Solar Rooftop Analysis

Find the roofs worth fitting with solar panels. We map every roof across a town, a property portfolio or a whole country, and give each one its suitable area, indicative peak power and yearly generation. In the UK and across Europe, including Serbia and the Western Balkans.

Ask about your area Product sheet (PDF, 2 MB)

Background: roofs of 1,000 m² or more in Novi Sad, Serbia, coloured by indicative peak power. Buildings: Overture Maps Foundation, © OpenStreetMap contributors (ODbL), Microsoft building footprints (ODbL).

1,910

buildings screened roof by roof from 1 m LiDAR in one run, in Great Baddow, Chelmsford

See the example

790

roofs of 1,000 m² or more mapped in 28.5 km² of Novi Sad, Serbia, from open data

See the example

26%

more yearly output per kWp in Novi Sad than in London, for the same 10° south-facing frame (PVGIS)

See the example

Which roofs are worth it?

Roofs are the easiest place to add solar: the building is already there, the power is used on site and no land is taken. The hard part is knowing which roofs are worth a visit. We apply one method to every building, combining building outlines, the shape of each roof where a surface model exists and solar yields from the European Commission's PVGIS, and rank every roof by what it could generate, with a plain note of what was assumed.

It is the same work for a solar installer looking for commercial roofs to approach, a landlord deciding which buildings in a portfolio to fit first, or a council planning panels across its schools and offices.

What you get

A ranked list of roofs with suitable roof area, indicative peak power (kWp) and yearly generation (kWh), each with a low to high range

Building use where it is recorded (factory, warehouse, school, offices, flats), so you can filter to the roofs you work with

Roof suitability by pitch and direction when a 1 m surface model is available

The same data as a GeoPackage for QGIS or ArcGIS, and as a spreadsheet

An interactive web map to share with your team, investors or clients

A short report: method, data sources, assumptions and limitations

How it works

Three stages, each ending in something you can use

Start with the stage that answers your question. Most projects begin with an area screen and go deeper only on the roofs that make the shortlist.

Area screen

Every roof in a city, region or country, from building outlines and PVGIS solar yields. Large flat roofs (factories, warehouses, schools, retail) hold most of the power and screen well without 3D data, so this works even where there is no national LiDAR.

You'll see: a ranked roof list and a map of the whole area.

Roof by roof

The pitch and direction of every square metre of roof, from 1 m LiDAR (free across England) or a surface model made from aerial or satellite stereo imagery. South, east and west faces, flat roofs and unsuitable parts, with shading from nearby buildings and trees as an option.

You'll see: a roof suitability map and figures for each building.

Site report

For the buildings you want to pursue: a detailed roof survey (a drone survey flown by a partner drone operator where needed), usable area around plant, skylights and edges, an indicative panel layout and yield, and the checks still needed on roof structure and grid connection.

You'll see: a site report per building, ready for your installer's design.

What the ranked roof list looks like

The top five non-residential roofs from the Novi Sad example below, straight from the area screen. A full list has every roof, with building use, location and both ends of each range.

Five largest non-residential roofs in central Novi Sad with indicative peak power and yearly generation
Rank	Use (as mapped)	Roof area (m²)	Indicative peak power (kWp)	MWh a year
1	Commercial	43,900	3,070 to 5,270	5,100
2	Retail	25,730	1,800 to 3,090	2,990
3	Commercial	24,960	1,750 to 3,000	2,900
4	Commercial	22,130	1,550 to 2,660	2,570
5	Commercial	18,960	1,330 to 2,270	2,200

Peak power at 0.07 to 0.12 kWp per m² of roof; MWh a year at the central 0.10 kWp per m² and 1,162 kWh per kWp (PVGIS). Upper bounds before structure, shading and grid checks.

Worked example 1: United Kingdom

Great Baddow, Chelmsford: every roof, from 1 m LiDAR

Our Rooftop Solar Screening model worked out the pitch and direction of every roof cell inside 1,910 building outlines from the Environment Agency's 1 m LiDAR, then gave each building its suitable roof area, indicative peak power and yearly generation.

