A 250-hectare ground-mounted solar farm with a co-located battery energy storage system (BESS) is surveyed most reliably by a single fixed-wing UAV sortie carrying both a LiDAR sensor and a metric camera, flown over a GNSS ground-control network tied to OSGB36 and Ordnance Datum Newlyn, with independent check points withheld to verify the result. On the Lincolnshire site described here that method delivered a 5 cm bare-earth DTM, a 3 cm orthomosaic, a classified LiDAR point cloud and a full topographic CAD set, verified at a vertical RMSE of ±3.1 cm and a horizontal RMSE of ±2.3 cm against a ±5 cm specification.
This guide is drawn from our 250 ha solar farm and BESS topographic survey in Lincolnshire and supports our aerial survey, aerial LiDAR and topographic survey services. It is written for developers, planning consultants, drainage and electrical engineers and EPC contractors who need to specify or check a survey of this scale.
What the survey has to support
At 250 ha a solar generating station exceeds the 50 MW threshold and is consented as a Nationally Significant Infrastructure Project under the Planning Act 2008, through a Development Consent Order. The survey is therefore not one deliverable but the common evidence base for several workstreams, each of which draws on it differently:
| Workstream | What it takes from the survey |
|---|---|
| DCO application (Planning Inspectorate) | Site-wide layout plans, land plans, orthomosaic base mapping |
| Environmental Impact Assessment | Terrain and surface models for landscape and visual impact, shadow flicker, noise contours; georeferenced viewpoint photography |
| Grid connection (DNO, 33 kV) | Cable route corridor levels, cross-sections and DTM for the Acceptance to Offer process |
| BESS compound design | High-resolution DTM and sections for platform levelling, ground-bearing assessment and fire-separation layout (BS EN IEC 62933) |
| Flood Risk Assessment | 5 cm DTM for surface-water flow paths and the 1-in-100-year plus climate-change envelope |
| Internal Drainage Board consent | Field-drain headings, culverts, outfalls and ditch inverts |
| Agricultural Land Classification | Orthomosaic base mapping and parcel context |
| Biodiversity Net Gain baseline | Hedgerow lengths, watercourse margins, grassland parcels and tree positions for the ecologist’s habitat assessment |
| Highways Section 278 | Access junction, visibility splay and carriageway levels; swept-path base |
| Public rights of way | Verified footpath alignments for temporary diversion applications |
Specifying the survey around these uses, rather than around a generic “topographic survey”, is what determines the resolution, the contour intervals, the sections and the formats below.
Why a fixed-wing platform with LiDAR and a camera together
Two things drive the platform choice at this scale. The first is efficiency: a fixed-wing UAV covers 250 ha across multiple field parcels in a single extended sortie, so the whole site is captured under one set of light and ground conditions and one control network, without separate mobilisations for imagery and LiDAR. The second is vegetation. Arable land is divided by hedgerows, tree lines and scrub, and photogrammetry alone models the top of that vegetation as if it were ground. For drainage design and earthworks quantities a false surface along every field boundary is not a small error; it is the wrong answer in exactly the places the drainage engineer cares about. Carrying a LiDAR sensor on the same payload records ground returns through the canopy gaps, so the bare-earth model along boundaries comes from LiDAR while the open arable ground takes its levels from the photogrammetric solution where the two agree. Our guide to drone LiDAR versus photogrammetry covers the decision in general terms; on a large solar site the answer is usually both.
Flight parameters
On the Lincolnshire site the aircraft flew at 120 m above ground level on parallel transects with 80 % forward overlap and 60 % sidelap for the photogrammetric component, giving a ground sample distance of 3 cm. The LiDAR component achieved a point density above 25 returns per square metre with first and last returns recorded, which is what allows ground and vegetation to be separated in classification. Operations were flown under a CAA Operational Authorisation permitting beyond-visual-line-of-sight flight across the multi-parcel site, with a ground safety crew in radio contact and coordination with the local air traffic control zone.
Those parameters are the consequence of the accuracy band and the deliverables, not the starting point. A brief that specifies the RICS band, the DTM resolution and the contour interval lets the surveyor choose height, overlap and density to meet them; a brief that specifies the aircraft does not. See how to brief a UAV topographic or corridor survey.
Ground control and independent check points
Absolute accuracy on a site this size comes from the control network, not from the aircraft’s own positioning. Before flying, 24 ground control points were established and observed by static GNSS with dual-frequency receivers, tied to the Ordnance Survey National GNSS Network and levelled to Ordnance Datum Newlyn, at a maximum spacing of 350 m so that every parcel was inside the control. Eight further points were observed as independent check points and deliberately withheld from both the photogrammetric adjustment and the LiDAR strip adjustment. Their residuals are the only unbiased statement of what the deliverables achieved, which is why they are reported with the data rather than the specification being quoted back as if it were a result. The reasoning behind that practice is set out in our guide to ground control points for drone mapping.
