Predictive maintenance processes are easy to implement

With food manufacturers being continually squeezed on price by retailers and asked to fulfil orders for supply that can seem, at best, challenging and at worst highly unrealistic, improving productivity is a priority. Tight timescales mean many lines are already running on a near 24/7 basis, leaving little leeway for even scheduled maintenance, let alone an unexpected breakdown. This can lead to overcautious service and maintenance regimes, which are expensive to support, but preferable to unscheduled downtime.

Short supply or delayed delivery due to plant failure damages a business’s reputation and impacts on the relationship with the customer. And as many food and beverage manufacturers find to their cost, customers such as supermarkets can’t support empty shelves, which makes them very demanding customers indeed.

Avoid grinding to a halt

Many production line failures are not characterised by a sudden fault that results in immediate line stoppage, often it is a gradual degradation. So, before the line eventually grinds to a halt, it might have spent a considerable period producing inconsistent goods – that add to bottom-line costs, due to waste.

“The good news is that random equipment failure doesn’t have to be a fact of life,” explains John Rowley food & beverage industry sales manager at Mitsubishi Electric: “Modern condition monitoring sensor technology can be easily retrofitted to rotating plant and equipment, while many of today’s plant and machine controllers have advanced monitoring and diagnostics functions built in.”

Taking advantage of these technologies can quickly take food and beverage companies into the realm of predictive maintenance, where businesses can see advanced warning of impending equipment failure, with enough time to plan repairs during scheduled maintenance periods rather than being hit with an asset failure out of the blue.

From preventative to predictive

A conceptual and technological leap forwards from preventative maintenance, intelligent predictive maintenance ensures an asset is serviced only when needed, helping to increase both productivity and efficiency. Predictive maintenance spots equipment problems as they emerge and develop, providing ample warning of impending failure, and so helping to maximise asset availability.

“Importantly, these predictive maintenance solutions are not complex,” Rowley says. “Frequently they are simple and cost-effective to implement, and often can be built from functions that already exist within the plant’s control equipment.

“Take, for example, the add-on sensors that have been developed to monitor the increases in operating temperature, excessive current draw, changes in vibration characteristics and significant shifts in other operating parameters that can all be indicative of impending problems in rotating machines. Today these sensors come with embedded ‘smart’ functionality, revolutionising condition monitoring.”

A simple add-on to pumps, motors, gearboxes, fans and more, these sensors use a simple traffic light system of red, amber and green lights to provide at-a-glance monitoring of the condition of the machine. They can also be connected into wider factory automation networks using Ethernet and a managing PLC for a smarter solution.

From traffic lights to telemetry

In isolation sensors offer a great start point to implementing preventative maintenance strategies, but there are limitations to the traffic light warning system. While it indicates that a problem is developing, it gives no real clue as to what the problem might be or just how serious it is; it offers no practical recommendations as to how the problem should be addressed; and, while it shows problems developing on individual machines, it fails to provide an overview on the asset health of the plant.

“It is these limitations that Mitsubishi Electric has addressed with the Smart Condition Monitoring (SCM) solution,” states Rowley. “The kit provides an integrated approach to monitoring the condition of individual assets and enables a holistic approach to be taken to monitoring the asset health of the whole plant. Individual sensors retain the traffic light system for local warning indication at the machine, but at the same time information from multiple sensors is transferred over Ethernet to a Mitsubishi Electric PLC for in-depth monitoring and more detailed analysis.”

The SCM kit is a plug-and-play solution with sensors that can be added to machines as and where required, and a simple teach function allowing the sensor and controller to learn the normal operating state of the machine, generating a memory map of key parameters.

Rowley adds: “Once set up, the SCM provides 24/7 monitoring of each asset, with functions including bearing defect detection, imbalance detection, misalignment detection, temperature measurement, cavitation detection, phase failure recognition and resonance frequency detection.”

Linking multiple sensors into the control system enables the controller to analyse patterns of operation that are outside the norm and alerts the operator when attention is needed. The SCM analysis provides detailed diagnostics, offers suggestions for where additional measurements should be taken, and provides maintenance staff more precise error identification. It is even said to make recommendations as to what rectification actions should be taken via text messages. This information can be networked to higher-level systems for ongoing trend analysis across all the assets around the plant.

How it should be done

Muntons Malt, one of the UK’s largest producers of malted barley uses the SCM system to protect fans and motors in its production process. It had previously experienced issues with difficult-to-reach bearings inside a large fan housing and was forced to make an unscheduled stop to a line to make repairs.

The SCM system was then installed on two 315kW and a 90kW fan sets. The company now has a very clear picture of any maintenance needs in advance of needing to make physical changes.

The system can work autonomously of any other automation, with multiple sensors located and recognised by unique IP addresses. However, at Muntons Malt the visual information as well as the alerts were connected into the existing automation software platform.

For multi-site businesses, this ease of connectivity can aid in quickly changing over production schedules from one plant to another to fulfil the most pressing orders or can alert remote maintenance teams of the need to perform more detailed diagnostics.

It may already be in your drives

This information isn’t just coming from external sensors. Modern drives, PLCs, SCADA systems and other automation products have comprehensive diagnostics capabilities inbuilt, monitoring not only their internal workings but also parameters such as current draw, voltage and temperature in connected motors, pumps, fans and compressors. All of this helps to build a detailed picture of the health of plant assets.

With a simple plant network backbone, this information can be shared around the plant and beyond. Indeed, this sort of functionality is a key aspect of Industry 4.0 and is at the heart of the benefits of the digitalisation of production.