The control room at the Talvivaara nickel mine in Finland. (Photo courtesy of Metso).
IN CONCENTRATORS AND MINES ALIKE, CONTROL SYSTEMS ARE INCREASINGLY PROVING THEIR WORTH AS OPERATORS STRIVE TO STREAMLINE FUNCTIONS AND IMPROVE PRODUCTIVITY
By Simon Walker, European Editor
Two key facts often stand out in relation to today's industrial processes: they are becoming increasingly complex, and there are fewer people available who have the skills and dedication needed to get the best out of them. Little wonder, then, that the traditional rule-of-thumb approach to process operations is steadily being replaced by control systems that can be fine-tuned to optimize them in terms of a variety of parameters, either singly or in unison.
Technology has moved on dramatically since the first programmable logic control (PLC) systems became available in the early 1970s. As the 1985 edition of the SME Mineral Processing Handbook noted, "No facet of the art of mineral processing is changing faster than the use and application of automatic controls.
"The application of computers to control mineral recovery processes is being investigated in varying degrees by many companies," the handbook went on, before providing a few words of caution to plant operators who still had to take the plunge. "The decision to apply a computer to an ore processing operation can only be taken after a careful analysis of the potential economic benefits, plant location relative to equipment servicing, and the degree of enthusiasm expressed by the operating personnel.
"In most cases, the introduction will involve months of effort by the plant metallurgist and at least one computer technician. The expense will also be considerable."
Perhaps the one comment that is missing from this snapshot of the state of affairs some 30 years ago is this: in today's price, energy cost and skilled labor availability market, an operation can't afford not to.
As Martin Knabenhans, head of process control engineering at ABB, told E&MJ, the early PLC systems bear little resemblance to those available today. "There is a large range of automation levels, from the basic ability to monitor and interact with specific components (such as valves, pumps or motors) to complete process optimization systems, where the automation system takes over control of the process like an ‘autopilot,'" he said. "One can safely assume that a highly automated system will provide a 20% improvement in productivity, product quality and energy efficiency, compared to a purely manual operation."
As Knabenhans inferred there, there is a whole raft of componentry that can be applied in whole or in part to the industrial processes involved in mining and mineral processing—not forgetting, of course, that control systems are increasingly finding their way into the underground environment for applications such as ventilation and pumping. E&MJ asked a number of the leading suppliers of control and instrumentation systems for their views on current practices.
RAPID DEVELOPMENT HISTORY
The first question that E&MJ asked looked at the major advances that have taken place in plant control and instrumentation technology over the past 10 years, and the main technological achievements that have helped this to happen. "Ten years ago, control systems were not as open, integrated and object-orientated as today," Knabenhans explained. "Many solutions were either PLC + HMI (human-machine interface)—based with lots of effort for interface engineering between them, or proprietary—often closed—DC (distributed control) systems. Alarm and asset management, historian and advanced process control, nowadays seamlessly integrated parts of modern automation platforms, could only be implemented by adding separate systems running on dedicated hardware and requiring different engineering environments.
"The introduction of standardized wired and wireless, vendor-independent communication protocols connecting intelligent electrical and instrumentation devices to the automation platform has been another game changer," he added.
Chris Culph, electrical engineer at Gekko Systems, noted that user-friendly interfacing, distributed components and data storage have been pivotal in gathering the required information from around process plants. "Historically, instrumentation had a feeling of a ‘black art,' which often meant that devices were difficult to install, calibrate, and get connected into the control system," he pointed out. "Centralized systems also meant that required power and control cabling had to be run over large distances at great labor and capital expense. The time and cost expense to get more information has become the key factor as to whether or not that information was absolutely necessary.
"By suppliers making instruments easier to connect to, and providing some universal or commonality of communications, and distributing data capture locations (such as field-based PLC input and outputs), much of the cost has been reduced. This allows for further instrumentation to be installed," Culph said. He also noted that over the past 10 years, data storage has increasingly become cheaper, so more information can be retained, processed and developed to provide better clarity on how a plant is performing.
