6 min read
Ask five workshops and you will get twelve months five times. It is a reasonable starting point and it is also the single least examined number in most maintenance systems. A torque wrench used twice a month on a clean bench and one used every shift on a pit face in the wet season are on the same annual cycle in almost every operation we walk into, and there is no engineering reason for that.
What actually moves a tool out of calibration
- 01Cycles. A click wrench has a spring and a mechanism, and both wear with every operation. Cycles matter more than elapsed months.
- 02Load as a fraction of range. A tool habitually worked at the top of its range ages far faster than one worked in the middle of it.
- 03Environment. Dust, water, vibration and heat all degrade mechanisms, and a wet underground mine is a different world from a workshop.
- 04Events. A dropped tool is out of calibration until proved otherwise, regardless of the date on the sticker.
- 05Storage. A click wrench stored at load rather than wound down loses its rate, quietly, while nobody is using it.
Start from consequence, not from convenience
The first question is not how often, it is what happens if this tool is wrong. A wrench used on non-critical fasteners and a wrench used on a mill liner or a haul truck wheel assembly carry different risk, and that difference should show up in the interval.
| Situation | Reasonable starting interval |
|---|---|
| Occasional use, clean workshop, non-critical joints | 12 months |
| Daily use, workshop conditions, specified joints | 6 to 12 months |
| Daily use, field conditions, dust and wet | 3 to 6 months |
| Safety-critical or high-consequence joints | 3 to 6 months, and after any suspected event |
| Any tool dropped, shocked or behaving oddly | Immediately, whatever the sticker says |
Then let the as-found data settle it
This is the part most operations never use, because their certificates do not carry it. A certificate with as-found readings tells you how far the tool had drifted in the interval that just ended. That is the only real evidence about whether the interval is right.
- A tool that comes back year after year comfortably inside tolerance is telling you the interval could be longer
- A tool that comes back marginal, or out, is telling you it should be shorter, and that the work it did recently deserves a look
- A tool that swings unpredictably between certificates is telling you something is wrong with the tool, not the interval
What to do with a lapsed tool
A tool found past its calibration date has not necessarily done any harm, but you cannot know that yet. Take it out of service, calibrate it, and read the as-found result. If it was still inside tolerance, you have a record saying so and the exposure is closed. If it was not, you now know which period of work needs reviewing, and roughly how far out it was.
Operations that throw the lapsed tool straight onto the calibration bench without recording as-found readings destroy the only evidence that could have answered the question.
Common questions
Is twelve months a rule?
No. It is a widely used default. Intervals should be set from use, environment, consequence and the tool’s own as-found history, and reviewed as that evidence accumulates.
Does a dropped torque wrench need recalibrating?
Yes, before further use. A shock event invalidates the certificate regardless of its date, and the as-found reading will tell you whether it caused a problem.
Should click wrenches be wound down after use?
Yes. Storing a click wrench at load leaves the spring compressed and it will lose its rate over time, which is drift caused entirely by storage rather than use.