Comply with 17025 or Customer?

[qaadrian]

Our calibration lab is working towards implementing ISO 17025. The cal lab is pressuring me (QA) to back off on some of the requirements. I responded OK, but you cannot take credit for compliance to 17025 on the calibration certificate. They objected. My question is this: If you comply with the customers calibration requirements, which do not include compliance to 17025, can you still take credit? Is there such a thing as partial compliance???

This issue is frustrating me. How do you meet both customer and 17025 requirements in a QA Program if the customer asks for something less than 17025 or if the cal lab intentionally does not comply with all of the requirements? How do you word this condition on the certificate? How do you answer an auditor when confronted with noncompliance to 17025, but comply with customer requirements? Thanks for the feedback.

 

[Al Dyer]

...Your questions

...I responded OK, but you cannot take credit for compliance to 17025 on the calibration certificate. They objected.

What is the position of upper management?
Who is going to issue this "certificate"? I don't think a registrar would.

...If you comply with the customers calibration requirements, which do not include compliance to 17025, can you still take credit?

Credit for what? Meeting customer or registrar requirements?

...Is there such a thing as partial compliance???

Is there such a thing as almost pregnant?

...How do you meet both customer and 17025 requirements in a QA Program if the customer asks for something less than 17025 or if the cal lab intentionally does not comply with all of the requirements?

You are either 17025 compliant or not. Just make sure all customer requirements are included in the 17025 scheme. If the lab intentionally won't comply you have a bigger problem than can be handled here. Where is upper management?

...How do you answer an auditor when confronted with noncompliance to 17025, but comply with customer requirements?

You need to define what you want to do. It sounds as if there is a lack of communication and direction in your company.

If you are audited by a registrar to comply with 17025, you will fail, not much to answer. You will probably get a pat on the back for meeting customer requirements (if they are defined) and a report telling you where you do not comply.

ASD...

 

[Graeme]

If your calibration lab implements ISO 17025 and are successfuly accredited, they will have made a major achievement. You can tell your customer that you are giving them everything they want plus a lot more. Top management can be confident that having the accreditation will draw even more customers. The people in the lab will have proof that their work really is as good as they have always told you it is.

By the way, ISO 17025 is a formal accreditation scheme, just like ISO 9000. (In the US, these accreditations are done by A2LA and NIST's NVLAP group.) A self-declaration of conformity (if that is what you are talking about) has only as much value as a customer is willing to give it.

 

[Ryan Wilde]

Even under formal accreditation, you do have an out here. You are always allowed to deviate from 17025 in order to fulfill customer requirements, as long as you explain to your customer that what they are requesting is not in compliance with 17025 (preferrably in writing). After this, remove any references to 17025 from the calibration certificate and feel good that you did what the customer wanted.

The second part, with the cal lab intentionally not complying with all of the requirements. There is a term used quite often when you deal with government contracts known as 'legal metrology'. In other words, if you claim compliance with 17025, and do not comply, you are legally responsible for your actions, and yes, there are actual court cases that have covered this exact topic. If you can't get management to actually commit to backing 17025, explain that it is also a legal issue, and that if they claim compliance, especially on the calibration certificate (which is a legal document) then they had better ensure that it is so. If they still do not want to commit, I would personally advise them to scrap the idea of complying with 17025. Remember, the very first part of Management Responsibility is 'legally identifiable'.

Concerning how you word it on the certificate - you do not. If the calibration and system do not comply with 17025, then you cannot state it on the cert. The customer that requires you to deviate from 17025 must take the brunt of their auditor's wrath, but you have done your responsibility to comply with everything possible, and you can sleep well at night.

Ryan

 

[qaadrian]

Thanks you for answering my questions. If I may, I'd like to get specific. Section 5.10.1 of 17025 indicates that the results shall be reported.... The Lab is telling me that they are going paperless and that a certificate will not be issued if a customer does not ask for it. The Lab will use MetCal/MetTrack (a Fluke program) to store calibration data only. It may lack some of the information required in section 5.10.2, including a signature of the person authorizing the calibration results. I've heard that going paperless is acceptable. Can the Lab take credit for complying with 17025? Is there more that has to be done to support a paperless system and comply with 17025?

 

[Ryan Wilde]

Originally posted by qaadrian:
Thanks you for answering my questions. If I may, I'd like to get specific. Section 5.10.1 of 17025 indicates that the results shall be reported....

In my experience, providing actual calibration data to the customer is not a requirement, as long as the data is available. The caveat here is that the customer requirements rule this area - if they want the data, they get it.

The Lab is telling me that they are going paperless and that a certificate will not be issued if a customer does not ask for it. The Lab will use MetCal/MetTrack (a Fluke program) to store calibration data only. It may lack some of the information required in section 5.10.2, including a signature of the person authorizing the calibration results.

Electronic signatures (or marks - whatever you like to call them) are acceptable as long as only the person whose "signature" is shown can provide that signature. If I'm not mistaken, MetCal/MetTrack requires passwords for data entry, which suffices in this regard. They may also be able to not provide a paper certificate, but again, that must be acceptable to the customer, and "acceptable" generally means "in writing".

I've heard that going paperless is acceptable. Can the Lab take credit for complying with 17025? Is there more that has to be done to support a paperless system and comply with 17025?

Yes and no. All paperless systems I've seen involve the issuing of electronic certificates, generally using PDF files, which can be digitally signed. I would contact the organization that you will (would) use to become accredited for the answer to that. As for more work, I believe that the non-issuance of any documentation will be a nightmare to control, because I guarantee that some customers will require it, the only way I can see it not being a problem is if you are an internal lab with no external customers and the CEO says it's peachy.

Ryan

 

[rock]

Check the third paragraph of 5.10.1. Specifically "written agreement" and "readily available". If you have both of these, there should be no finding.
Mike