Committee Interview talking about ISO/IEC 17025

BACKGROUND If you couldn’t attend the annual NCSLI conference in Orlando in July, then you missed the luncheon plenary session regarding the (potential) new revision of ISO/IEC 17025:2005, which was led by Steve Sidney, Director, National Laboratory Association South Africa (NLA) and Jeff Gust, NCSL International’s Representative to the ILAC Laboratory Committee. Their slides can be downloaded from the NCSLI Standards Writing Group (ASC Z540) Committee webpage. However, the key takeaway points from their presentation are:

  • ILAC members voted over 80% in favor to recommend that ISO/IEC 17025:2005 be revisited for appropriate changes since the standard was first published, i.e., not simply be extended (renewed) another five years.
  • Of course, ISO/IEC 17025:2005 is officially owned by ISO CASCO (Committee on Conformity Assessment Standards), so any new revision would have to come from them.
  • ISO CASCO members have 90 days to vote/respond to a June 15 ballot initiative to revise the standard.
  • If the ISO CASCO ballot passes, then a CASCO working group will be convened to begin work.

Metrologist: I understand the plenary session created quite a buzz at the Orlando conference. I have three committee chairs who have graciously agreed to comment on what their committee members are saying: Welcome, Bob, Jim, Ryan

Bob Stern NCSLI 174 Standards Writing Committee, Keysight Technologies

Jim Wachter NCSLI 171 Calibration System Resources Committee, APT Research, NASA MetCal Program Office

Ryan Fischer NCSLI 146 Accreditation Resources Committee, Laboratory Accreditation Bureau

Metrologist: Before we dive in, I understand that your three NCSLI committees work closely together on a range of topics. Can you comment on that?

Bob: Sure, I’ll start. Folks come to the 174 committee to get early news/warning of new standards for testing and calibration. Our members generally prefer international standards, but we are ANSI accredited so are able to develop ANSI/NCSL standards with board approval as appropriate.

Jim: The 171 committee develops handbooks and recommended practices to provide guidance for the standards developed or adopted by the 174 committee. These guidance documents provide application guidance for their corresponding standards and are useful for seasoned professionals or newcomers looking to comply with the standards.

Ryan: When it comes time to assess a given testing lab or calibration lab for compliance to ISO/IEC 17025:2005 or other standards like Z540.x, then the 146 committee works to ensure consistent assessments to those standards. RP-21, “The Assessment of ANSI/NCSL Z540.3-2006 Sub-clause 5.3,” is an example of a recent work product from our team.

Bob: Or, in short: the 174 for “what are the rules?”, 171 for “how do I comply?”, and 146 for “how are the rules enforced?”

Metrologist: I understand that some people regularly attend all three committee meetings.

Ryan: (Laughs). Certainly some people do attend all three. Nevertheless, the audience is different in each of our committees, and we all work well together. Our committees are very diverse so that no one interest group dominates.

Metrologist: Are there any common objectives of the members of your committees?

Bob: Yes. Our members are passionate about the need to make accurate measurements in general. And specifically to comply with international and/or national standards with a minimum of extra work, as assured by uniform outside assessment.

Metrologist: Jim, we understand that your committee is presently working on a handbook for the application of ANS/ISO/IEC 17025:2005. Is that correct?

Jim: Yes, the request came from the 174 committee over five years ago and was approved by the NCSLI Board of Directors. The project has lost momentum a couple of times, but Working Group 2 is currently holding weekly telecoms to refine and edit the draft handbook.

Metrologist: But Jim, wouldn’t it be best to wait until ISO revises the standard before developing the handbook?

Jim: We’ve discussed this topic with NCSLI leadership and the plan is to proceed with the handbook. We really have no idea what the timeline is for revision; I think it’s conceivable that the current version of the standard could be in place for another three to five years. So, I believe a handbook for the current revision of ISO/IEC 17025:2005 is needed, and I think is will be a very useful tool for anyone seeking to comply with the standard.

Metrologist: Bob, your Standards Writing Committee is, in fact, accredited by the American National Standards Institute (ANSI). How do you respond to those who say your committee focus lacks an international viewpoint?

Bob: I’m really glad you asked that question! Turns out I heard this (criticism) second hand in Orlando just after our committee meeting had ended. While I couldn’t explore it in open committee, I was able to approach many of the committee folks. I think it is important to note that the committee membership is not restricted to US members. Further, organizations represented on the committee include those with interests outside of the US and those viewpoints are represented in everything the committee does. Most members, even those with no international affiliation, prefer to reference standards that are not only nationally recognized, but also internationally recognized.

Metrologist: What suggestions would your committees offer to a CASCO working group that is convened to revise ISO/IEC 17025:2005?”

Ryan: Accreditation assessments always start with compliance to ISO/IEC 17025:2005. Then additional requirements such as ILAC-G8, ILAC-P14, or ANSI/NCSL Z540.3-2006 are assessed. Today, the industries are confused with guidance being treated as requirements, and requirements not assessed consistently. The fewer standards needed to conduct an assessment, the better!

