The manning reductions seen in mining over the last few years have meant engineers are no longer available to perform many of these maintenance tasks, giving rise to the outsourcing of this all-important function.
According to Tony Healy, principal consultant with maintenance contractor, CCI Pope, at mine site level the biggest impediment to realising the full potential of tools like condition monitoring is the lack of a systematic approach.
“When this work was done in-house engineers looked for opportunities and they tied together all the various strands that were investigated along the way. As the techniques are contracted out they become compartmentalised,” Healy said.
“They’re now getting what they order rather than what they wanted.”
Broadly speaking condition monitoring is one of three basic approaches to equipment maintenance. In the past Run to Failure (RTF) was the ‘default’ for equipment, meaning little or nothing was done to prevent or pre-empt failure. RTF gives rise to high equipment downtime but low maintenance costs; business risk is high.
Preventive maintenance, a statistically based approach, is commonly used today. Maintenance costs are high, downtime is reduced but risk to the business is low. The method is not always applicable as it can actually increase downtime and generate over-maintained equipment.
Condition based repair/replace methods are a third option. Condition monitoring encompasses three basic phases: detection, diagnosis, and prognosis.
A measurable parameter of equipment condition is identified, which is routinely monitored and trended over time. Like preventive maintenance, downtime is reduced and risk to the business is low, while spend on maintenance is lower.
The primary advantage of condition monitoring is that it provides an early warning of developing faults; opportunity to plan repair work; reduction in consequential damage; and reduced risk to workers. The cost of setting up such a program and the high skill levels required are two disadvantages. Also, a suitable measurement parameter is not always available.
CCI Pope design and carry out maintenance programs for equipment in many diverse industries, including for large mining houses such as BHP Billiton and Anglo Coal.
There are some interesting challenges at the interface between maintenance philosophies at a mine and the introduction of a systematic approach, Healy has found.
He said many maintenance programs are driven from the bottom by focus on breakdowns or by what site engineers perceive to be problem areas.
“For instance, a mechanic says we should be doing oil analysis on this gearbox. It’s probably a good call and it’s not necessarily wrong but there’s no mechanism to say it’s justified.”
Another thing that typically happens is a ratcheting up of maintenance programs as people include more equipment and increase the monitoring frequency. This increases the spatial scope of the program, (more machines), and increases temporal detail, (more measurements in a given time period.)
“So, as the program grows there’s no discipline to bring it back to its root values. Because equipment is being over-maintained the cost is not noticed because breakdowns rarely occur,” Healy said.
CCI Pope’s methodology is to clearly link any maintenance program to business goals.
“With commercial businesses such as mines you can look at what the business aims are, examine underlying hazards and associated risks. Everything from that point on - maintenance activities, safety programs, condition monitoring programs - only exist to manage the risks you’ve identified,” Healy said.
“There has to be an auditable trail between an activity, say a vibration monitoring reading, back to some business goal. This trail is not always immediately obvious but should exist.”
The approach CCI Pope takes begins with a detailed analysis of a productive unit, say a wash-plant, which is broken down into assets that provide a major function. These are broken down further into sub-units as the process drills down from the machine unit to its component parts.
The likely failure mode is identified and meantime between failure (MTBF) estimated.
The process incorporates a thorough risk assessment related to a particular failure mode; could the failure of this piece of equipment shut down the business?
This risk ranking allows allocation of resources to maintenance to be made against clear parameters. Without this kind of process items that present the most ‘nuisance value’ are the ones that get the most attention, Healy said.
A by-product of a good system is it provides information to make continuous improvement.
The mining industry is most familiar with condition monitoring techniques such as vibration analysis, wear debris analysis, thermography and others. Already used widely in manufacturing and process industries, mining is beginning to adopt process monitoring, according to Healy.
This is a measure of how a machine is running when it is running under optimal conditions.
The use of Distributed Control Systems (DCS) and tools like 'smart' valves in modern wash-plants means that much of the necessary hardware for process monitoring is already in place as part of the DCS. Interrogating the system for maintenance purposes can sometimes be done with little cost or inconvenience.
A simple example often used in process plants is to trend the drive frequency against the flow 'set point' of a pump set and use the increasing speed as an indicator of 'wet component' wear. By establishing threshold values it is possible to trigger replacement activities for the pump set based on monitored parameters that indicate the part condition.
Healy said process monitoring as a Condition Monitoring tool has considerable promise. This is because the process measures tend to be 'primary' measures in that they indicate the condition of the asset function.
“This is after all what we ultimately wish to control,” he said.