Machines get health checks online

Condition monitoring is a proven, valuable tool for optimising maintenance strategies. Ian Liebler, Rockwell Automation integrated conditioning monitoring solutions architect, explains how integrated condition monitoring can provide even greater production and maintenance efficiencies through streamlining real-time information flow.

In order to ensure the ongoing efficiency and operation of plant, machines must be maintained and serviced. Yet, while too little maintenance will result in an increased number of breakdowns, too much will cause needless machine downtime and reduced productivity. The optimum maintenance schedule lies somewhere between, demanding that plant operators come up with ways of predicting when servicing might be required, so that maintenance can be carried out more strategically.

Yet, while this approach may minimise unscheduled downtime, there are inherent drawbacks. Servicing might be conducted needlessly frequently to accommodate a margin for error, resulting in excessive unwarranted downtime. Moreover, such a program does not guarantee that breakdowns will not occur, because the ‘health’ of the machine is not being monitored between services.

Regular health checks

Regular monitoring of machine health - or condition monitoring - has proven a valuable tool for providing operational continuity, regulatory compliance, safety and reliability of plant operations. Fundamentally involving the collection of key machine data (such as vibration data), condition monitoring techniques can be used to identify any changes in machine operations and instigate strategic condition-based maintenance (CBM) as required.

Conventional condition monitoring is carried out at regular intervals, often using handheld devices, and is by nature separate and isolated from other plant control systems - including machinery protection systems, which are usually also present. However, the emergence of integrated control technologies is promoting a shift towards ‘integrated condition monitoring’ (ICM), where the control, protection and condition monitoring elements are merged into a single, integrated architecture.

The aim of ICM is to leverage existing infrastructure for the incorporation of CBM strategies and improve the flow, accessibility, and ‘actionability’ of information. Ideally, ICM provides full connectivity between the condition monitoring system and the plant enterprise - including computerised maintenance and monitoring systems (CMMS) - and supports multiple types of data and data collection strategies.

In the realm of ICM, there are nevertheless different levels of integration, depending on the infrastructure that is present: conventional ICM and fully integrated ‘online’ ICM, as shown in Figure 2.

In many plants, data collection strategies are likely to follow two distinct paths - one from maintenance and one from operations. Maintenance personnel will collect and analyse condition monitoring data and then feed the results into a database for historian trending. Concurrent with this, control data will also be collected for operational requirements. The two types of monitoring are distinct and separate, although integrated into the same system.

Taking ICM online

It is possible to take integration one step further, however. Fully integrated online ICM removes the need for separate data collection. Here, all field sensors - both maintenance and operations - feed signals through to the controller to allow it to perform auto-diagnostics. All data is online, providing real-time access to current information. The results can be viewed on any SCADA or HMI interface and relayed back to a database for historian trending. Real-time data can, alternatively, be routed to the person who needs the information, or passed back to a software interface for non-time-critical functions, such as maintenance management scheduling and data analysis.

The ability of online ICM systems to handle machine vibration data is a fundamental requirement. Vibration data can be collected in two ways: an overall measurement across a wide array of frequencies; or discrete measurement of vibration magnitude at specific frequencies. While an overall measurement strategy will still highlight vibration increases, and thereby flag a forthcoming maintenance issue, it will not necessarily be able to diagnose the cause. Discrete vibration measurement can be used, however, measuring key narrowband frequencies associated with aspects of interest for a given machine - such as imbalance, misalignment, blade pass and blade pass harmonics.

The frequency associated with alignment issues may be monitored, for example. If a vibration is measured above the designated threshold for that frequency - indicative of a specific alignment fault - an alert can be set up to produce a real-time message for the machine operator, specifying an alignment problem. This will allow the operator to facilitate targeted corrective action, resulting in less machine downtime.

Integrated intelligence

ICM systems can also leverage the intrinsic intelligence of the controller. As the diagnostics are managed at the controller level, the controller can be programmed to do more than provide real-time alerts to the operator. It can be further tasked with providing protection or even taking remedial control action.

Excessive vibration on a conveyor, for example, might result in the controller being tasked to reduce the belt speed, thus minimising the vibrations while still allowing the unloading of material currently on the conveyor. In a similar way, vibration alerts from a bearing shaft, caused by excessive friction, can be used to task an auto-lube function through the controller, thus rectifying the likely cause of the problem. Control actions that minimise or rectify machine faults are especially valuable where, due to process or safety considerations, machines cannot simply be shut down.

The ability of online ICM to transmit real-time data also provides enhanced levels of monitoring and diagnostics. This will provide a more accurate picture of the health of key machines, including machines that are critical to production or safety, have minimal spares, are difficult or expensive to replace, have a history of inherent faults, and are prone to catastrophic failure.

The provision of this real-time data will enable maintenance activities to be provided only when required, and targeted to the cause of the monitored symptom. Furthermore, ICM has the functionality to provide increased levels of protection and control of machines and processes, extending the working life of a machine and ensuring maximum productivity for minimised downtime.

By Ian Liebler, Rockwell Automation