Problem Solving: A look at analyzing mining data to improve performance
Maximizing production, while increasing efficiency and safety, are paramount in today’s mining operations where companies are grappling with some of the toughest commodity pricing conditions in memory.
Consider that at the end of last year, the price of coal was down 32 per cent from the end of 2014; iron ore was down 24 per cent; palladium down 30 per cent; copper down 25 per cent; zinc down 30 per cent, and aluminum down 19 per cent, with only gold recovering slightly from the lows of 2015.
To help mitigate this price challenged environment, IBM is now putting the full force of its analytics capabilities behind the mining industry in an effort to help companies increase production and make fundamental shifts in the way they operate efficiently.
Working with a large, multi-national mining company, our company developed a new proof-of-concept that is able to analyze existing data and produce a novel set of data visualization reports that show where inefficiencies are occurring in a mine’s operations.
Combining data from mine dispatch systems, fueling, location and other available data, the solution provides a single, fact-based view of the orchestration in the mine, and uses an analytical model to identify where the inefficiencies are.
This information then allows mine personnel to determine possible causes and rectify them quickly.
In the proof-of-concept testing phase, the solution focused on a core mining process – hauling cycle times.
Typically a mine’s mobile fleets are continuously moving throughout the mine: queuing, loading, dumping, and hauling both in full and empty modes. And the hauling process is complex due to many factors, like weather conditions; regulations, other processes like blasting, the minerals being mined, human behaviour and inherent uncertainties like mechanical failures.
These complexities also lead to inefficiencies, some of which a mine’s dispatch system is designed to capture, but which a dispatch system alone is not suited to analyze for recurring patterns that are outside the norm, and pinpoint where a supervisor needs to go and investigate further.
In one Brazilian operation, for example, our software solution was able to visualize repetitive patterns in the truck/shovel hauling process that were previously undetected.
By analyzing many, many days of truck/shovel cycle times – both historic and real time – it was able to determine that an inordinately high number of trucks in the shovel que were being passed by other trucks prompting the question, “what is going on here?” The mine management team was then able to take these visual analytical results and determine the possible root cause.
In this particular case, mine staff discovered that older operators were letting younger drivers pass by and load up, thereby getting extra break time.
The solution not only identified a critical inefficiency in the mine’s operations, it was also able to help mine management put a dollar value on this particular truck/shovel inefficiency and make the appropriate adjustments.
Truck/shovel operations generally involve high costs. In surface mining, for example, truck haulage is the largest item in the operating budget, constituting about 50 to 60 per cent of the total mining costs. So, in a competitive market environment, coupled with the current price-challenged environment, closing the gap and maximizing potential truck/shovel capacity is of critical importance.
This Operational Effectiveness Analytics solution, developed by IBM’s Haifa Research Lab, was tested over a full year in an open-pit environment and can be customized for any type of mining operation or mineral.
Not only does it address operational excellence in the mining industry by identifying defects and minimizing process variabilities, as was the case at the Brazil mine mentioned above, the solution has broader implications for health and safety, as well as compliance in general. It can help verify proper evacuations before blasting, identify circumvention of safety equipment, verify pre-shift checks and training, and identify deviations from recommended speeds and routes.
In total, it allows a mine to be productive, efficient and safe; all key imperatives for the sustainable operations of today’s modern mines.
By mining the data, IBM is helping transform the mining industry – improving operations, democratizing key processes by creating transparency, and lowering costs while expanding margins – even in a downturn. C
Doug Hanson is Director at IBM Oilsands and Natural Resource Solution Centre, Calgary.
Comments