According to Bob Jonas, mining and metals market development manager at Honeywell Process Solutions, multivariable predictive control (MPC) has had a considerable impact on directly improving operations and delivering significant financial benefits. Honeywell now has more than 300 individual mineral processing units that are being optimized by its MPC, he said, with the technology being used in 20 different mineral processes—of which grinding and flotation have been the most successful.
"Another paradigm shift we've seen in the industry is the use of remote Web access to provide an increased level of collaboration. This has led to the enhanced support of automation systems, as well as enabling real-time monitoring of mining and processing equipment," Jonas said, adding, "remote support centers and virtual centers allow geographically distributed experts to act as a central support center.
"These help customers achieve a higher implementation rate and provide better support of automation technologies, from instrumentation to advanced performance monitoring."
Jonas also mentioned that mining companies are increasingly using active monitoring for the performance and health of mobile equipment like trucks and shovels. Honeywell's Mobile Equipment Monitoring (MEM) system captures and communicates data on fleet health and operational status, with operators using the system to define rules to alert to real or impending conditions, and to automate procedures for recommending remediation. "MEM users are seeing savings through the prevention of equipment failures (such as engine or transmission), as well as opportunities to improve operational safety," Jonas told E&MJ.
SEEING INFORMATION IN THE RIGHT WAY
Jonas also explained that the inclusion of human factors into the control system has been progressing in the mineral processing industry. For example, standards developed by the Abnormal Situation Management (ASM) consortium, of which Honeywell is a member, define how operator displays should be built, including the best use of colors and shapes to optimize control-room operator effectiveness. The ASM has also addressed the problem of more alarms being generated by today's intelligent instruments and controls.
Rockwell Automation's industry marketing lead for mining, minerals and cement, Andrew Bagley, noted that since 2003, the major technological shift in plants has seen proprietary networks being replaced by Ethernet as a common standard. This has allowed much more enterprise-level networking in plants, leading to better network management and security, he said. The integration of instrumentation networks into the control system has also been simplified. The ability to upload plant and configuration data directly into the controller has tremendously accelerated configuration and asset management as well as the ability to leverage smarter instrumentation for data validation and diagnostics.
The controllers in traditional PLC-type architectures have increased in performance that allows far more control, with less reliance on the HMI layer. This significantly improves software maintainability and the cost of ownership, Bagley explained. "In addition, more deterministic MPCs are replacing expert systems as the primary advanced control technology," he said. "MPCs identify process models to drive optimal operations and deliver higher value in a much more sustainable (easier to maintain) platform."
Janne Kytökari, product manager for mining industry applications at Metso, looked at the overall benefits associated with sophisticated control systems. "With accurately controlled mineral processes, such as the crushing and grinding circuit, flotation and water management, a mine can get a steady yield without any severe deviations. By using an automation system that is designed for plant-wide control, it is also easy to deploy advanced process controls (APCs) that help to solve more complex challenges like SAG-mill optimization. The benefits of correctly tuned control loops and APCs include stabilized production, energy savings and better quality.
"Machine wear is often hidden until a machine breaks down," Kytökari added. "With a single automation system, which has integrated machine-condition monitoring such as Metso DNA, it is easy to measure condition data such as vibrations from a machine and use that with other maintenance information in scheduling predictive maintenance. This can reduce process downtime significantly."
Turning to typical plant performance improvements that can be gained through the use of automated plant control systems, Bagley stated that the biggest of these is the long-term steady-state control that they bring. In any hydrometallurgical operation like flotation and milling, the ability to maintain a steady state control over a longer period of time is the biggest advantage, with quality variability being reduced by 30%–75% when using an MPC.
