Modeling systems, connectivity and shared expertise drive costs down
Steve Fiscor, Editor-in-Chief
The primary crusher and conveyor on Cerro Verde’s dry side demonstrates the scale of mega projects.
The scale of some of the modern mineral processing plants is immense and the time frame from concept to commissioning has been compressed to as little as four or five years. The engineers that design and build these projects rely on experience and skill sets amplified by modeling and project management software that automates repetitve tasks, which shortens the timeline. Increasingly, engineering, procurement and construction management (EPCM) firms are reporting that they are delivering projects on time and sometimes under budget.
Some of that success could be attributed to fewer projects in the development pipeline, allowing more experts to focus on fewer projects. After all, the mining industry has switched gears more recently from project development to optimization of existing operations. Here too, technology is playing an increasingly important role. Buzz words, such as Big Data and the Internet of Things (IoT) are bandied about regularly. While mining executives talk about improving efficiencies and lowering production costs, are all the mines using all the tools to their full advantage?
For sure, the connected world has forever changed how engineers solve problems. They literally have access to all the information they need in the palm of their hands. The sensors and devices that provide machine intelligence are becoming smaller and cheaper, allowing them to be placed in more and more locations. Some of the more progressive concepts, such as advanced pattern recognition and machine learning, will only improve diagnostics for preventive maintenance.
The mills and concentrators around the world rely on a finite group of professionals. While this collective pool of knowledge is not growing, connectivity allows more plants to access the knowledge base or, vice versa, a talented metallurgist could oversee or consult on several operations from a central location. Until recently, the engineering and metallurgical expertise was expected to be available on site. With remote access and the ability to more effectively monitor everything inside the plant from the performance of a flotation cell to the temperature on a pump, these professionals will no longer be confined to remote outposts.
Bringing a Mega Mineral Processing Project Online
One of the largest recently commissioned mineral processing projects was the expansion at Freeport McMoRan’s Cerro Verde copper mine in Arequipa, Peru. It is not only considered a world-class operation, but Fluor refers to it as the largest one-time built copper concentrator on the planet. They designed the original 120,000-metric-ton-per-day (mt/d) copper/molybdenum concentrator in 2007. At the time, it was the first concentrator to use high pressure grinding rolls (HPGRs) for tertiary comminution.
When it was time to expand, Freeport turned to Fluor again. The plan was to increase capacity by 240,000 mt/d to 360,000 mt/d. The expansion would allow Cerro Verde to triple production at the mine, explained Brad Matthews, Fluor project director for the Cero Verde expansion. “Fluor’s scope on the project was to provide engineering services for the concentrator with responsibility for procurement activities,” Matthews said. “All equipment and materials were specified by Fluor, including contract administration and commissioning.”
The expansion design was based on the same conventional copper sulfide flotation process, but included larger process equipment at a separate location on the mine property. It has a dry side (comminution) and a wet side, which consists of six stages of single stage ball milling and six rows of rougher flotation that produce a copper and moly concentrate, which is reground before it is run through cleaner circuits to produce a final copper-moly concentrate that is exported to Japan.
Fluor initiated a feasibility study, engineering and procurement services from its Vancouver, Canada, office in May 2010. The company’s office in Lima, Peru, supported the project’s construction management operations. “More than 50% of the workers on this project were recruited from Arequipa,” Matthews said. “We trained and coached these people, and the project was completed with an impressive safety record.” In fact, Cerro Verde received a President’s Award for its safety achievement.
Fluor leveraged its project knowledge and Peruvian project execution experience gained from the original Cerro Verde concentrator project, which is considered one of the industry’s most cost-effective concentrator plants. A phased startup of the new concentrator commenced on September 1, 2015, on schedule and under budget. The startup of the last of six grinding circuits commenced ahead of schedule in November 2015.
“The most impressive aspect is that Fluor executed a mega project on time and under budget within a five-year time frame,” Matthews said. First ore was pulled in late 2015, followed by project mechanical completion in March 2016.
