Interpretation - Must Calibration Procedures be Published Internationally?

Jerry Eldred

Forum Moderator
Super Moderator
I think if your calibration procedures cover the details from Recommended Practices from the National Conference of Standards Laboratories, it would be difficult for any auditor to find a problem. I would recommend you review ISO 10012 and NCSL Recommended Practices for calibration procedures. My guess, from reviewing, is that it is referencing 4.11. The wording is overly vague in that. If you go to 4.11.2, and refer to the note at the end of that, NOTE 18, it references ISO 10012. In ISO 10012, para 4.7, there are a few paragraphs. THE NCSL recommended practices are very detailed, and are considered a best practice for how to write calibration procedures.

I must apologize, as nearly all of my library is packed for a forthcoming office move, or I could give you a quote from the recomended practice. The web address for the NCSL is, I believe www.ncsl-hq.org. I'll have to check to be sure.

Your procedures must cover some basics:

1. Document control/revision control
2. List of the measurement standards to be used
3. Detail of each measurement uncertainty for each calibrated parameter on the measurement standard.
4. Detail of each tolerance limit for all calibrated parameters of the device to be calibrated
5. Step by step instructions for all of the process steps
6. Environmental requirements, warm up requirements for all equipment

One note of caution is that a manufacturer's adjustment procedure is not acceptable. There must be instructions for making a set of measurements prior to making adjustments to verify the unit was received in or out of manufacturer's specified tolerances.

A second note of caution is that the procedure must be specific enough that it is unambiguous as to specifically what measurement standards may be used.

For example, on a DVM, a meter calibrator is not sufficient. For a specific model DVM, a specific model meter calibrator that may be used.

I have written in the past, generic procedures for DVMs. In that procedure, I was very specific about what models of meter calibrator could be used. The reason for this is for example, that if a Fluke 5100B calibrator is the meter calibrator in the lab, if the procedure just says meter calibrator, there is a loophole which may allow a technician to unintentionally use that meter calibrator for a meter of too high accuracy.

I also deliberately made a statement in all of my procedures that the 4:1 test uncertainty ratio must be maintained, I even documented in the procedure what the actual ratio was in some instances where it could make a difference.

A good resource for well written procedures (if you do government contract work) is GIDEP (Gov't Industry Data Exchange Program).

I don't know if I have covered the right information to help in your resolution. Please let me know if I can help further.


Art Kitkowski Interpretation

Paraphrased "The laboratory shall use calibration methods, which are appropriate for the calibration it undertakes, preferably the current issue of those published as international, regional, or national standards".

Our in-house cal labs perform cal or simple instruments such as calipers, DVM, thermocouples, digital thermometers,etc. The cal procedures are documented and follow the cal procedures recommended by the O.E.M. of the articles. We considers this sufficient to meet the requirements of

A recent audit interpreted that the cal procedures (ignored the word preferably)must be published international, etc. standards. We tried to find these standards for some simple cals, on the web, etc. and could not. Any thoughts on the interpretation and any ideas where standards for cal can be found?
Top Bottom