British researchers are using flying drones to help oil companies map subsea reservoirs more economically.
The drones, operated by geoscientists at Aberdeen University, do not study the seabed itself but instead gather information on rock formations in cliff faces.
This can then be used to create a database of rock structures all over the world that geologists can then deploy to generate models showing what specific subsea rock layers probably look like.
Similar work has previously been carried out using laser systems mounted on helicopters. But the researchers, working with colleagues at Uni Research in Bergen, Sweden, have found that equipping drones with 3D cameras dramatically cut the cost and time such studies take while still producing highly accurate data.
‘We’re revolutionising the way people do field geology,’ research leader Prof John Howell told The Engineer. ‘This is equally applicable to mining, to civil engineering, to mapping avalanche hazards.’
Oil companies can gain snapshots of the geological structure of the seabed at specific points by drilling boreholes.
But for the area between boreholes they either use seismic data – which only produces image resolution of between 10m and 50m – or build a statistical model of what the rock probably looks like based on the borehole data and information from comparable formations as seen in cliff faces.
The remotely piloted octocopter drones used to gather this information cost £10,000, whereas a helicopter costs around £5,000 an hour.
The downside is that the drones are limited to a payload of 1.5kg, meaning they can’t carry the LIDAR (the laser equivalent to radar) equipment previously used to create 3D models of the cliffs.
However, computer processing power is now sufficient that the researchers can recreate similar models using 1,000 optical photos taken at slightly different angles and positions and stitched together with GPS coordinate data to create a 3D image with cm-level resolution.
‘The majority of the work we did last year was shooting cliffs with both the drone and the LIDAR and then work out whether the errors between the systems were acceptable,’ said Howell. ‘We’re getting down to very small percentages of errors between the two systems.’
The research is part of the ongoing SAFARI project, which is sponsored by 24 oil and gas companies.