Automation and the Connected Processing Plant
Once a plant achieves commercial production, the engineering aspect of the project converts to a program of maintaining or sometimes optimizing operations. The plant will likely recover metal or produce concentrate for 10 to 20 years, and conditions such as grade and metal prices will probably change during that period. However, today’s plant engineer and manager has more tools available to solve the problems that will surface.
Maintenance strategies, for example, have evolved from basic maintenance to preventive, planned and condition-based maintenance. This has changed the way plant equipment is operated and maintained from one extreme (run-to-failure) to the current regime of predictable breakdowns.
A key part of a maintenance strategy is asset management, a systematic process for maintaining, upgrading and operating physical assets, explained Dave Almond, general manager of FLSmidth Automation, Americas. While there is no single, unique definition of an asset management system, he said, there is a general agreement that it is a framework for measuring the health and performance of physical assets to identify potential problems before they escalate, allowing short-, medium- and long-range planning.
“Automation and technological advances provide the opportunity to develop and implement cost-effective asset management solutions and strategies that increase effectiveness and reduce operational cost,” Almond said. “In today’s competitive environment, a properly implemented and applied asset management system can minimize operational and maintenance costs by predicting the failure of a critical part and facilitate prevention, thereby minimizing the risk of breakdowns—and it can at the same time, maximize availability and utilization by increasing the time between failures and decreasing the time for repair.”
Almond believes that nearly every asset in a processing plant can be categorized into two major classes: production and automation. Asset management for production assets focuses on monitoring heavy machinery, electrical equipment and motors. Asset management for automation assets focuses on field measurement devices or sensors, the networks that connect these devices, and process analyzers.
“When it comes to productivity, a plant engineer wants to make sure they are getting the most out of production machinery,” Almond said. “They should identify key pieces of machinery that must remain healthy from a mechanical and processing standpoint. Then there is the automation aspect, which needs to be maintained as well. It’s more than just connectivity. It’s also ensuring the devices that provide information for the automation systems, such as sensors and instruments remain calibrated and function properly.”
Plant information management systems allow the integration of process data, business data and personnel data, which are the foundation to measuring plant performance. Real-time information management systems gather and historically record data from all the different sources, and include the interfaces to connect with other systems. Tools to convert data into information are required to calculate the key performance indicators (KPIs) by applying business rules.
A well-implemented asset management system checklist includes:
- Preventive, predictive and condition-based maintenance;
- Automatic notifications (process, alarms, events, etc.);
- Advanced diagnostics (instrumentation, devices, actuators, etc.);
- Downtime reporting and tracking;
- Performance monitoring (KPIs) and web visualization;
- Integration with computerized maintenance management system (CMMS); and
- Asset/object information.
“Progressive asset management systems enable the performance monitoring system to work together with the asset management solution to provide a framework to measure the performance and health of the plant assets,” Almond said. “This performance measurement includes not only individual machines and process areas, but also the process control performance, including advanced process control systems.”
Establishing such systems will create greater transparency on operations performance and identify areas for improvement, Almond explained. Integrated operating systems should also free people and resources to focus on operational excellence and productivity.
A team in the U.S. monitors daily operation at a nickel concentrator in South Africa.
Enhancing Process Performance
Process control systems can be integrated with maintenance systems, enabling access to real-time data for assets, which is key for condition-based maintenance operations. Plant engineers can combine the original equipment manufacturers’ (OEM) data on how the equipment should perform with knowledge about the actual wear rates to schedule maintenance. Understanding the risk of continuing operations, embedded sensors can provide better intelligence on actual wear rates allowing some repair jobs to be postponed to the next relining period.
“The fusion with more and more additional IoT data streams, from process, maintenance and wear parts, enables us to estimate ‘remaining useful life’ of the machines’ subsystems and components more accurately—and thereby improve operation and maintenance practices,” said Steen Christian Knudsen, technical manager, R&D, for FLSmidth.