22.7 m²

suitable roof on the typical (median) house, about 4.1 kWp

37%

of all suitable roof faces south (23,489 of 63,334 m²)

10,100 MWh

a year if every suitable square metre carried panels (an upper bound)

5.4 s

to screen all 1,910 buildings in the 1.2 km² test area

Roof suitability map of part of Great Baddow, Chelmsford: south-facing roof in orange, east or west facing in light yellow, flat roofs in blue, north-facing or unsuitable in grey — Roof faces by pitch and direction, part of the test area. Orange: south-facing. Light yellow: east or west. Blue: flat. Grey: north-facing, too steep or roof edge. Open full size

Map of Great Baddow with each of 1,910 buildings shaded by indicative solar generation per year, darker green for more — Indicative generation per building per year, whole test area. Open full size

Settings: 0.18 kWp per m² of suitable roof, 950 kWh per kWp a year, east and west roofs 17.5% lower; shading and roof structure not modelled in this run. Contains Environment Agency LIDAR Composite DSM 1 m, © Environment Agency copyright and/or database right 2022, OGL v3.0. Buildings © OpenStreetMap contributors (ODbL). Yield setting from PVGIS 5.3 (European Commission JRC). Model run 10 November 2026.

Worked example 2: Serbia

Novi Sad: an area screen from open data

Serbia has no free national LiDAR, so we screened from building outlines alone. In 28.5 km² of Novi Sad, 790 of 16,604 buildings have roofs of 1,000 m² or more, and 320 of those are non-residential, with 0.97 million m² of roof between them.

Map of central Novi Sad and the Danube with every roof of 1,000 square metres or more coloured by indicative peak power, from 100 kWp to over 1 MWp — Indicative peak power = footprint × 0.10 kWp per m² (screening factor). Buildings: Overture Maps Foundation, © OpenStreetMap contributors (ODbL), Microsoft ML Building Footprints (ODbL), downloaded 13 November 2026. Open full size

68 to 116 MWp

indicative peak power on the 320 non-residential roofs (0.07 to 0.12 kWp per m²)

79 to 135 GWh

a year at 1,162 kWh per kWp, before structure, shading and grid checks

1,155 to 1,184

kWh per kWp a year across Belgrade, Novi Sad, Subotica and Niš, against 920 in London

Yields: PVGIS 5.3 (European Commission JRC), SARAH3 radiation data, 14% system losses, 10° south-facing frame, queried 13 November 2026.

Who it is for

Solar installersSolar developersCommercial landlordsProperty portfoliosCouncilsHousing associationsSchools and universitiesEnergy consultantsInvestors

Tools we use

QGISGRASS r.sunPDALGeoPandasDuckDBPostGISpvlibPVGIS APIMapLibre

Data we use

Building outlines

UK: OpenStreetMap, Ordnance Survey where licensed

Serbia and Europe: Overture Maps (OpenStreetMap and Microsoft building footprints), national cadastre where licensed

Roof shape

UK: Environment Agency 1 m LiDAR (England), national LiDAR in Scotland and Wales

Serbia and Europe: National LiDAR where it exists and can be licensed, stereo satellite or aerial imagery, drone surveys

Solar yield

UK, Serbia and Europe: PVGIS 5.3 (European Commission Joint Research Centre), checked for every location

Questions

How accurate is a solar rooftop analysis?

Each stage says what it assumes. An area screen is an upper bound: it assumes the roof can carry panels and is not shaded. Roof-by-roof analysis measures pitch and direction from a surface model, and a site report adds the roof's real obstructions. No stage replaces an installer's structural survey or the grid operator's connection offer, and every report says so.

Do you need LiDAR to map rooftop solar potential?

Not for the first stage. An area screen needs only building outlines, which exist for most of Europe. England has free 1 m LiDAR for roof-by-roof work; elsewhere we use national surface models where they can be licensed, surface models made from stereo imagery, or a drone survey for single sites.

Which countries do you cover?

The UK and Europe, including Serbia, Bosnia and Herzegovina and the rest of the Western Balkans. PVGIS covers all of Europe and open building outlines cover most of it; we check coverage for your area before we start.

Can I run a rooftop solar screen myself?

Yes, for one area at a time: our Rooftop Solar Screening QGIS model does the roof-by-roof step from a LiDAR surface model and building outlines. It is the model behind the Great Baddow example on this page.

What does it cost?

Every job is priced for your project. We agree one fixed price in writing before any work starts, with no hidden fees.

Start with your area

Tell us the area, the buildings or the portfolio, and what you need to decide.

Email us Book a call Product sheet (PDF, 2 MB)

Or write to semir@bozmaps.com