
Suggested citation
Solar Surveys Ltd (2026). 2026 Structural Risk Report: structural feasibility outcomes across 575 UK commercial rooftops assessed for solar PV, 2024 to 2026. https://solarsurveys.co.uk/risk-report-2026-data
The full report, with the analysis behind these figures, is at the 2026 Structural Risk Report.
Wondering which side of the 35% your roof sits on? Find out before you commit.