“With sensors becoming smaller and cheaper, we can apply sensors in more places today,” Knudsen said. “The rapid development in information technology and network topology allows the machines to send more signals, more frequently. Today, we are getting more high-quality information at lower costs.” The experts at FLSmidth see interesting potential for using intelligent collaborative environments employing automation and sensor technologies to exploit opportunities that increase the plant’s operational effectiveness, and reduce its operational costs.
Remote systems also allow mining companies to measure how the machines are performing not just relative to the machines’ history at one mine site but relative to the performance of an entire group of plants. “Better diagnostics and trouble-shooting by fault symptoms signature, severity identification through data classification, and pattern recognition based on neural networks all enhance process performance, and provide better access to engineering services, and proactive maintenance services,” Knudsen said.
Engineers have studied neural networks for years, but the speed with which modern computer systems can employ algorithms to enhance pattern recognition and machine learning is constantly increasing. Microsoft’s Machine Learning is the most common neural network. The system’s neural network algorithm, according to Microsoft, tests each possible state of the input attribute against each possible state of the predictable attribute, and calculates probabilities for each combination based on the training data. Engineers can use these probabilities for both classification or regression tasks, to predict an outcome based on some input attributes.
Simplifying the explanation, Knudsen said the algorithm detects a pattern in the data. “The computer remembers that the last time it saw this data, this event occurred,” Knudsen said. “The key is the human interface with the data. The algorithms are providing more useful interpretations at a much faster pace.”
Intel believes the scientific community is just beginning to understand the potential of machine learning. Data scientists, developers and researchers are using machine learning to gain insights previously out of reach. The company said that engineers can now scale machine learning and deep learning applications quickly—and gain insights more efficiently—with the existing hardware infrastructure.
The principles behind these concepts are not new, Knudsen explained. “Now, however, we see the development and the possibility of data storage and faster computers,” Knudsen said. “This allows engineers to connect the dots and make predictions faster and automate the process. With the connectivity, everything is moving very fast now and we are limited only by our imagination.”
Harnessing the Intelligence
Tomorrow’s exploitable deposits will be found at deeper depths or in remote locations often with extreme conditions. With a shortage of experienced staff, and high turnover rates, mining companies will need to rely on new techniques to operate and maintain remote processing facilities. FLSmidth currently offers remote access systems that permits tasks traditionally performed at the plant location to be carried out at any location with a network connection to the plant.
As an example, FLSmidth cited a copper concentrator located in the Democratic Republic of Congo that had installed an Advanced Instrumentation and Process Control System for its grinding and flotation circuits. A remote access server provides secure internet access, allowing FLSmidth engineers to access data and to execute programming and process strategy adjustments from support offices located in Denmark and in the U.S. The mining company’s technical staff consists of expatriates and the location results in high turnover. With the new system, the operations no longer needs local level support for the control systems and they benefit from the use of advanced systems using less qualified personnel to maintain them. Furthermore, remote engineering eliminates the need to fly in contractors to do this work.
“So far, the mining companies that have embraced this opportunity have done so by circumstance,” Almond said. “They have been forced to be a little more innovative because they do not have the same access to skills. With travel headaches and restrictions, such as work visas and other paperwork, they simply cannot get people onto the site as fast.
In another case, a nickel concentrator in South Africa needed a comminution circuit comprised of a primary autogenous mill, pebble crushing, and a hybrid pebble/ball secondary mill. But the complicated circuit was difficult to stabilize and operate at high efficiency using manual operation, so the mining company turned to FLSmidth for Advanced Process Control and Remote Support to automatically operate the grinding circuit. To ensure a high degree of technical support to maintain this system, FLSmidth installed a remote access server for secure internet access and a team in the U.S. monitors the daily operation, adjusts the maintenance plans and optimizes the processes. In addition to a 4.3% production increase and a 6.7% decrease in circuit energy consumption, the setup has reduced instability across mills, pumps and classification circuit and resulted in a more stable operation with a more efficient pump operation. Remote connectivity enabled high-quality technical support without incurring the costs of regular site inspections and maintenance visits.
