17025 Para. 5.6.2 - Calibration certificates - Measurement results - Uncertainty

[joecalibratio]

17025 states that when using external calibration services, traceability of measurement shall be assured by the use of calibration services from laboratories that can demonstrate competence, measurement capability and traceability. The calibration certificates issued by these laboratories shall contain the measurement results, including the measurement uncertainty and/or a statement of compliance with an identified metrological specification.

My question is, finally, when our calibration lab sends its primary standards to an external calibration lab, do we need to specifically request As-Found/As-Left Data with our calibration certificates? or Is As-Found Condition as stated suffice to meet 17025?

Joel

[Kevin H]

Hi Joe, it's been about 5 years since I was working with calibration requirements for a lab. (At that time it was guide 25 & we were transitioning to guide 17025.) For calibration certs, we needed to receive as found and as left conditions, to be compliant with guide 17025. We also required an instrument/device condition to be reported - an example might be: Used, in good condition, repaired, new, etc.

[Randy Stewart]

Joel,
According to our auditor we needed to specifically request 17025 calibrations and have it noted on the certificate. The "As-Found" and "As-Left" are needed to verify measurment based upon your standards in the event that one is damaged or is giving errors.

[joecalibratio]

Do I take that to mean that if we specify a 17025 calibration on our primary standards sent to an external calibration lab, that those units will (should) be returned to us with data?

[Hershal]

It is highly desirable to report readings, whether in or out of adjustment.....however, it is not specifically required unless the item is out of tolerance (OOT).

As for measurement uncertainty, yes it needs to be reported.

Hershal

[Randy Stewart]

Joel,
If you put that requirement on your PO, the info should be returned to you. I haven't had an issue receiving the data, the calibration house we use was giving us the 17025 requirement without it being requested.

[Graeme]

For many calibration labs you do have to specify before and after data if you want it, otherwise you probably won't get it. And it usually costs more.

Why do you want both sets of data for a calibration standard? If the standard is subject to drift or wear, the "before" data quantifies how much since the prior calibration. The "after" data sets a baseline value for the upcoming interval. Both values get plotted on a run chart, and once you have "enough" data you can start to calculate the actual uncertainty of your standard based on its recorded calibration data. You also get information that may allow (or require) adjusting the calibration interval for best performance. Just for those two types of knowledge the data is worth the money.