Mining Company PV Invest based in Velenje, Slovenia, Uses Drone-Generated 3D Model to Calculate Volumes of Surface Subsidence.
Mining and Surface Subsidence
Several types of sub-surface mining, and specifically methods which intentionally cause the extracted void to collapse will result in surface subsidence. Mining-induced subsidence is relatively predictable in its magnitude, manifestation and extent, except where a sudden pillar or near-surface underground tunnel collapse occurs.
Mining-induced subsidence is nearly always very localized to the surface above the mined area, plus a margin around the outside.
Where mining activity is planned, mining-induced subsidence can be successfully managed. This is accomplished through a combination of careful mine planning, preventive measures, and the carrying out of repairs post-mining.
This project covers the area above ‘Pesje’ cave, right between the ‘Druzmirsko jezero’ and ‘Velenjsko jezero’ lakes. Intense underground mining has caused surface subsidence in several locations. To prevent water spillage from one lake to another the area needs frequent monitoring and repairs. Those are carried out by filling the damaged terrain with by-products from a nearby Sostanj Thermal Power Plant.
Drones in Mining and Site Rehabilitation
This case focuses on the drone-based methodology for producing precise geo-referenced data that can be used to measure the volumes of surface subsidence and therefore rehabilitate the area with the right amount of by-product from the power plant.
Flying the Site
PV Invest had no need for contracting a local drone services provider to fly the site as their team have been trained to be able to pilot drones themselves. They flew a DJI Phantom 4 and used a mobile app to plan and automate the flight. Due to the size of the site (44.5ha), multiple flights were needed. Total flight setup and execution took about 2 hours.
Because the map would be used for analysis and volume calculations it was important for the map/model to be accurately orientated in space. To ensure a high degree of accuracy, PV Invest used 16 ground control points (GCPs).
After flying the site, PV Invest uploaded the captured imagery and GCP coordinates to 3Dsurvey for point cloud and orthophoto generation.
Results - Calculating Volumes of Surface Subsidence
Upon generating 3D point cloud and calculating an orthophoto in 3Dsurvey - acting as the basis for corrections and re-alignments - a local (country-specific) database was used for sourcing original documentation and pairing (overlaying) it with the digital orthophoto map to produce a topographic map.
From there on, digital elevation maps and contours were calculated to provide a reference for post-repair site investigation and comparison.
Automatic volume calculation feature in 3Dsurvey empowered PV Invest engineers to get all the numbers on cuts, fills, surface changes, and subsidence volumes in minutes.
With achieved accuracy of 4 cm, ordering and transporting the right amount of by-product material for carrying out the repairs was a simply a matter of looking at final volume calculation numbers and - action.
Faster Data Collection Leads to Better ROI
The whole project, from flying the Phantom to finished documentation, took 2 days, nearly 5 times faster than classical methods usually take, with the benefit of highly improved data quality and quantity.
More Surveying Projects w/ DJI Phantom 4 and 3Dsurvey
DJI Phantom 4 Geospatial Accuracy Tested >
Mapping Cadastral Boundaries w/ Phantom 4 >
No longer is manual fieldwork and classical measuring method the most efficient and accurate way of data acquisition. Speed, data quality and quantity, ability to capture information elegantly with no need to revisit the site, all of these stand as enormous benefits to any surveying operation.
Project Tools & Specs
Drone & Software: DJI Phantom 4 + 3Dsurvey
Flight details: 3x 20-min flights at 100m AGL
Camera: Integrated DJI w/ embedded rolling shutter
Positioning & Orientation: Integrated GPS (EXIF data), 16x GCPs measured w/ Leica GPS Viva GS10
Photogrammetric processing & Outputs: 3Dsurvey
Number of Images: 1378
Image overlap: 80%
Area size: 44.5 ha