“The operations located relatively close to high tech centers have been a bit resistant because they believe they can get the support they need from traditional means,” Almond said. “Some have also voiced concerns about secure Internet connections, viruses, hacking, etc.”
FLSmidth has developed ways to communicate with these systems that are much more acceptable to plants as far as granting secured access to systems over the internet. “We have developed ways of managing the access that are traceable, as far as who entered the system, what they did and what time they did it,” Almond said. “People can be held accountable.”
The systems are also more reliable and robust than they used to be, and they are also more technically complex. “In the past, the plant had to have someone onsite who was properly trained and could tune the systems for changing strategies,” Almond said. “It’s not easy for every plant to have a person like that on site with today’s shortage of skills. Those people are not abundant. Using the internet, multiple plants can have access to expertise from a central location.”
“The rapid collection of real-time data is not worth much if it is not used real time—and above all you have to know where you want to go,” he said. “Operators often concentrate on improving one or two variables, such as reducing costs, lowering capital intensity or increasing throughput. However, a holistic focus on drivers of productivity that is shared at multiple levels is what is needed. With all the new technologies, the winners in the industry will be those who integrate their operations and rationalize their infrastructure.
“A lot of mining executives are talking about productivity improvements,” Almond said. “Unfortunately, that process is not moving at the pace with which it could move.” Looking toward the future, Almond sees automation specifically as an area where mining companies the greatest gains associated with optimization.
Detailed Engineering for the Tanami Expansion
For an expansion project to upgrade plant capacity at Newmont’s Tanami gold mine, located in the Tanami desert in Australia’s Northern Territory, Tetra Tech Proteus accepted the challenge of installing additional equipment in a constrained footprint. Basically, the scope of the work included the addition of a new ball mill, two new gravity concentration circuits, a pre-leach thickener, a new electrowinning cell and a deslime and tailings filtration system.
Ultimately, the objective of the project was to improve the processing plant’s throughput by 300,000 metric tons per year (mt/y) from 2.3 million mt/y to 2.6 million mt/y.
Using the OpenPlant PID software developed by Bentley Systems, Tetra Tech generated all the piping and instrumentation diagrams (P&IDs) required for the detailed engineering phase of the pro-
ject. Multiple design changes occurred throughout the life of the project and the system provided an efficient means to retain and modify the information attached to the P&IDs.
Details for the existing 20-year-old plant were provided by Newmont in terms of previous 3-D models generated via multiple software platforms and a detailed point cloud of the current plant. Tetra Tech used Bentley tools to convert this into a comprehensive and detailed 3-D model of the plant. Piping was modeled in OpenPlant Modeler and structure and concrete were modeled in AECOSim. Using saved/cached views from these packages allowed Tetra Tech to automatically generate annotated plans and elevations of the plant for the required client deliverables.
An Isometrics Manager platform was used to auto-generate close to “no-touch” piping isometric drawings directly from the 3-D piping model. The intelligence from this was used to automatically generate Material Take-Offs (MTOs) for the piping, concrete and structural steel. Using these tools, Tetra Tech brought an old plant back to life within a smaller footprint than would have otherwise been feasible.
These various design and modeling platforms allowed the engineers to share pro-
ject information and data between the engineering office and the field, and through various phases of the project. Engineers with Newmont and Tetra Tech could review designs, visualizing potential operational and maintenance issues, as well as clash detection.
The OpenPlant PID significantly reduced drafting time as altering components or splitting drawings did not require components to be re-tagged. As the design information changed, the engineering data could be batch uploaded into the P&IDs directly and then the drawings were re-synched with the database. This reduced the P&ID drafting hours by a few hours per drawing.
Using the Isometrics Manager, Tetra Tech reduced its isometric generation time from 3 hours to 1 hour per isometric. With more than 1,500 isometrics, this saved Newmont about $250,000.
Moreover, the technology also improved safety. The 3-D model was used for HAZOPs and design review sessions. The accuracy of the model enabled clear discussion on access and egress methods, maintenance methods, crane access and crane lifting studies within a tight footprint.