Bob: To paraphrase George Orwell from the novel Animal Farm, “All paragraphs are equal; Some paragraphs are more equal than others!” Jeff Gust sent out a questionnaire to all NCSLI members in January regarding issues/challenges/confusion meeting various paragraphs of ISO/IEC 17025:2005 and the perceived urgency to make changes now 14 years after the original standard was published. Our members would like to engage in discussions about real world use cases, problems and confusion with the current standard, and suggestions to avoid future confusion on key paragraphs of ISO/IEC 17025:2005. Why our members? Because they are the experts, within their respective organizations, that spend the most time and energy focusing on meeting the requirements of ISO/IEC 17025:2005.

Jim: We all know there are certain paragraphs in ISO/IEC 17025:2005 that cause confusion and are most in need of revision. I believe Jeff’s questionnaire clearly identified many of those paragraphs and a byproduct of our work on the ISO/IEC 17025:2005 handbook will hopefully be recommendations for revisions to those “heartburn” paragraphs.

Metrologist: Okay guys. Let’s follow that thread. If you could target a single area for revision what would it be?

Bob: For 174 members, that’s an easy one. ISO/IEC 17025:2005 paragraph 5.10.4.2 states “When statements of compliance are made, measurement uncertainty shall be taken into account.” Yet, there is no further guidance on how to do that! We know that this topic was controversial back in the 90’s when this standard was first drafted. Fourteen years later, everyone now recognizes that any measurement is only valid to the extent the associated accuracy is known. In our arcane metrology world, accuracy is measurement uncertainty (per JCGM 100:2008). There is no such thing as a perfect measurement with zero measurement uncertainty. So, when a measuring instrument/device has a specification and you measure very close to that limit, what is a test lab or cal lab supposed to report? Consumer risk of incorrect Pass/Fail decisions is now well understood. In fact, this, more than any other reason, drove, the creation of ANSI/NCSL Z540.3-2006 and the <2% PFA requirement (probability of false acceptance). So we’d like to see specific guidance in the revision for addressing consumer risk.

Ryan: Certainly there are multiple, ISO/IEC 17025:2005 paragraphs our 146 members would like to see written more clearly, but on the topic of paragraph 5.10.4.2 I would like to say that confusion regarding that paragraph leads to different interpretations across ABs and across geographic regions. In Europe, most ABs insists on compliance to ILAC-G8:03/2009 (Guidelines on the Reporting of Compliance with Specification) to meet 5.10.4.2, but in North America there has always been concern that ILAC-G8 is a “guideline,” not a policy. Remember, we are looking for clear instructions on how to meet ISO/IEC 17025:2005 with a minimum of references to further documents, and preferably not “guidelines.”

Jim: Just to add to what Ryan and Bob have said regarding paragraph 5.10.4.2, I would like to see ISO/IEC 17025:2005 provide a more-specific requirement for taking uncertainty into account when making conformance decisions. It’s clear that this paragraph is vague enough to generate confusion and to necessitate ILAC guidance. A specific risk-based requirement would serve to standardize the implementation of this requirement.

Metrologist: Suppose that a new revision of ISO/IEC 17025:2005 did address requirements that are now covered in ANSI/NCSL Z540.3-2006. Would we still need ANSI/NCSL Z540.3-2006?

Bob: We haven’t had any open committee discussions about this. I can say that I have had some interesting conversations with a number of 174 members, Ryan, Jim and NCSLI board members. The thinking is this à if overall compliance to standards could be made simpler; many of our members are willing to consider changes to ANSI/NCSL Z540.3-2006. My personal knowledge mainly covers testing or calibration laboratories, so I also defer to Jim and Ryan for their broader perspective.

Jim: The two standards have different scopes. ISO/IEC 17025:2005 provides requirements for the competency of testing and calibration laboratories, while ANSI/NCSL Z540.3-2006 provides requirements for controlling the accuracy of measuring and test equipment used to ensure an organization’s products and services comply with prescribed requirements. Calibration competency (sub-clause 5.3) is an element of ANSI/NCSL Z540.3- 2006 but the standard includes additional requirements for maintaining the accuracy of the equipment throughout its use cycle within the organization. Think of ANSI/NCSL Z540.3-2006 as a standard for organizations that use measuring and test equipment to provide products and services to an end customer. However, I would say that any future inclusion of elements ANSI/NCSL Z540.3-2006 sub-clause 5.3 in ISO/IEC 17025:2005, the probability of false acceptance requirement for example, would reduce the difference between ISO/IEC 17025:2005 and the ANSI/NCSL Z540.3-2006 calibration competency requirements.

Ryan: We assess the national and international standards required by the laboratory customers, regulators and specifiers. As long as the consumers of the laboratory services require the standards, we will provide the resources necessary to complete a competent assessment.

Metrologist: As we come to a close, is there anything you would like to say to our Metrologist Magazine readers?

Jim: We welcome participation in the 171 committee from anyone who would like to contribute to the development of handbooks and recommended practices. Contact me at: james.wachter@nasa.gov

Ryan: I’d like to echo Jim’s thought for 146 committee participation – while our core members belong to accreditation bodies, we absolutely welcome anyone interested in furthering the goal of uniform assessments. Contact me at: Rfischer@L-A-B.com

Bob: I too, invite anyone to be on the Standards Writing committee. We have some international regulars, but we’d love to have more. Even if you can’t be at our meetings, please visit our committee website at: http://www.ncsli.org/I/i/c/sp/s/c/Committees/170/174- committee.aspx?hkey=9aa417c5-6641-4896-9443-f4d5249a11a6 Contact me at: bob_stern@keysight.com