Third party auditor mentions no grace period for calibration

[Galaxy32]

Hi Guys,

I have a question relating to this post.

Our internal procedure regarding calibration intervals states that our standard calibration interval is 12 month +/- 1 month. This is to allow an extra 30 days if an external calibration provider is unavailable to be onsite within the 12 month period. This has been our practice for at least 15 years, and we recently had an auditor question that practice. The auditor stated it was unacceptable, but was unable to provide any clauses in ISO17025, ISO9001, or ISO13485 stating that factoring in a "grace" period was unacceptable.

The quality manager and I have looked through other various specs and resources to find some verbiage that either say we can extend our calibration interval an extra month if needed, or that we can't extend the calibration interval, but it seems to be a giant grey area. We are conforming to our procedure, but the auditor believes that our practice is wrong. Does anyone know of any specs or standards that has a definitive statement regarding this?

 

[Jim Wynne]

Hi Guys,

I have a question relating to this post.

Our internal procedure regarding calibration intervals states that our standard calibration interval is 12 month +/- 1 month. This is to allow an extra 30 days if an external calibration provider is unavailable to be onsite within the 12 month period. This has been our practice for at least 15 years, and we recently had an auditor question that practice. The auditor stated it was unacceptable, but was unable to provide any clauses in ISO17025, ISO9001, or ISO13485 stating that factoring in a "grace" period was unacceptable.

The quality manager and I have looked through other various specs and resources to find some verbiage that either say we can extend our calibration interval an extra month if needed, or that we can't extend the calibration interval, but it seems to be a giant grey area. We are conforming to our procedure, but the auditor believes that our practice is wrong. Does anyone know of any specs or standards that has a definitive statement regarding this?

The person who makes the assertion (the auditor in this case) bears the responsibility to support it. You should stop looking for support for an idea that doesn't need it, and politely tell the auditor to put up or shut up.

 

[Ajit Basrur]

Also refer the below related threads:

Calibration grace period after due date

How to determine Preventive Maintenance and Calibration Grace Periods

 

[Galaxy32]

Thank you both for the quick responses.

 

[Sidney Vianna]

Calibration intervals, including potential grace periods can only be determined intelligently, on a risk based basis. As I mentioned in another, recent post

That's an area where the ill-defined RISK BASED THINKING should be used to the fullest. There are so many variables, such as tolerances involved, criticality of the characteristic being measured, criticality of the product, potential for product recall, ease of verification/calibration, use conditions, environmental conditions, etc...

I've seen processes where before every shift, all of the fixtures, inspection, test and measuring devices in an (high volume) assembly line were verified. Reason being, the organization wanted to contain any potential nonconforming product from further processing.

 

[Tagin]

Several years ago, our ISO 9001 auditor actually suggested to us that it would be prudent to include a grace period for our calibrations (we had none at the time).

 

[outdoorsNW]

While you do not state which standard the auditor is using, I am going to assume ISO9001.

The ISO9001 standard allows you to set the calibration period. The standard allows you to set a calibration period of two years or longer if reasonable (such as you have no calibration failures or the tool is not used for the final quality check on a feature or process). In your case you are effectively setting the period as 13 months. The risk is higher with longer periods, but if there are no failures you can save some money.

Since you documented the period as 13 months, the auditor is wrong.

I have worked in large systems calibration periods were 3, 6, 12, 18, and 24 months depending on the item. The 3 and 6 month periods were for heavy use items that could not reliably go 12 months without calibration.

As a side note, if you are having many calibration failures without a known special cause, such as a dropped tool (assuming the tool is not used until it can be checked), then the auditor might be able to write you up under ISO9001 section 7.1.5.1. But if you are having many failures, likely many tools are out of calibration by month 10 or 11, so the argument about the extra month is a moot point.

 

[dwperron]

Hi Guys,

I have a question relating to this post.

Our internal procedure regarding calibration intervals states that our standard calibration interval is 12 month +/- 1 month. This is to allow an extra 30 days if an external calibration provider is unavailable to be onsite within the 12 month period. This has been our practice for at least 15 years, and we recently had an auditor question that practice. The auditor stated it was unacceptable, but was unable to provide any clauses in ISO17025, ISO9001, or ISO13485 stating that factoring in a "grace" period was unacceptable.

The quality manager and I have looked through other various specs and resources to find some verbiage that either say we can extend our calibration interval an extra month if needed, or that we can't extend the calibration interval, but it seems to be a giant grey area. We are conforming to our procedure, but the auditor believes that our practice is wrong. Does anyone know of any specs or standards that has a definitive statement regarding this?

I'm curious as to what you put in your calibration records as the due date for the next calibration. Do you have it due in 12 months or 13 months?

17025 addresses this issue in that you need to have a due date or interval defined.:
6.4.13 Records shall be retained for equipment which can influence laboratory activities. The records
shall include the following, where applicable:
e) calibration dates, results of calibrations, adjustments, acceptance criteria, and the due date of the
next calibration or the calibration interval;

13485 is similar:
7.6 Control of monitoring and measuring equipment
As necessary to ensure valid results, measuring equipment shall:
a) be calibrated or verified, or both, at specified intervals...

Using your grace period method you don't define / specify a due date or interval.
If you mean to have a 13 month interval then state it, don't imply it. Nothing wrong with that.

Yes, you can have a QMS that allows you to extend the interval, but you still need to define when the calibration is due.

 

[Sidney Vianna]

Yes, you can have a QMS that allows you to extend the interval, but you still need to define when the calibration is due.

What if the interval is set in a non traditional manner? Something like "a micrometer must be calibrated, after 50 instances of normal usage?" So, if the device gets used once a week, it would be almost a year, before re-calibration is due. If the device gets used once a day (assuming 5 business days/week), the device would be re-calibrated after 10 weeks.

A pre-established "time interval" is not smart, in my opinion. I checked the ISO/TS 9002:2016 document and they offer the following pearl of wisdom, concerning this subject:

 

[dwperron]

Galaxy 32 was asking about the requirements in ISO 17025 and 13485 concerning calibration intervals.
Both require you to define the calibration interval, and 12 ± 1 month is not an interval.

If the interval is 50 uses then you need to keep a record of the number of uses, and the calibration is due after number 50. Not at the end of the month of the 50th use.

Your quote from ISO 9002 concerns the choice of an interval based on risk of the measuring device being out of tolerance, not on the risk of it being outside of its calibration interval. It is telling you that if the measurements you make are critical to your process then you should determine a calibration interval commensurate to that risk. But you still need to choose and define an interval.