"Any advanced control system typically leverages the existing control system infrastructure, which makes it not only a wise investment but it can be implemented in a relatively short timeframe," Bagley added. "MPCs' other benefits over traditional solutions include increasing capacity between 2% and 10% depending on the process, and reducing energy costs between 4% and 10% per ton."
From Honeywell, Jonas agreed. "Typically 1%–6% throughput and 1%–4% recovery (yield) increases are achievable, depending on the processing methods and technologies deployed," he said.
Gekko's Culph looked at the benefits in a wider sense. "If employed correctly, the system can benefit safety, helping to protect personnel and equipment around dangerous activities. It can also benefit predictive maintenance, leading to reduced downtime; optimal resource efficiency—with lower operational costs; and increased productivity, allowing the plant to achieve greater throughput."
SO YOU'VE GOT THE SYSTEM...
Clearly there can be a significant investment involved in installing a control system in a concentrator, and the obvious response will be to make the best possible use of it. That does not always happen, however, and E&MJ asked each of the respondents about the most common mistakes that operators make in relation to the control and instrumentation systems that are installed in their plants.
For ABB, Knabenhans pointed out that mistakes often happen in the conceptual design phase of a plant. "Extensive use of vendor package PLC solutions that are integrated later into the overriding process automation platform makes life difficult for operation and maintenance," he said. "Those subsystem PLC's do not match the plant's common operation and alarming concepts, they also require different spare parts and provide limited diagnostic data for the underlying process. This causes problems when fault-finding.
"In many mining plants, operators still control individual devices such as motors and valves," Knabenhans said. "Other industries have long relied more on functional groups or even process area control, where an overriding group function controls bigger parts of the process. In combination with properly tuned loop controls and advanced process control solutions, both up-time and productivity can be improved."
From Siemens Process Automation, business development manager for mining, aggregates and cement, Mark Yseboodt, looked at hardware issues that the company has noticed. "One problem is where operators look at the cost of instruments rather than the lifecycle cost (cost of ownership)," he said. "In addition, EPCs sometimes select instruments for new projects based on cost, not performance.
"Being unaware of the latest developments in technology might lead to equipment being installed that is not the best, and once in operation, if the calibration and maintenance of, say, continuous weighing devices is not done properly, they won't provide accurate data inputs to the control system. Then again, another problem that we see is where operators try to rely on information from instruments that are not suitable for the task, for one reason or another."
According to Jan Schilling, global product manager for process control systems at FLSmidth, the natural tendency is for operators to look for simplified patterns of operation to help them operate more easily. "This level of ‘expertise' means monitoring fewer signals more often, expecting other signals to be linked in some way to the few," he said.
"Most often this technique works fine, but there may be some potential optimization being missed. And while monitoring fewer signals would seem to leave more time to keep an eye out for other problems, unfortunately sometimes it results in a general lack of focus, resulting in missed warnings and alarms."
Honeywell's Jonas told E&MJ that the company's experience shows that many well-intentioned efforts fail due to their stepwise implementation. "In today's business environment, a stepped or phased approach is often needed for funding and efficient allocation of resources," he said.
"Where I see companies failing is in their planning to deliver and sustain business value. Rather, plans are made to do the first phase, often to repair controls or add instruments, with the intention that additional phases of the project will happen and support will be there when needed.
"The fault in this type of planning is that benefits might take several phases and years to deliver, and the results are mixed due to a lack of proper support. I often hear from plant managers that they need to fix a piece of equipment or add an instrument before they can make any automation improvements. Such delays can represent millions of dollars per year in missed opportunities."
...NOW MAKE IT WORK RIGHT
The investment has been made, but the system is not providing the results that were expected. What should an operator do to get it working effectively, E&MJ asked?
"One key to continuous improvement is to ensure benchmarking is in place to understand the automation and equipment performance, so that one can prioritize the limited resources used to fix the problems," said Jonas. "Today, there are tools available that provide insight and assist in the fixes. I believe these tools are under-utilized in our industry."
