This is what I have called "INACTIVE STATUS". When an instrument is placed in inactive status for an indefinite period, there are a few acceptable options. You can pick which ever fits your needs and requirements.
Back in my Navy days, that was an official status for an instrument. Their rule was it was to be cal'd immediately prior to placement in inactive status, then calibrated again upon removing from that status for active use.
As a minimum, the items need to be calibrated when you retrieve from inactive status.
Another option (I tend to favor currently) is to keep items in current calibrated condition (unless you expect them to sit for many years). If the normal cal cycle on an item is 12 months, and you expect perhaps it may sit (for example) for 18 to 24 months, there is something to be said for keeping it calibrated while in storage.
The reason is about what one of the other replies said... I've found that when items sit on a shelf, they sometimes tend to deteriorate as much as when they are used (sometimes more). This is not always true. Mechanical instruments with physical wear and tear in use (such as micrometers) may stay more stable. Electronic instruments with analogue switching, however, may tend to develop oxides on switch contacts when they sit. So those would possibly be in worse shape. DC Power supplies with large capacitors, the capacitors may dry out, PC board ground connections may oxidate, causing higher than normal ripple. Some stable handheld multimeters (such as Fluke 77) could sit for a couple of years and no noticable change (same as in normal use). All that to say that test instruments are all over the map in so far as what happens to them when they sit inactive for a couple of years.
Therefore, keeping them all in the calibration system (with some qualifiers below) maintains known operational status. I have a customer with about fifty or so power supplies who left them uncalibrated. They sit for a year or two at a time (sometimes a month or two, or a week or two). He had not calibrated any of them, as "they didn't need it." I recently went through them all. He was surprised to learn that a number of them had dead outputs, some had bad ripple, some had significantly inaccurate meters, etc.. As engineers who retrieve them (even though they monitor them externally with calibrated multimeters) want to at least be able to retrieve a working unit, the owner of this test equipment pool concurred that we should keep them all calibrated.
The qualifiers to my philosophy are:
1. High dollar calibrations. If you have to send back to the manufacturer (like a high end Hewlett Packard RF instrument) and need to save money, you could consider deferring that, and perhaps schedule a once a year op check.
2. Some mechanical instruments that mostly wear only in use (calipers, micrometers, etc.) could be deferred. You could possibly do a once a year walk through/inspection to be sure they had not oxidated or been damaged during storage.
3. Some process critical instruments that you need good R&R on that are in storage, you may want to consider keeping in calibration.
4. A final option one of my divisions considered - to save money they wanted to stop calibrating all the handheld multimeters used by maintenance people. I pointed out that part of the purpose of calibration is not just accuracy within tight manufacturer's specifications - it is reliability. If you use a multimeter "just to check line voltage", my contention is that if you did not care about any accuracy at all, why use a multimeter. You somehow inherently expect some level of accuracy when you use a multimeter. Even if you don't need the manufacturer's specs, you still want to know that that meter works - especially checking line voltage - people's lives may be at stake in the case of a bad reading. I recommended to the division that wanted to stop calibrating handheld multimeters that since they were already at 12 months, a more advisable saving might be to extend the calibration interval. The recommendation was to go to 24 months. Later, when cost savings were not so critical, they could review the percent in-tolerance on the meters, and determine reducing the calibration interval.
Back to your question. The simple answer is that an instrument that has been in storage ought to be calibrated prior to use on the manufacturing floor. Your application determines whether or not you need to take additional steps. If there is process critical, reliability, safety, or compliance to customer specs involved, consider whether or not you need to do an R&R, since you have lost continuity in the instruments in-tolerance performance.
