Condition-based Practices Allow Transition to Predictive Maintenance

Best practices are taking maintenance to new heights

Up to 90% of all machines inside mines are either unmonitored or only monitored on an intermittent basis. So, it´s no surprise that one of the most common sources of unplanned downtime is machine failure. Of course, not every machine failure leads to catastrophic damage, and monitoring for early detection of damage and defects does not necessarily make good economic sense for every machine or system.

Previously, the costs associated with implementing a digital maintenance strategy may have been a constraint for plant operators. However, a smart, data-driven approach to maintenance is changing industries around the world, revolutionizing how plant operators manage and maintain their machinery for optimal performance, and providing significant cost savings.

Switching to long-term predictive and condition-based maintenance practices allows users to cost-efficiently transform many tasks from small-scope reactive to large-scope proactive. But what exactly is predictive maintenance?

Predictive maintenance (PM) revolves around condition monitoring (CM) and goes hand in hand with data analysis. CM is the continuous monitoring of machines and systems to ensure their optimal utilization. CM results are incorporated into the planning of targeted maintenance measures with respect to the actual condition of components, taking into account various parameters. The efficiency of the monitored machine is increased and an overall reduction in downtime costs is achieved. Both handheld and remote CM devices are available. Remote online solutions are quickly gaining in popularity, as they allow users to monitor machine health from their desk or smartphone. And with wireless solutions, users can now monitor an entire mine without having to run a single cable.

The combination of around-the-clock vibration-based CM and PM measures has taken industrial maintenance to new heights. Machines that, in the past, could not be monitored at all or, at best, only infrequently, can now be monitored without involving a lot of labor or endangering maintenance personnel. Early fault identification and analysis for nearly every imaginable point in a mining plant allow users to take appropriate measures to reduce downtime and optimize maintenance activities. 

When considering a comprehensive monitoring program, it is important to remember that production equipment is subject to varying requirements:

Process relevance: CM is strongly recommended for machines with high process relevance. Even simple online systems can help to provide continuous feedback on a machine’s condition. With respect to the possibility of consequential damage, systems that offer a broad range of functions should be considered. Route-based manual monitoring is not suitable, as this only provides selective insights into the condition of the machine. 

Simple and inexpensive solutions are appropriate for less critical machines. To cost-effectively monitor the large quantities of devices associated with this machine group, it is particularly important that the selected CM solution can be installed, commissioned, and used with considerable ease and speed.

Operating Parameters: Machines with continuous speeds and loads often have a lower process relevance. Once again, due to the large number of machines in this category, rapid installation, configuration and use of the selected CM system are necessary features. The CM solution should perform most tasks automatically in these applications.

Flexibility is key for CM systems used in applications with alternating speeds and loads. Numerous interfaces are required for exchanging speed or other vital information with the machine’s controller. To provide the most precise analytical results, the varying operating conditions must be taken into account. Ease of installation, configuration and use are, once again, vitally important here. 

Ambient conditions: Operating conditions can vary considerably. For the variable and, in most cases, harsh ambient conditions involved in mining, online CM solutions — which should perform most tasks automatically in these applications — are a viable option. 

Example from practice: 

Southeast Asian coal mining operator Adaro Indonesia was looking for a predictive CM solution that would minimize annual maintenance efforts, reduce safety risks, and prevent unplanned downtime of their conveyor belt bearings. The miner opted for Schaeffler SmartCheck, a robust monitoring system that detects deviations and changes in vibration behavior to identify bearing damage at an early stage — which can prevent an unplanned shutdown. Using SmartCheck, the operator avoided a production loss of around $200,000 per hour for each of the customer´s three production lines.  

CM systems can be complex, which is one of the reasons why, in the past, many companies avoided comprehensive online vibration-based monitoring. In the majority of plants, vibration condition monitoring is still performed manually, in addition to some isolated wired online vibration measurement for the most critical machines. But due to the high number of assets in a mining plant, only a small number of them can actually be monitored in this manner. As a result, costly unpredicted failures still happen regularly. With today’s Artificial Intelligence (AI) technologies, however, this no longer needs to be the case. Experience has shown that a robust digital maintenance strategy can offer considerable savings in overall plant maintenance costs. The game-changing technologies that make this possible are mainly low-power electronics and wireless technologies, platform technologies, as well as AI. Choosing the right combination of suitable technologies plays a decisive role in making the solution easy for the user. 

How much expert knowledge is required for efficient condition monitoring? Thanks to modern IoT solutions, automated CM systems require literally no prior expertise. Sensors do all the necessary monitoring, while alarm thresholds are automatically set and adjusted via a self-sufficient learning mode. Automated analysis functions provide easy-to-understand information, enabling users to address the needs of the components that require attention. This means that monitoring plant assets with constant operating conditions — such as motors, pumps and fans — is possible without prior specific expertise. 

When an acute machine problem occurs that cannot be detected or eliminated by the monitoring mechanisms already in place, experts can provide on-site as well as remote support when you need it, e.g., via troubleshooting or borescope inspection, depending on the application. 

Additional expert knowledge may also be required, depending on the analytical tool being employed, to continuously monitor machines with dynamic process requirements such as alternating loads or rotating directions. The data collected by an online monitoring system can be viewed and analyzed remotely by an external partner that provides regular status reports. In the event of unwanted changes in the machines’ status, appropriate recommendations are provided as warranted.

Technicians could also obtain training from experienced CM experts and become acquainted with the broad range of easy-to-use, automatic, and technically sophisticated solutions. Training opportunities for vibration diagnosis, for instance, start with basic fundamentals and can be expanded to certification seminars on various levels according to DIN ISO 18436-2.

Can CM save money? In most cases, mines benefit from cost savings and even an initial return on investment after the very first alarm has been issued by the CM system — simply because impending damage was prevented. Additional savings are realized by avoiding expensive unplanned production downtime. Modern solutions for remote condition monitoring provide an array of benefits, such as:

Better maintenance planning: No more replacing of spare parts “just to be safe,” even though they are still working. Early detection of possible failures enables in-time replacement or repair, which facilitates better planning of maintenance activities and spare parts inventories.

Availability of professional resources: Minimize the time spent on monitoring to free up the maintenance team for other, more important tasks.

Easy and safe access in hard-to-reach locations and hazardous areas: Whenever manual monitoring is difficult to perform or poses a hazard to team members, the risk of neglect is high. Taking action only when indicated in the app on a smartphone not only saves time and effort, it also increases occupational safety. Maintenance by exception – from a safe distance. 

Reduced CO₂ footprint: There is a correlation between vibration and power consumption in rotating assets. Power consumption increases when equipment is out of balance, for instance, or when a bearing is insufficiently lubricated. CM helps machines run smoothly without unnecessary extra friction, which helps keep vibrations under control.  

The success of any CM system is largely dependent on how well the solution is tailored to the requirements. When choosing your CM partner, consider the experience and product portfolio of the solutions provider rather than continuously changing specifications. Pay attention not only to the available hardware and software, but also to the provider’s service offerings and training support — as well as proven experience. 

This article was authored by Schaeffler Lifetime Solutions. It was originally published in the November 2023 edition of Engineering & Mining Journal (E&MJ).