"In general, even well-designed process control systems would benefit from additional high-level control," said Schilling. "It may not always perform better than human operators in the short term, but being programmed to monitor many more signals simultaneously and continuously will result in better long-term stability in the process, less wear on the machinery, and in the end fewer, shorter production stops.
"Another, quite obvious thing to do, is to improve alarm handling," he said. "The aim is to reduce the number of alarms and make the few important alarms sharper and impossible to miss. Combined with root-cause analysis of the most common reasons for downtime, it should be possible to provide the operators with better prewarnings to help focus on problems before they become critical."
Yseboodt (Siemens) offered the advice that end users should be more demanding on performance and the right choice of technology with their EPC/contractor. This does, of course, require companies to have a better understanding of the latest technologies, and to know what to use, where and why.
Bagley (Rockwell) agreed. "Operators should focus on increasing standardization, not only in the platform (hardware layer), but more importantly on the application layer of control system software," he said. "They should also be open to investigating new concepts and be a critical listener when learning about potential advantages. In order for these mistakes to be addressed, they should clearly express any concerns about introducing a new technology into their specific operation."
"Problems can occur when experienced operators, who know how to adjust a process if needed, are replaced," said Kytökari (Metso). "New operators might not have the needed experience, so APCs and other expert systems can prove to be valuable tools in that situation.
"Don't forget that APC, analyzer and expert-system parameters can deteriorate over time as well, so it is important to do regular tuning to keep the process performance at a high level," he pointed out. "And it can also happen that maintenance crews and operators don't get the right training when new systems or instruments are installed—meaning that the equipment may be neglected altogether or its full potential not fully realized."
AND WHAT OF THE FUTURE?
Schilling (FLSmidth): At the moment, it is not completely clear where the development in technology is headed, but most equipment manufacturers are focusing on building smarter instruments and machines. While half the market seems to be a little reluctant to go all-in on these sophisticated solutions, the other half happily embrace it, moving toward fully automated systems with very few dedicated operators.
Knabenhans (ABB): Instrumentation and automation systems will become commonplace. In addition, there will be ore property sensors that will inform the control system in real time, which ore is being processed. This will increase the ability to optimize performance by better exploiting the tradeoffs between quality, recovery and production.
Culph (Gekko): Global connections that allow remote monitoring is the short-term future for many processing plants. This ability allows experts from anywhere in the world to review a plant's performance, and to provide advice on how to get the best out of it. Same services can already perform this function, and the concept is developing rapidly.
Bagley (Rockwell): Process control systems will increasingly become the focus of mining companies trying to get more production from existing operations. We anticipate increasing connectivity between systems so that the control and instrumentation of concentrators will be connected via higher-level advanced control and predictive systems to provide smooth, plant-wide control. The centralization of control systems in regional centers will become the norm.
Kytökari (Metso): Future plant control systems will have so much computing power available that more and more complex process challenges can be solved in real time. Eventually this will lead to self-controlling systems where there is no need for the operator to do any parameter changes. However, this will still take some time, and the development of the control algorithms is not going to be easy.
Yseboodt (Siemens): We anticipate seeing more demand for accurate and reliable measuring in the field. Operators must know all the conditions in the process at every stage in order to automate to the highest level. The development of smart instrumentation for ease of use and performance will certainly continue.
Jonas (Honeywell): We see the use of newly emerged technologies as a competitive advantage for mineral processors. As an example, Honeywell's Lean Execution of Automation Projects (LEAP) utilizes cloud engineering, virtualization of control hardware platforms, and application driven systems with universal I/O. The result is faster and more flexible project deployment that reduces costs and risk.
Human factors and usability are core values, allowing mineral processors to do more with less effort and resources. Honeywell's Experion Orion console helps to improve operator effectiveness through touch panels, alarming through ambient lighting, and clusters of ultra hi-def monitors, while the Experion Collaboration Station assists decision-making using a display that brings together data from widely differing sources.