DesignSpark Electrical Logolinkedin

Predictive maintenance: spend a little to save a lot

RS Components looks at maintenance, specifically how to justify predictive/preventative solutions.

Spending money to save money doesn’t come easy to some. But in the case of predictive and preventative maintenance regimes – especially for high-capital assets – a relatively modest spend could potentially offset repair or lost-production costs that run into millions.

The problem with maintenance spend is that it is very hard to quantify its benefits if nothing actually goes wrong – accountants tend to like tangible figures as opposed to those generated by “what-if” scenarios. Couple this to budget competition from other disciplines and the job of justifying the expenditure becomes even harder.

As it is tough to define an “opportunity cost/loss” in terms of maintenance statistically, it is far easier to look at the issues simplistically by determining, for example, the cost of maintenance vs. the cost of repair/replacement and the expected life of a machine or asset vs. its extended service life. In all cases, the cost of lost production, goodwill and customer confidence must also be factored in to this equation. Even with these simple analyses, the need to have maintenance as a strategic element of an operational plan, becomes very clear, but what is the next logical step?

For those with complex operations, the obvious route is the deployment of a computerised maintenance management system (CMMS), which will help remove much of the guesswork and workload by defining a strategy and timetable as well as the actual work to be undertaken on individual machines and assets. More often than not, this type of proactive approach is combined with technological solutions that offer a real-time snapshot of machine health.

These solutions range from simple temperature sensors and vibration probes all the way up to bespoke, fully featured condition monitoring (CMS) software solutions. How they are selected, deployed and subsequently interrogated depends very much on the application.

In the case of electronic systems, most are “clever enough” to know when they are going wrong or are much easier to interrogate using intelligent systems, although cabinet-based components do need careful monitoring. Mechanical and electromechanical systems, on the other hand, tend to be more demanding and can fail in a much wider (and noisier) variety of ways.

In the majority of mechanical systems, wear is the primary culprit, resulting in (to name but a few) reduced tolerances, elevated heat levels, slower and jerkier movement and the inevitable loud squeal and grinding that foreshadows something going seriously wrong. The good news is, with proper servicing and regular observation, mechanical systems can easily match and even exceed the predicted overall life of the machine.

One of the primary predictive measurements in the case of wear is heat generation. Lubricant failure and additional friction almost always equate to elevated heat levels. In these cases simple, inexpensive temperature sensors come into their own.

One example is the PN151 NFC Infrared Temperature Sensor from Calex. (905-8768)

 

By mounting a sensor like this near potential wear points, operators can be quickly made aware of any issues that are normally preceded by elevated temperatures.

Another approach would be to use thermal-visualisation solutions (893-8489), such as those offered by FLIR. These deliver a ‘heatmap’ in an easy-to-decipher format and are designed to highlight specific hotspots, delivering the means to target maintenance more precisely.

Flir AX8 (893-8489) is a unique thermal sensor with imaging capabilities.

Wear is also normally accompanied by vibration, especially in rotating equipment.

Specialist vibration sensors exist, such the CMSS 200 (818-6947) pictured below from SKF, which can measure multiple frequencies in order to differentiate between normal and abnormal vibration.

 Indeed, some sensors and their associated software can even highlight the potential failure mode by interrogating the frequencies being measured and comparing them against those generated by known failure modes.

Earlier, we mentioned maintenance as a strategic tool. With this and a longer-term view in mind, plant managers have to cater for the changing demographics of their workforce, from one that comprises a greater number of older, specifically skilled workers, who know their machines more intimately and can therefore interpret subtle wear signals, to a smaller, younger multi-skilled workforce who rarely build up the same level of technological relationships. This younger workforce is also much more comfortable with digital solutions.

From a strategic operational perspective, maintenance expenditure should be seen as something that can increase overall equipment effectiveness (OEE), create faster return on investment (ROI), lower reactive repair costs, reduce secondary damage and contribute to increased or maintainable product quality. These are the tangible benefits that have to be at the forefront of any expense justification; and with a modest expenditure potentially having a massive positive impact on virtually all strategic goals, it is very hard to create a cogent counterargument.

Predictive maintenance products help contribute to the range of RS solutions for customers looking to establish an effective plant maintenance programme.

rbhullar has not written a bio yet…