Calibration vs Verification of Measuring Instruments
Looking for collaboration on a standard that prescribes gauge management activities in our business
Is anyone familiar with NSF-14?
10 Replies
I work in an industry where we have to meet NSF-14.  What paragraph are you looking at?
Hi Ryan,
It is section 9.4.1 and 9.4.2.  We are complying with the weekly verification, but it seems excessive.  The instruments are calibrated, say, quarterly, but then are subject to an additional record weekly. 
In NSF-14 Section 9, where gauges are required to be ‘verified’ weekly, checking zero and a reading ‘close to’ the critical part dimension that they’re used for….  Our NSF agency contact states that calibration is not needed on the instrument, only on the gage block used to verify it.  I’ll note that threads are exempt from this.  But it’s in effect for, say, a caliper.
The ISO 9001 standard doesn’t spell out calibration, but I’ve always developed a calibration check that uses the whole range of an instrument.  For example, a 6” caliper may be checked with blocks as .5000”, 2.000”, 3.000”, 5.000” and 6.000”.  In this NSF explanation, if the caliper is only used on parts that are 2” and smaller, you would never test the high end of the range of the instrument.
I wanted to know if this implementation seems most efficient, or if you have considered it differently
Barb

 
7.1.5.2 of ISO9001:2015 talks about calibration, but very light compared to earlier versions of ISO.  We treat the weekly verification like you described - one block or pin, but we also do an annual calibration across the range of the instrument.  If you don't do the full calibration at some frequency then you are essentially only doing a limited calibration and I would expect an external ISO auditor to question whether that caliper would be used for more than just 2".  If it is or could be then you might run into a nonconformity from an ISO standpoint.  
Hi Ryan,
It is section 9.4.1 and 9.4.2.  We are complying with the weekly verification, but it seems excessive.  The instruments are calibrated, say, quarterly, but then are subject to an additional record weekly. 
In NSF-14 Section 9, where gauges are required to be ‘verified’ weekly, checking zero and a reading ‘close to’ the critical part dimension that they’re used for….  Our NSF agency contact states that calibration is not needed on the instrument, only on the gage block used to verify it.  I’ll note that threads are exempt from this.  But it’s in effect for, say, a caliper.
The ISO 9001 standard doesn’t spell out calibration, but I’ve always developed a calibration check that uses the whole range of an instrument.  For example, a 6” caliper may be checked with blocks as .5000”, 2.000”, 3.000”, 5.000” and 6.000”.  In this NSF explanation, if the caliper is only used on parts that are 2” and smaller, you would never test the high end of the range of the instrument.
I wanted to know if this implementation seems most efficient, or if you have considered it differently
Barb

  
Barbara,

As I mentioned earlier - We treat the weekly verification like you described - one block per week.  We will also be doing an annual calibration across the range of the instrument using multiple blocks.  This is because we use this instrument across its full range.  If you don't do the full calibration at some frequency then you are essentially only doing a limited calibration.  This is fine if you never use the caliper above 2".  Some auditors may expect that this to be noted on the gauge so if someone were to pick it up and want to use it for say 3" they'd know it wasn't calibrated up to 3".  It really depends on your ISO auditor.
ISO 9001 does not address calibration.  It tells us to control our monitoring and measuring resources and ensure they are suitable for the job ISO 17025 prescribes the methods for calibration.
I do not know the standard that you are referencing.  Nothing in the ISO 9001 standard prescribes a frequency for verification.
Hope this helps.  
Debra,

You are absolutely correct that ISO9001 doesn't specify a frequency.  I apologize if my earlier comments suggested that it did.  However, NSF14 which is a standard some of us are required to follow does require a weekly verification of some measurement equipment.  ISO simply says when measurement traceability is required the measuring equipment shall be calibrated or verified, or both, at specified intervals or piror to use ...  
A common 483 finding is "calibrated instrument used outside of it's calibrated range".  It is best to calibrate any instrument over it's entire useful range.   This avoids the situation were someone picks up the instrument and uses it for a novel GMP purpose - outside of the calibrated range.  Alternatively a sticker can be place on the instrument specifying the calibrated range.  This calibrated range information needs to be captured in a notebook, logbook or LIMs, whenever the instrument is used.  
NIST requires that as a minimum you calibrate at 25%, 50% 75% and 100% of the range of the instrument. This covers the entire range of the instrument. You can do more calibration points of course, the 4 are a minimum.
Hi, Larry. Can you provide the source for the NIST requirement? Thanks!