Simply put, a criticality analysis identifies potential risks. That means that maintenance tasks can then be assigned according to each asset’s criticality, which ensures a good operational condition of all equipment. In fact, a risk assessment is a major step to:
- Prioritise assets for total productive maintenance;
- Determine priorities for risk-based maintenance;
- Plan a condition monitoring program;
- Decide which pieces of equipment need a root cause analysis.
How to Make a Criticality Analysis in Maintenance
There isn’t a single, foolproof recipe to make a criticality analysis. Most maintenance managers come up with a criticality matrix, a 6×6 grid where the x axis represents the severity of the potential failure and the y axis its likelihood.
We’ll do our best to explain a 3 step process to organise your own criticality matrix. However, bear in mind that this process is flexible itself, and you can make it as lengthy or as thorough as you wish. (More on a few shortcuts later.)
1. Put together the pieces of the puzzle
If you don’t have one already, start with an asset list. Then, put together a team with staff from different departments – maintenance, engineering, operators, or anyone else who deals with equipment.
Now, how much do you know about each asset? If it fails, what are the consequences? Remember to account for the impact on safety and the environment, not just for customers. What’s the lead time for spare parts? This information will allow you to calculate the failure’s severity.
Then look at data to calculate the probability of failure and assess equipment condition. Check the asset’s log and reactive maintenance history, single-point failures (parts that, should they fail, shut the whole system down), and mean time between failures.
At this point, you may also estimate the asset’s lifespan and assign it a category from A to D, where:
- A – VERY GOOD: the equipment is in optimum condition and is expected to function properly until the end of its normal lifespan;
- B – GOOD: the equipment is in good condition with only some signs of deterioration;
- C – OPERATIONAL: the equipment is operational but needs major reparation works or even some parts replacements;
- D – UNSTABLE: the equipment is unstable and in danger of breaking down;
- There is also an intermediate category B/C, which means the equipment is currently in category B but will switch to category C within 5 years.
💡Pro tip: organise your assets using a functional hierarchy (for example, production floor > production unit > individual pieces of equipment in the unit). That way, when you find low criticality equipment, you can skip its “subordinates”.
2. Get everyone on the same page
Some people can’t imagine booking a hotel without breakfast included. Others can’t imagine waking up before 10 AM while they’re on holiday. It’s all a matter of priorities.
It’s important to agree on a criticality matrix. Define your evaluation criteria and combine dimensions (health & safety, environmental, operational) to create your matrix. You can follow ISO guidelines or adjust things if you feel one dimension is more detrimental to your business than the others, for example.
3. Assess failure risk
If your only concern is the operational and financial impact you can calculate the criticality score with the following formula:
Failure frequency (per year) ✕ Cost consequences (€)* = Risk (€)
cost of the consequences = cost of lost production + repair cost
If you want to go ahead with the criticality matrix you need to look into the data you’ve collected.
First, you’ll need to figure out the risk priority number (RPN) of each asset, which determines the severity of the failure.
How to calculate the risk priority number
Evaluate each dimension (Health & Safety, Environmental, Production and Operational) from 1 to 5, where:
- 1 = insignificant
- 2 = minor
- 3 = moderate
- 4 = severe
- 5 = catastrophic
For reference, you can use the ISO14224 Asset Criticality Consequence Ranking:
Then, multiply all the scores to calculate the RPN. For example, if a failure has a rating of 3 in “Safety”, 1 on “Environmental”, and 4 on both “Production” and “Operational”, its RPN will be 3 ✕ 1 ✕ 4 ✕ 4 = 48.
Estimate the frequency of failure
Finally, estimate the probability/ frequency of failure. Or, again, give it a score from 1 to 5, with some guidelines for reference:
- 1 = rare (less than 0.5 per year, on average)
- 2 = unlikely (0.5 – 1 per year, on average)
- 3 = possible (1 – 2 per year)
- 4 = likely (greater than 2 failures per year)
- 5 = certain
Make the final Math
Now, you should have a clear grid to assess criticality. Or you can multiply the severity and probability scores (from 1 to 5) and come up with something like this:
The colours represent each risk level: low (green, ≤ 7), moderate (yellow ≤ 10), significant (orange ≤ 20), and high (red ≥ 20).
For instance, if the probability of a failure is “possible” (3) and the potential consequences are “critical” (4), then the risk equals 3 ✕ 4=12 (significant risk).
Now, we did say we’d discuss some shortcuts, so let’s get to it. If you don’t have the time or the resources for a thorough assessment:
- Understand that failure relates to events that trigger it. Since it’s impossible to take into account all the events that can lead to all the failures, focus on the likeliest one with the worst outcome in terms of risk.
- If you don’t have enough time to assess the score for each dimension (safety, environment, operations) focus on the one with the highest risk level.
The risk assessment is complete. What’s next?
Once you complete the risk assessment, the maintenance plan for equipment or facilities should be prepared. Naturally, priority should be given to those that carry significant or high risk.
It’s recommended to perform an FMEA analysis on the top 20% of critical assets. This way, you can be sure each preventive maintenance task matches a failure mode. Not only is this more efficient at preventing failures, but it will also save you time and resources on unnecessary maintenance.
👉 You may also be interested in this actionable guide on how to prevent failures and decrease reactive maintenance.
And, of course, remember that in maintenance… nothing is ever truly over. Try to review your criticality analysis at least every 5 years.