The control room at Xstrata Nickel’s Nickel Rim operation
Poor little Johnny; he had excitedly assembled his first bicycle all by himself only to discover that it just won’t go, even though it looks fine. Then again, the instructions were not exactly crystal clear. Frustrated, he declares his shiny new bike a lemon and asks his parents to replace it. After a little investigation, Johnny’s father found that his son had made a few mistakes while assembling the kit — the brakes were not tight enough and it was impossible to switch gears. When fixed by a competent shop, the lemon was transformed into the dream bicycle his son had dreamed of.
But what does this have to do with process control in mining, you might rightfully ask? However fanciful or seemingly simplistic Johnny’s scenario might sound, it is not dissimilar to the frustrations that process control personnel may encounter at a mill. An technician might spend his days wrestling with a stubborn PLC (programmable logic controller), trying to stabilize the grinding circuit to stop the level of flotation cells from oscillating, only to discover that nothing works. As in the case of Johnny, he may wrongfully think that he has a lemon or that the process is too complex to be handled by the PLC using PID (proportional integral derivative) controllers, and an expert system is required.
Asking the right questions
In such process control situations several questions need to be addressed:
- What kind of instrumentation is in place? Is the right technology always used? Do the installations always follow best practices?
- Are the signals used in the PLC, for process control purposes, properly sampled and filtered? Are the PID controller configurations consistent?
- Are the control strategies appropriate? Do the controllers require sporadic manual adjustments by the operator? Are the process constraints and interactions properly handled in the PLC? What is the method used for PID tuning? Do existing control strategies take advantage of cascade, feed forward and multivariable configurations?
In other words, the root causes of the problems need to be examined. One needs to determine whether the problem is related to instrumentation, control systems or process control. Where to start investigating and who to consult are also important considerations.
There is no such thing as an omniscient expert. Instrumentation, control systems and process control are closely interrelated, and expertise in any one of these fields will imply some fundamental knowledge in the others. However, proper design and efficient troubleshooting generally requires more than basic knowledge — this means it will often take more than one expert to tackle the problem. Where does one start? Here are some tips.
These problems are so common that it is estimated that 30 per cent of control valves exhibit mechanical problems and that 15 per cent of instruments and control systems are not installed properly. To avoid these problems, one should look for multiple references — suppliers, users, consultants, integrators, etc. — with whom to develop a network and share your own experience. It is also useful to carry out trials with new technologies. You can always team up with manufacturers and suppliers to share the risk and even develop your own in-house expertise. Although it seems obvious, it is worth stressing that you must always follow the instrument manufacturer’s installation specifications. Finally, it is highly recommended that you work with recognized integrators — they spend 90 per cent of their time solving issues that you encounter 10 per cent of your time. They could save you a lot of trouble.