Repairable spare parts (rotables) are high-value components you overhaul and return to service, while consumable spare parts are low-cost items you use once and throw away. The distinction matters because each type follows a completely different financial and logistical logic. Consumables are simple to reorder in bulk, but rotables live inside a repair loop that ties up capital, requires a "float" of ready spares, and rewards careful tracking. Getting this split wrong means either mountains of dead stock or a critical line idle while a €12,000 gearbox sits waiting for repair.
The line between the two categories is defined by economics and physics, not by any fixed price threshold. A part is a good candidate for the repairable pool when it meets several of these conditions:
Consumables sit at the opposite end: seals, filters, gaskets, fuses, belts, bearings under a certain size, and lubricants. They are cheap, non-restorable, and consumed in volume. You manage them with straightforward stocking policies rather than a repair workflow.
When a repairable part fails, you cannot wait for it to be overhauled before the machine runs again. Instead you swap in a ready spare from a pre-built pool, send the failed unit out for repair, and receive it back into the pool later. That circulating pool of interchangeable units is the rotable float.
The float exists to absorb the repair turnaround time (TAT). The longer and less reliable your repair loop, the bigger the float has to be. Consumables need no such float because there is no return leg: you buy, install, and discard, so a simple reorder point and safety stock policy covers them.
Suppose a plant runs 8 identical filling heads, each driven by a servo actuator that is repairable. History tells you:
Expected failures during one 45-day (0.123-year) repair loop equal 5.3 x 0.123, which is about 0.65 units in for repair at any moment. Because failures cluster, you add a safety buffer. Using a simple Poisson coverage rule for a target service level around 95 percent, you round up to 2 float units. So the plant carries 8 installed actuators plus 2 spares, a total pool of 10, with roughly 1 unit typically out at the repair vendor.
Now flip the numbers. If the repair vendor slips to a 90-day turnaround, in-repair demand doubles to about 1.3 units, and you would size the float to 3 to hold the same service level. This is the core lesson: float size is driven far more by repair turnaround than by the raw failure rate. Shortening TAT is often cheaper than buying more spares.
Every rotable should carry a status at all times. A workable set of states is: installed, serviceable on shelf, awaiting repair, at vendor, and condemned. Without this visibility, units vanish into "somewhere in the repair process" and the float silently shrinks until a failure catches you short.
Serialising each unit is what makes this possible. A serial number lets you build a repair history per physical part, spot the "bad actor" that fails repeatedly after every overhaul, and decide when to condemn it. Feed removal and repair events into your reliability metrics so you can watch MTBF and MTTR trend over the part's life. Rotables belong in the "A" tier of an ABC analysis, warranting tight control, while consumables usually fall into B and C tiers managed by simpler rules.
Because consumables have steady, predictable demand and no return leg, they respond well to classic inventory formulas. Use economic order quantity to set batch sizes and a reorder point tied to lead time. Watch inventory turnover so you are not sitting on years of filters. Where a supplier can manage replenishment, vendor-managed inventory keeps consumable shelves full with little effort.
Rotables ignore that math. Their "demand" is failure-driven and their "supply" is a repair queue, so you size the float against turnaround and service level as shown above. The right posture also depends on maintenance strategy: parts on a condition-based maintenance program can be pulled and rotated on a planned schedule, while a pure run-to-failure approach demands a larger float to survive random breakdowns.
Managing rotables well depends on trustworthy, current data about parts, assets, and work, and that is exactly what Fabrico provides as a real-time data foundation. Fabrico is an EU-built CMMS with EU data residency, so its spare parts, work order, and asset modules let you record each rotable against the asset it serves, log removals and installs through work orders, and schedule preventive pulls before failure. Linking spare-part consumption to real OEE and downtime data, captured live (including from machines with no PLC via computer vision), means your float decisions rest on what is actually happening on the floor rather than guesswork. Fabrico gives you the operational record; the repair-loop policy and float sizing remain yours to set. Moving from reactive to proactive maintenance gets much easier when the part history lives in one place.
Compare the cost of the overhaul against a fresh replacement, then weigh reliability. A common rule of thumb is to condemn once repair cost exceeds roughly 60 to 70 percent of new, or once a serialised unit becomes a repeat "bad actor" that fails soon after each overhaul. Tracking per-serial repair history makes this call obvious rather than emotional.
Sometimes the same physical item is treated differently by context. A mid-size bearing might be consumable on a low-criticality conveyor but part of a repairable assembly on a critical drive, where you rebuild the whole assembly rather than scrap it. Classify by the assembly and asset criticality, not by the raw part number alone.
Repair turnaround time. As the worked example shows, cutting turnaround roughly halves the number of float units you need to hold at a given service level. Negotiating faster vendor repair, or bringing high-volume repairs in house, usually beats simply buying more spares.
Ready to give your rotable pool and spare-part data a single live source of truth? Book a Fabrico demo and see how real-time OEE and a field-ready CMMS keep maintenance running while parts move through the repair loop.
Login
Blog
