The most common mistake we see when organisations adopt reliability modelling tools is ambition. They try to model the whole plant: every pump, every valve, every instrument. Eighteen months later they have a beautiful, exhausted team and a model too big to maintain and too shallow to trust.
The principle that prevents this: analysis effort should follow criticality, not completeness. Rank your equipment by risk, consequence of failure, cost of failure, safety exposure, using the same framework described in our article on risk matrix modelling. Build detailed RBD, simulation and RCM models for the critical few, where the value of a better decision is measured in hundreds of thousands of dollars. For the non-critical many, OEM recommendations and generic strategies are not a compromise. They are the correct allocation of scarce engineering attention.
A Tiered Model Estate in Practice
- Tier one, the critical few: full availability simulation, RCM with cost-benefit analysis, fitted failure models and optimised spares. Deep, defended, and kept current.
- Tier two, the important: targeted analysis of the dominant failure modes, library-based failure models, and strategies reviewed on a cycle rather than continuously.
- Tier three, the balance of plant: OEM strategies, generic tasks, and monitoring through normal work order metrics, with promotion to a higher tier only when the evidence demands it.
This is how the tools have been applied at gold-standard level across Australian heavy industry since the late 2000s: deep models where the money and the risk live, light touch everywhere else, and a model estate small enough that people actually keep it current. It is the asset-level expression of the argument in our white paper Working on the Right Things: reliability engineers should move to the bottleneck, not the job description. And it guards against the mismatch our paper Setting Expectations documents, where analysis effort and organisational need quietly part company.
For projects heading toward commissioning with no operating history, this discipline matters even more. Every hour of analysis effort is precious. Spend it on the equipment whose failure would hurt most, and let the model grow outward as real data arrives; our article on reliability without failure data covers the starting position. A criticality ranking done in week one pays for itself many times over by directing everything that follows.
How the Tools Help You Discover the Benefit
Availability Workbench supports the tiered approach naturally: consequence and risk structures rank the fleet, full AvSim and RCMCost treatment goes to the top tier, and parts libraries plus drag-and-drop reuse let proven component models propagate as the estate grows. The result is a small, sharp set of models that people trust and update, which beats a plant-wide model nobody opens, every time. Our asset reliability and reliability program assessment services help organisations rank their fleets and scope the right first models, and the podcast episode One Size Fits All, or Not makes the case in conversation.
Go Deeper
- Software: RCMCost, AWB Availability Simulation, Reliability Parts Libraries
- Services: Asset Reliability, Reliability Program Assessment, Vulnerability Assessment and Analysis
- White papers: Working on the Right Things, Setting Expectations, Backward Blueprinting
- Podcast: One Size Fits All, or Not
- Back to the start: Reliability Concepts series index
How can we help? Talk to us at contact@mantua.group.