ISO 17025 requirements not enough to prove NIST Traceability?

[Slighter]

Has anyone ran into this? How did you answer it? I did some calibration for a new customer, and they sent it back saying they need NIST test numbers. I asked where they were getting that from, since I thought we stopped using them in the late 90's, and this is what they sent me:

Our nuclear requirement comes from ASME NQA-1-2008, Requirement 12, Paragraph 301, “Calibration shall be against and traceable to certified equipment or reference standards having known valid relationships to nationally recognized standards, or to international standards known to be equivalent to and verified against corresponding nationally recognized standards.”

The Nuclear Regulatory Commission (NRC) compared required nuclear requirements with 17025. The reviewers found that 17025 addresses all but two of the nuclear administrative requirements for calibration. The following two requirements must be included in purchase orders to commercial-grade calibration suppliers.

1. Calibration report/certificate shall include identification of the laboratory equipment/standards used, including NIST traceability

2. Calibration report/certificate shall include as-found and as-left data. (17025 requires that as-found data be reported only if the item requires adjustment or repair.)

To meet the above requirement NIST numbers shall be documented or the calibration is found deficient and data taken with the equipment is suspect until NIST traceability can be positively confirmed or data is re-taken with equipment having correctly documented traceability.


We are an ISO17025 lab and this calibration was within our scope...

 

[John Broomfield]

Has anyone ran into this? How did you answer it? I did some calibration for a new customer, and they sent it back saying they need NIST test numbers. I asked where they were getting that from, since I thought we stopped using them in the late 90's, and this is what they sent me:

Our nuclear requirement comes from ASME NQA-1-2008, Requirement 12, Paragraph 301, “Calibration shall be against and traceable to certified equipment or reference standards having known valid relationships to nationally recognized standards, or to international standards known to be equivalent to and verified against corresponding nationally recognized standards.”

The Nuclear Regulatory Commission (NRC) compared required nuclear requirements with 17025. The reviewers found that 17025 addresses all but two of the nuclear administrative requirements for calibration. The following two requirements must be included in purchase orders to commercial-grade calibration suppliers.

1. Calibration report/certificate shall include identification of the laboratory equipment/standards used, including NIST traceability

2. Calibration report/certificate shall include as-found and as-left data. (17025 requires that as-found data be reported only if the item requires adjustment or repair.)

To meet the above requirement NIST numbers shall be documented or the calibration is found deficient and data taken with the equipment is suspect until NIST traceability can be positively confirmed or data is re-taken with equipment having correctly documented traceability.


We are an ISO17025 lab and this calibration was within our scope...

Slighter,

As well as being accredited to ISO 17025 your lab management system will require conformity to customer requirements (provided these requirements do not diminish the calibration).

Apparently your customer did not specify these additional requirements for the calibration report or certificate:

1. Identification of the laboratory procedure used, including NIST traceability?
2. As-found and as-left data?

Accordingly you could ask the customer to pay for the calibrations to be completed according to their changed requirements (particularly with respect to item 2).

Your customer may argue that as your lab is the expert in calibration that the NRC requirements should been determined as a result of your lab doing contract review.

Whatever way it pans out you need to update your management system to elicit these requirements during the sales process.

John

 

[Slighter]

Thanks for the reply, right now the issue is their nuclear auditors not excepting the fact that we don't list NIST test numbers. I calibrated some thermocouple wire, so the as found is the as left. We list a cal procedure adding the revision/date wasn't a big deal. I thought though being ISO17025, we had already proven traceability. I just don't know why the NRC doesn't accept it. Should I try and get NIST numbers from the people who calibrate our standards?

 

[John Broomfield]

http://www.iaea.org/inis/collection/NCLCollectionStore/_Public/27/063/27063493.pdf

Thanks for the reply, right now the issue is their nuclear auditors not excepting the fact that we don't list NIST test numbers. I calibrated some thermocouple wire, so the as found is the as left. We list a cal procedure adding the revision/date wasn't a big deal. I thought though being ISO17025, we had already proven traceability. I just don't know why the NRC doesn't accept it. Should I try and get NIST numbers from the people who calibrate our standards?

Slighter,

Yep, here is some more information on NIST numbers.

You will see NIST numbers used in this example here.

John

 

[Slighter]

John,
Thank you for the link and the example.

 

[Hershal]

NRC is still struggling I suspect with understanding ISO/IEC 17025:2005 and also traceability and legacy contracts.

NIST numbers by themselves are not valid for traceability. If you only provide the NIST numbers then the calibrations are not considered traceable. To be traceable the evidence must make it clear that the NIST number is descriptive of a specific calibration of a specific item at a specific time, and the calibrations provided to the customers must be traceable by unbroken chain to that.

But, your customer is painted into a corner.

 

[Slighter]

Fortunately for me, our ISO auditor was here, a couple days after this issue came up. I also talked to the calibration manager for the company that does our standards.
In the end, we all agreed that for the test number, I should put the calibration certificate number for the standard, and if NRC want to chase the train then that is on them.
I really appreciate all the help everyone has given.
Thanks
Aaron

 

[jwright]

Well, I would not rely on an answer from an ISO auditor. You need to understand that NQA-1 (or 10CFR50 Appendix B rather, since this is the hard requirements) is a Nuclear Quality Program, not be confused with an ISO QMS.

Pure and Simple: NQA-1 is written for Compliance, while ISO is written for Customer Satisfaction. Confused? While ISO is focused on the process and to make sure the process is controlled in order to meet Customer Requirements, NQA-1 is focused on making sure the product is compliant, meaning inspection.

Here is the real answer: NQA-1 does not apply to you. If you are a Cal Lab, then to your Customer, you are a Commercial Service. You do not have an Appendix B program (I am assuming) and are working from PO requirements. Your Customer needs to define how they are meeting the Commercial Grade dedication to justify how they are using your service, not you.

 

[Norm Moreau]

Having audited a number of calibration shops over the years for the nuclear and other regulated industries we have always asked if the calibration of calibration equipment is traceable to NIST or other recognized organization. The fact that the requirement was not in your customer's P.O. is the first place to go to. The NRC letter that the customer is citing is accurate and the NRC is expecting that information. What this letter has done for the nuclear industry is, if done properly, reduced the number of audits utilities have to do by requiring the calibration facility to be ISO 17025. Great idea, however the utilities have to do a couple of other things to make this work and passing on requirements in their P.O. is one of the them! And yes, there may be a cost for this additional service.

 

[Hershal]

Fortunately for me, our ISO auditor was here, a couple days after this issue came up. I also talked to the calibration manager for the company that does our standards.
In the end, we all agreed that for the test number, I should put the calibration certificate number for the standard, and if NRC want to chase the train then that is on them.

That is actually the best method to document the traceability.