When Peter Joynt talks about data management practices in the mining industry, he likes to start with this example:
“Let’s say, you’ve just joined a company as the new resource geologist and have been asked to update the resource estimate. First, you'll want to find the most recent version of the resource model. So you start asking around, looking at handover notes from the previous geo, to see where you can find it.
“After some digging, you locate a zip file on the network drive, and unzip the project into a new directory. The directory has a ton of file types you have to click through, and even more zip files. You find a few block model file versions – but it’s hard to tell which one is most recent. There are also a few parameter files, though the ones related to estimation seem to be missing. And a whole bunch of other files that you really have no idea about.
“Eventually you find the geological model – it’s contained within a software's proprietary binary format. But unfortunately, you are not a competent user of that software, and the software you plan to use doesn’t support import of that file format directly. So you get the mine surveyor to open the project and export the files as DXFs. The files import okay to your new software, although with errors that will block you from directly working with the solids. You find the polygon strings that the solids have been constructed from, but these clearly contain later edits.
“Then you open up the block model files and run some reports. You can’t initially reproduce the grades and tonnages published in the resource report, though they do match grades and tonnages that were reported in a press release.”
For some of you reading, this situation will sound frustratingly familiar.
Peter, who shared the scenario above, is a Product Manager at Seequent. He worked jointly with Mike Stewart, a Technical Domain Expert at Seequent, to explore what it means for a mining company to have good geological data management practices.
Mike and Peter believe there is an interesting contrast in how raw data (like drilling data) and interpreted data (like resource estimates) is managed at mining companies.
“At the moment, the mining industry is really good at managing raw data,” Peter said. “There are well-established procedures for data collection and QA/QC of that data. But mining companies aren't valued on their raw data. They're valued on the interpretation that's applied to them.”
Management of that interpreted data is generally ad-hoc, without permission or version control.
“There's no good governance around it. It's very reliant on people and that means that there's quite a high risk associated with it,” Peter said. “Luckily, this is not a unique situation to the mining industry. We can look to parallel industries for solutions.”
Peter Joynt will be presenting findings on good geological data management practices at this year’s Seequent Lyceum 2018 events in Vancouver and Perth. You can view the full presentation below, and find out more about Lyceum 2018 here.