Quality Magazine

Other Dimensions: Cal Labs Aren't Always Right

January 5, 2012


A reader wrote me recently of a situation he was trying to sort out with a couple of his calibration sources. Some of their comments left me shaking my head in wonder, but that’s not uncommon these days when so many are still not accredited. Here are some examples from my reader and others.



Due dates on reports

My reader sent out some adjustable thread rings for calibration and requested that a due date beyond one year be shown on the report. The lab advised him that they were “not allowed” to do this. A discussion ensued and my reader had trouble making the lab understand that it is up to the customer to make such decisions-not the cal lab.

A quick look at ISO 17025 confirmed this to be the case. It clearly states that any such statement on a report must reflect what the customer wants, not what the lab thinks it should be.

This makes sense because the gage owner is privy to the conditions of use and the lab is not. And the owner knows what the potential use is going to be for the gages as well. That doesn’t make the owner an expert in these matters, but he or she has more critical information needed to make the call that the lab does not have.

It goes without saying but I must-some gage owners look at calibration frequency based on economics rather than performance. This can be very dangerous and I can understand a laboratory recoiling in horror. The best thing a lab can do is advise the customer when their request is off the wall. Yes, some auditor will read the report, and not understanding ISO 17025, will think the lab is out to lunch when the due date decision was made by their customer.

If the customer doesn’t want to change the due date, the lab could put a footnote on the report indicating that the due date was determined by the customer. Alternatively, the due date could be flagged with a note suggesting that it be reviewed by the customer.

I find it interesting that so many people want a due date on a report, or cal labs put one on without discussing it with the customer. The creators of ISO 17025 were concerned that disreputable labs could use this as a method to drumming up business. As a result of this situation, the standard doesn’t require a due date to be put on a report.

A calibration report is a summary of the state of the item calibrated at that point in time. After it leaves the lab anything or everything can change. A due date, however created, could be made obsolete by a number of factors such as:

An accident that may require an item to be calibrated sooner than what the cal report says to confirm it is still within the limits.

A review of the calibration data may indicate to the customer that the due date should be changed. That will only become apparent after the calibration report with a now inappropriate due date is received.

Disputes between departments or manufacturers and their customers often can trigger recalibration.

Your customers may indicate they want changes to your due dates.



Too many gage users want a due date on the report so they don’t have to justify their choices to an auditor. If it’s on a report from a reputable lab, an auditor is unlikely to question it, so the user gets a free ride. At the end of the day, however, the gage user has to justify calibration frequency irrespective of who made the judgment call.

I believe the due date should be part of the gage record so its history is traceable along with reasons why changes may have been made. Under this scheme, you don’t have a report stating one thing while your frequency has been extended or compressed.

The item itself will have a status sticker on it with a due date and when that sticker is missing, damaged or illegible, the gage record will note what it should be.

Now, all you have to figure out is how to express the date so it will fit in the space on the sticker and the format you use is understood by all who need to know. Being a simple kind of guy, I think it’s hard to beat “May/12.”