Asset Management & Reliability: What is the math telling you?

And organisations have very detailed asset maintenance plans including shutdown and turnaround scheduling, resourcing etc that have often been in place for years. But the question is – are they properly optimised?

In our experience, most organisations would say they could be better optimised, but they don’t have the resources or time to take a closer look and “do the maths”.

When it feels like everything is working well enough, the appetite to optimise often falls down the priority list unless some compelling event forces a shift in thinking.

In this blog post, we’d like to talk through the high-level considerations that could be the push you need to take a closer look.

Asset Cost of Ownership

This is the first area business should review and one that is commonly overlooked. It’s fundamentally about getting the maximum return on your assets – maximum uptime, productivity and throughput with minimised costs.

Every time you work on a piece of equipment, it costs you in 2 ways:

• Time and materials plus labour
• Operational downtime and productivity

So, it’s hitting you on the expenses and revenue sides of your balance sheet and here’s the first place that traditional operational thinking can misguide you. The default approach is that uptime must be maximised so very detailed maintenance schedules are developed to keep everything running as much as possible and minimising breakdowns.

But is all your equipment actually worth maintaining to that level? What if you do the maths looking for equipment that is cheaper to “run to fail” than maintain? i.e. You put it in the field and leave it to run with no inspections or maintenance and when it fails, you replace it with a new one and continue operating. No worrying about ordering or storing spare parts, training maintenance staff etc.

This could include older equipment that has been fully depreciated, or equipment that just makes no sense to fix. Just like that bunnings drill where the cost of a new aftermarket battery is almost the same as an entirely new drill + battery.

But that’s just the start? There are many variables that can then come into play.

For example, you might not have any equipment that explicitly falls into a “run to fail” category, but you might be constrained by the number of maintenance staff you have, or the length of maintenance windows your operational schedule will allow.

You might need 6 maintenance staff to do everything expected, but you only have 3. You enter that into the calculations to get an optimum maintenance strategy based on your current position for all assets (or a specified group of assets). It becomes an optimisation problem with multiple variables such as resources, time, productivity, cash etc.

What if….

The next part of the calculations comes from exploring different scenarios that may deliver a better result.

• What if I hire an extra person, or 2 or 3?
• What if I outsource some functions on a fixed contract basis?
• What if I add more equipment to my run to fail list – how does that change the outcome and where is the optimum point?

What about preventative maintenance schedules and looking at the mean time-to-failure data.

• Can I increase the intervals without impacting productivity?
• If we change from 99.9% availability to 99.5% availability, will the resulting decrease in productivity be more than offset by maintenance cost savings?

What role does safety play in this analysis?

• While it could be most cost-effective to decrease a maintenance schedule on a piece of equipment, the safety or regulatory requirements might not allow it.

Or environmental impacts perhaps?

• It might be cheaper to throw it away each time it fails, but is that in alignment with any corporate social responsibility policies you have if it can’t be recycled?

Each time you add variables, the maths obviously gets more complicated, but the results can be eye-opening and worth the investment of time.

Other variables to consider

What effect would additional IoT sensors and monitoring tools have on the calculations in terms of more accurate predictive maintenance?

Let’s look at a simple example with a pump. You might set your preventative maintenance schedules on 200 operating hours, or 100,000 litres of product. But better predictive maintenance capabilities could allow you to schedule preventative maintenance on-demand. That might sound a bit counterintuitive, but if you have a current fallback schedule of going out and looking at the pump once every 12 months, you could use monitoring sensors and equipment in conjunction with AI to automatically determine if there’s a change in operating parameters without anybody having to go out in the field.

All this IoT information (temperatures, pressures, vibration, noise etc) can feed into predictive maintenance calculations to drive these models. Also, IoT sensors tend to be cheaper, simpler and more convenient than the types of sensors used in SCADA networks. They allow you to pick and choose what you want to monitor separately from your SCADA system, or your process control system.

Anecdotally, the need for more IoT sensors seems to be rising as the number of long term (30+ years) employees starts to fall. While it’s not possible to draw direct conclusions, long-term employees often develop a “sixth sense” about equipment that helps them spot potential issues. Established maintenance schedules may have been developed with this knowledge in mind, but when these employees leave, that knowledge walks out with them. While trying to capture that knowledge is vitally important, it could also be a perfect time to run some of these calculations to develop new, optimised schedules due to the fewer (relatively) experienced operators remaining in place.

In any IoT discussion though, it is important to not always buy into the hype. The promise can sometimes outweigh the reality of what IoT can actually deliver, but it’s a rapidly improving area. However, as with everything we’ve discussed in this post, it’s obviously important to do the maths as the cost of implementation might have too long a pay-back period given your operations and not be practical.

Conclusion

This blog is taking a simplified, high-level view of Asset Management, Maintenance and Reliability, but the underlying principles still apply. When we discuss this with clients, we often hear, “there’s a lot here to consider - what should I focus on first?”.

In our opinion, it’s analysing your preventative maintenance schedules to find where you’re wasting money. You may be doing the wrong work at the wrong frequency to the wrong equipment and an optimisation exercise will deliver a quick payoff.

Our Digital Maturity Assessment can help. We utilise a Capability Maturity Model to assess our client’s current state of maturity - often benchmarking against other organisations of similar size and industry - to provide recommendations for improvement and where you should focus to move closer to an optimised state.