The Future of Laboratory Accreditation

Thomas G. Smith

When I began to write this article, I had just finished responding to an e-mail in a Guide 25 chat room. I responded to several pointed statements on laboratory accreditation. The purpose of this article is to share with you, the reader, one perspective on the future of laboratory accreditation.

The American Association for Laboratory Accreditation (A2LA) is one of the two national laboratory accreditation bodies. As the owner of an A2LA-accredited commercial calibration laboratory and an assessor for the association, I have been faced with preserving the level of accreditation through many trials and tribulations. My passion for metrology runs deep with the conviction of a teenager racing his first street rod at the drag strip.

Once certain of all elements of the race and of the data from three previous test runs (proficiency tests) down the track, the car and driver together run a predictable time (bracket racing), or best measurement capability. The rest is left to chance.

The Driving Force

It is important to understand what drives metrology. (Picture a B-body 1965 Dodge 440 with 650 thumping horses sitting at the starting line.)

Who is racing? Industry is the driving force because many fields need calibration and testing services. One of those service providers, metrology, is a loosely regulated industry driven primarily by client trust. Some calibration labs even operate as cottage industries (out of home, garage, car, or van). Others operate from converted office spaces using simple window air conditioners (if any), up to prefabricated environmental rooms that used well-documented guidelines for calibration rooms.

The Track

For many years, manufacturing and calibration followed a set of military standards. MIL-Q-9858 was used for quality systems requirements, MIL-I-45208A for inspection systems requirements, and MIL-C-45662 for calibration program requirements.

The audits that followed these requirements have become legendary for their auditor-dictated, highly interpretive criteria. (Some of these were documented and some were not. The mood of the auditor seemed to dictate the outcome.) Auditor fear was firmly instilled and, for the most part, the distance industry had to travel was established.

Not long ago, the new guys overseas built a new track. All racers that wanted to do business internationally were required to use their track and obey new rules. This was the beginning of ISO 9000 Grand Prix racing. Team Manufacturing and Team Service racers were having to learn the new rules. Most of them wanted to become officially registered.

In 1978, the International Laboratory Accreditation Cooperation (ILAC) developed ISO/IEC Guide 25 to provide universal criteria for accreditation, that is, formal recognition of competence to specific tests and calibrations. Team Metrology now had its rules, called a guide, but this guide was written like a standard, using words such as shall and must rather than should and may. Using Guide 25 as a core for content, another document, ANSI/NCSL Z540-1-1994, was created in 1994 to support U.S. Team Military.

This document is the U.S. version of Guide 25, but it is tailored to calibration laboratories only. The document was embraced and it replaced MIL-C-45662, which the military canceled soon thereafter. The military's race committee then determined that a handbook would be beneficial. In October 1995, the Handbook for the Interpretation and Application of ANSI/NCSL Z540-1-1994 was published. Behind the scenes, Team Metrology felt the rules were somewhat vague and lobbied for ISO-9000 and QS-9000.

In 1997, the international race officials committee, the ISO Conformity Assessment Committee (CASCO), was given the authority to develop international standards. The decision was made to publish a revision to ISO/IEC Guide 25 as ISO 17025.

Draft International Standard (DIS) 17025 was published in July 1998 (the seventh revision). A five-month voting and comment period by ISO and IEC members followed. The U.S. ISO representative, ANSI, voted negative and provided extensive comments. The standard-drafting group met in February 1999 to review the comments. Working Group 10 met in April 1999 to consider revisions proposed to the drafting group. It will likely be late 2000 or early 2001 before this document is official and formally replaces Guide 25 and possibly ANSI/NCSL Z540-1-1994.

The Cars

Guide 25 provides one of the key cornerstones to the foundation on which to build mutual recognition agreements (MRAs) among accreditation bodies. Internationally, accreditation bodies are a function of government agencies. Government agencies grant accreditation based on a uniform set of rigid criteria and performance standards. Any calibration lab that wants to sell services must be accredited.

The United States has two internationally recognized accreditation bodies: the National Voluntary Laboratory Accreditation Program (NVLAP), which has links to NIST, and A2LA, a nonprofit organization. These two groups have MRAs with each other and with international agencies. All mutual recognition partners use ISO Guide 58 to establish operation-system requirements. Guide 58 provides for a process in which each partner observes actual audits of that partner's clients. They audit one another's quality systems, control mechanisms, and personnel competence.

After the on-site visit, the auditing partner provides documented feedback and the audited partner formally responds, where applicable, with corrective action. Participating labs are also subject to proficiency testing. They are measured on how well they make their measurements on a common artifact compared to other labs making the same measurements on the same artifact. Their measurements must be within stated uncertainty estimations, which are commonly referred to as uncertainty budgets.

This is where the rubber meets the road for commercial and accredited labs. Are the lab's uncertainty elements defined and estimated well enough to produce consistent measurements? This first estimate of lab-only measurements defines large bias components. How do different lab technicians compare when making similar measurements on a common artifact using a common calibration device, a common method, or both?

Once those data are derived, the artifact is sent to other labs to make the same measurements. The result compares those measurements to a population of others making the same measurements. The results are important for improving measurements, addressing lab bias, and continuing training. A known provider of proficiency-testing (PT) programs for calibration labs is the National Association for Proficiency Testing, a nonprofit organization specializing in calibration PT programs (http://www.proficiency.org).

Traceability requires calculation of uncertainty, a link to the primary standard by an unbroken chain of calibration by using national or international accredited labs, as well as intrinsic and consensus standards. ISO 17025 also includes "traceable to SI units" and a statement of compliance to metrological specification.

The Drivers

A2LA has mutual agreements with 11 third-world governmental organizations and cooperatives, with three in the final stages. Along with rigorous program requirements for specific fields of calibration, assessors are trained intensively. Normal credentials include ten years bench experience plus industry experience and intensive training, staff assessment witnessing, client assessment of auditor performance, and peer review.

Beware of the scams. Accreditation is a serious business founded by metrologists, for metrologists. If any calibration service says it is Guide 25 accredited just like A2LA or NVLAP, it is highly suspect. An accreditation audit by A2LA or NVLAP has value.

The Race

The race to keep accreditation a legitimate and coveted honor will only happen if everyone is informed and asks tough questions. Ask if the organization has mutual recognition with A2LA or NVLAP. If they lack such recognition, no A2LA test or calibration lab can use your service without a waiver (if granted).

Be proud to say you're accredited. If you use calibration services, ask the same questions. Ask if they have in place all the programs mentioned in this article.

What often happens is that some organizations will say they comply to Guide 25 without earning it. Such misrepresentation is common.

Once 17025 is in place, the landscape will change even more. As my passion for excellence in measurements runs deep, my conviction to always take the high road is never easy. I often am told that "the other guys were cheaper and they said they were accredited," only to find problems with them or worse, get deficiencies during an audit. Stay on top of this, people. Your measurement integrity starts here. Hold accreditation as the highest honor bestowed to a calibration lab.

Thomas G. Smith is president and CEO of National Calibration and Testing Laboratories (Golden Valley, MN), an accredited calibration laboratory he founded in 1984. He has been active in metrology for 25 years and sits on five standards-writing committees for metrology. He is a member of three working groups and a member of the measurement advisory council and accreditation council for the American Association for Laboratory Accreditation (A2LA). He can be reached via e-mail at tsmith@nctl.com.

Back to May/June Table of Contents