Wash Plant Optimization
Powering Processing Improvements at Fording River Operations
Fording River Operations (FRO) has embraced innovation for years, automating processes to achieve operational efficiencies. And now, through RACE21™, they’re taking on the ‘Connect’ pillar of the program and developing digital systems that leverage existing sensors and instrumentation and incorporate advanced analytics, to help drive improvements in throughput and yield.
This next phase of technology advancements at FRO is targeting wash plant optimization—a key step in the steelmaking coal processing cycle—by creating an operator advisory tool that recommends ideal set points in the wash plant, based on an analysis of incoming material and historical data. By reacting sooner to changes in the material and adjusting set points accordingly, yield is improving which is expected to achieve sustainable EBITDA value over time.
In addition, insights provided by the advanced sensing and analytics have also unlocked additional value, by identifying improvements to the thickener process, allowing the FRO wash plant to process additional coal from Greenhills Operations, thereby optimizing its processing capacity.
“Working with the RACE21™ teams’ resources and expertise has been instrumental in getting us set up on Google Cloud so that we can harness the power of big data,” says Shane McColman, Senior Engineer, Process Supervisor at FRO. “Working with RACE21™, we’re much better able to accelerate how we assess, model and implement changes to our systems.”
“This is an exciting ‘next step’ for us and an area that can drive real results in our business.”
Big Data Drives New Gains at Highland Valley Copper
Mill optimization requires continuously evaluating and improving a number of subprocesses to ensure maximum productivity; ore is processed efficiently, and copper recovery is optimized. At Highland Valley Copper (HVC) Operations, big data and machine learning are opening up cutting-edge opportunities in this area, and with the backing of RACE21™, HVC recently took on two optimization projects with the biggest potential to deliver value.
Semi-autogenous (SAG) mill and bulk flotation optimization were selected as the processes to target, based on the readiness and quality of data to enable the application of advanced analytics. By working with the RACE21™ team and with data scientists from Boston Consulting Group (BCG), HVC developed a new, powerful tool that uses machine learning to deliver real-time recommendations to operators on optimal grinding and flotation operating conditions.
Early estimates project these improvements will create significant value by increasing copper throughput and recovery.
“We’re able to better use the data coming from the mine to determine optimal operating set points,” says Murray Cruickshank, Deployment Specialist, Technology and Innovation. “On top of that, we’re using Cloud-based technology to share these insights with operators, via a dashboard, so they can continuously evaluate and make real-time decisions.”
And a key aspect of this project has been working closely with operators throughout the process—keeping them informed of upcoming changes and receiving continuous feedback along the way.
“Having the support and input from operators—the people who use these tools everyday—has been critical to ensuring the testing goes smoothly,” adds Murray, “The operator buy-in to field-test the tool has been strong, which will help guide our next steps as we move to future phases and roll out the tool more broadly.”
Khushaal Popli, Specialist, Process Control, HVC, who has been working closely with data scientists from BCG, is encouraged by the early results and even more excited about how these investments can help pave the way for step changes in predictive maintenance—predicting and mitigating equipment failures, and thereby minimizing unexpected downtimes.
“We are at the leading edge of revolutionizing our processes so that we can work smarter, faster, and most importantly, safer,” says Khushaal.
Condition Based Monitoring
Advances in Data Science Boost Predictive Maintenance at Trail Operations
Predictive maintenance presents enormous opportunities for operations to achieve innovation-driven efficiencies, by using sensors and powerful tools to analyze data in real time so that equipment is used optimally. At Teck’s Trail Operations, where they’ve been using machine learning for several years, RACE21™ is providing the resources and expertise to accelerate work in their predictive maintenance program, allowing the team to react more quickly and reduce maintenance costs.
“Several years ago we started with basic condition-based monitoring; using data derived from sensors to establish trends with our equipment. Now, with the support of RACE21™ and McKinsey we’re able to apply advances in data science, namely advanced analytics, to make our predictive maintenance software even more powerful,” says Gordon Kavaloff, Senior Reliability Specialist, Trail Operations. “Now, we can detect a failure, have a work order planned and a fix ready, all in one go—this type of efficiency allows us to work in a whole new gear.”
“RACE21™ is empowering us to think differently about how we can enhance performance in all areas—from safety, environment and production. The opportunities are pretty exciting,” adds Gordon.