Energizer Resources Inc., has started a major flagship project, a large graphite mining project at coast of East Africa, in Madagascar. it is named the Molo Graphite Project, or simply ‘the Molo’. Part of the 1,000 km2 Green Giant Graphite Project in the arid south of the country, the Molo deposit itself spans 5 km2 and is thought to contain between 80 and 120 million tonnes of high-quality, all-flake graphite — used in refractories, batteries and consumer electronics — making it one of the largest such deposits in the world.
The need for full land survey to create a 3D contour map of the site for preliminary economic assessment is a big challenge for the Energizer Resources Inc.. The team would then use this to determine the location of the mine’s dam — since such mines are highly dependent on water — plus the location of its accompanying pipeline, as well as the positions of plant assets such as buildings and equipment.
Energizer Resources contracted a mining engineering company, DRA, to perform a full Bankable Feasibility Study.
DRA has two options for full land survey either they go for flying a large-scale LiDAR survey using manned aircraft or to fly eBee drones.
Using UAV LiDAR survey would have been very expensive — involving the importing of aircraft from South Africa — running to hundreds of thousands of dollars.
The 150 km2 to be surveyed comprised several parcels. Three large, square areas were flown to ascertain the best location for the site’s dam. An additional corridor covered the route of the favored dam site’s proposed pipeline. Plus, three further parcels were flown to the west of the site: the mine site itself; the plant site; and a rectangular parcel, the possible route of a road heading to the nearby town of Fotodrevo.
Eric Steffler, the geomatics manager at Energizer Resources Said:
“We chose to employ drones for two reasons,” Steffler says: “One was the price. This was very low compared to a LiDAR survey, which would have cost hundreds of thousands of dollars, even with our reduced survey area. The second was that by making a capital investment in two eBee drones, we could then fly these over our other properties in Canada. Since we now have the technology we need to produce digital elevation models and air photos ourselves, we don’t have to hire third-party companies to do this for us.”
In terms of creating the required outputs, Steffler and his team would first create a digital terrain model (DTM), in grid format. This would then be used to generate 3D contour maps of the site (eight sets in total), featuring half-metre and two metre contours.
The team also wanted to utilise the individual RGB shots its UAVs acquired. “Air photo interpretation was required to determine the locations of villages, roads, crops, and culturally significant items such as tombs and historic trees,” Steffler says. “We also needed to examine the quality of the roads that we would use for transporting material to and from the site, and our RGB shots would also form part of our environmental assessment.”
Steffler calculated that for an area of 150 sq. km and with an average flight time of 35 minutes, the UAVs would need to complete total coverage over 300 flights, and capturing some 150 RGB images per flight (15,000 photos in total).
“For our flight planning, we set a ground resolution of 9.9 cm per pixel in the drone’s eMotion software,” Steffler explains. “We chose this figure to ensure that, after processing the images in Postflight, we could still achieve the 20 cm product accuracy that DRA required. This GSD meant eMotion set a flight altitude of 292 metres above the eBee’s take-off point. We also used eMotion’s multi-drone function to manage flying our two UAVs in the same region at the same time.”
For georeferencing the data Steffler’s team used their GPS system to locate and post-process 70 ground control points across the project area. These GCPs were made from plywood, painted white, with crosses marked in black duct tape.
“With 10 to 12 flights flown each day we had 16 to 18 gigabytes of raw data to process every night,” he says. “We needed to buy a more powerful computer to deal with this data. Then once the entire survey had been flown, we had to merge the data together. For this we used open-source SAGA GIS software. This took us over two months.”
The final data products Eric and his team produced achieved a GSD of 9.75 cm, and accuracy within DRA’s requested +/- 20 cm range. “Our mean re-projection error was 0.179 pixels,” Steffler adds.
Having successfully completed the Molo’s UAV survey, and having used its outputs to inform DRA’s feasibility study, Steffler is now effusive in his praise of drone technology. “The return on our investment has been amazing. We saved ourselves hundreds of thousands of dollars by using the drones in place of airborne Lidar, plus we have sub-contracted our UAVs and personnel on several other jobs,” Steffler says.