ISO 17025 5.10.2(e) - Test Method

[drturbo]

Hi everyone

We've got a condition in a recent ISO 17025 audit because we did not reference the test method as it appears on our scope of accreditation.

The test method in question is AS 1391 (Tensile Test) - which appears on our scope of accreditation. However the client requests testing in accordance with AS 3678 or AS 3679.

AS 3679 section 9 nominates AS 1391 as the test method for Tensile Tests with additional requirements.

We list AS 3679 as the test method on our tensile report because, we believe, it is traceable to the scope of accreditation by nominating AS 1391 as the method.

Our accreditation body's response was that AS 3679 is not a test method, and isnt listed on our scope of accreditation. However they said that we could get AS 3679/AS 3678 listed on our scope through a desktop variation audit.

This is quite alarming to us, as we perform testing to a multitude of material specifications, and having each one potentially listed on our scope of accreditation will lead to a massive scope. Also listing each material specification seems to be a bit overkill.

My question is, [FONT=&quot]If we document the traceability between the test specification on our reports to the test specification on our scope of accreditation would this satisfy ISO 17025 5.10.2e? [/FONT]

It might be worth noting that we have been performing these tests under the scope of accreditation since 1996 (20 years) and this is the first time we have seen this condition raised.

Thanks

 

[dwperron]

This looks like a valid finding against you.

I come from a different 17025 world, calibration, but we use the same standard. Scopes of accreditation for testing labs usually list the standards to which they are accredited, we list the actual measurement types. This often does lead to a "massive scope". My current lab has a modest 20 page scope, a previous lab I worked in had a 62 page scope.

For accreditation you need to demonstrate competency in your test methodology. For a lab that performs multiple disciplines that leads to demonstrating a lot of competencies, but that is the purpose of accreditation: showing that you have demonstrated competency in your fields so that a customer / regulatory body can accept your results with confidence. That is why the scopes have to list the specific test, to assure that you have demonstrated technical proficiency, repeatable results, have trained staff, have adequate test methodologies, etc. So if you actually test to AS 1391 you need to demonstrate the competencies of that method. You have not done that with AS 3679, thus you cannot claim that to be "accredited" in that method.
For the customer's satisfaction I think that I would have listed AS 1391 as my method with additional verbiage indicating that it is the method nominated by AS 3679. That might make everybody happy. I used to do that with customers who required calibrations to AMS 2750.

 

[drturbo]

Thanks for your reply dwperron.

I understand your point, and after some further internal discussions I think we disagree more with the direction our accreditation body is taking with the scope. After reading the ILAC guidance on laboratory SOA i can see why they are trying to change, but the process is going to cripple us commercially.

Current wait time to have an equivalent standard added to our scope is 7 weeks now. Customers will not wait that long, and I fear that this will lead to issuing reports without the endorsement, or customers taking their business to another lab with the specific code on their SOA.

 

[dwperron]

I hear you, brother.

It has always been a real pain to try to anticipate what a customer will be looking for.

I always thought it would be great to send them a letter in advance of your accreditation letting them know that it is coming up and requesting them to let you know what new requirements need to be added for the next 2 years...

But I've always been a dreamer.

 

[Hershal]

Disclaimer: Not knowing the AB or assessor who wrote the finding, I can only give guidance from the Standard itself, and my experience as an assessor.

Your position is correct, but were I doing the assessment I would have expected to see the test method per the SOA with addition of AS 3679 Section per customer requirements. That would then comply with Clause 5.10.2.e and also 4.4.1.c.

Hope this helps.