Processing
The LiDAR point cloud was classified by automated ground filtering followed by manual editing along hedgerows, watercourse banks and woodland edges, removing vegetation, overhead lines and structures to leave a clean ground class. The imagery was processed against the same control network by structure-from-motion and dense matching to produce a true-ortho corrected orthomosaic and a photogrammetric surface model. The two datasets were then fused: LiDAR elevations were used for the DTM in vegetated and boundary zones and photogrammetric values across open arable ground, with the overlap between the two providing a continuous internal check on the solution. Everything was delivered on OSGB36 with heights to Ordnance Datum Newlyn, the reference frame the design, drainage and grid-connection teams all work in; see our note on OSGB36, OSTN15 and OSGM15 for drone survey data.
Deliverables for a solar and BESS site
| Deliverable | Specification on this site | Who used it |
|---|---|---|
| Bare-earth DTM | 5 cm resolution, ASCII Grid and GeoTIFF, OSGB36 / ODN | Drainage engineer, FRA, structural engineer, earthworks contractor |
| Contours | 0.25 m for planning submission; 0.5 m for earthworks design | Planning, earthworks |
| Digital Surface Model | 5 cm, including hedgerows, trees, overhead-line sag and structures | Shadow flicker, line-of-sight, LVIA photomontage |
| Topographic CAD plan | 1:2500 site-wide, 1:500 for the BESS compound and substation; DWG and PDF | All design disciplines |
| Orthomosaic | 3 cm true-ortho GeoTIFF | EIA, LVIA, BNG mapping, ALC, land plans |
| Cross-sections | BESS compound, substation platform, DNO cable route, access road, watercourses; 1:500 H / 1:100 V | Electrical, civils, drainage |
| 3D site mesh | Georeferenced | Visualisation, earthworks simulation, programme planning |
| Cut-and-fill volumes | DTM against proposed formation levels | Contractor tendering, material balance |
| Classified LiDAR point cloud | LAS | Developer GIS; pile location and obstacle clearance |
The field-drain inverts, culvert positions and outfalls that the drainage consent depends on were captured by the ground crew, not from the air; no aerial method lifts a lid. What a complete pack should contain, and how to check it on receipt, is covered in what a drone survey deliverable pack contains.
Accuracy achieved, and what “within specification” means
The project specification required ±5 cm. Against the eight independent check points the delivered data achieved a vertical RMSE of ±3.1 cm and a horizontal RMSE of ±2.3 cm. In the language of the RICS Measured Surveys of Land, Buildings and Utilities, 3rd edition (RICS, 2014) band table that is the range a well-controlled UAV survey is specified in; it is not the millimetre accuracy of a terrestrial scan or total-station survey, and a solar site does not need it to be. Where a particular element does, for example setting-out control for the substation platform, the aerial survey is supplemented by ground observation to a tighter band rather than the whole site being surveyed to it. Our guide to how accurate a drone LiDAR survey is explains the bands and what governs them.
The combined LiDAR and photogrammetric DTM was measurably better in the hedgerow and boundary zones than photogrammetry alone would have been, because the false raised ground that vegetation produces in an image-only model was replaced by LiDAR ground returns. That is the practical case for carrying both sensors on a vegetated agricultural site.
Limitations to plan for
- Leaf-on hedgerows and woodland reduce LiDAR ground-return density; winter capture, slower flight lines or higher density recover more, but sparse zones are reported rather than smoothed over.
- Open water in ponds, ditches and the main river returns little usable data from either sensor and is flagged.
- Drain inverts, culvert dimensions and anything under a lid are ground-crew observations.
- Weather, wind and controlled airspace constrain when the sortie can be flown; a programme should carry contingency.
- The survey records the site as found on the survey date; subsequent land drainage works or cropping changes are outside it.
How to commission a survey of this kind
Send the site boundary as a digital polygon, the intended uses from the table above, the datum and grid required (normally OSGB36 / ODN), the accuracy band or tolerance, the DTM resolution and contour intervals, the formats your design team works in, any existing control or previous survey data, access arrangements and dates, and site constraints such as overhead lines and public rights of way. With those we can scope the capture, the control network and the deliverable pack, and quote a programme. Speak to a surveyor through our contact page.
Related reading
- 250 ha solar farm and BESS topographic survey, Lincolnshire — the project this guide is drawn from
- BNG baseline surveys — the habitat baseline delivered from the same capture
- Anglian Water reservoirs aerial survey — the same fixed-wing LiDAR method across 70 km² for twin DCO applications
- Weekly UAV earthworks monitoring — what the construction-stage surveys that followed look like
- Our UAV survey fleet — the WingtraRAY, WingtraOne GEN II and Zenmuse L3 platforms used on sites like this