Volume 5 · Issue 11 · November 2000

Contents

ILAC Conference Focuses on ISO/IEC 17025
36 Laboratory Accreditation Bodies Sign ILAC Arrangement

On November 2, 2000, 36 laboratory accreditation bodies from 28 economies worldwide signed an "arrangement" in Crystal City, VA, that will promote the acceptance of technical test and calibration data for exported goods.

From October 30 to November 4, 2000, the International Laboratory Accreditation Cooperation (ILAC) held a technical conference and then a General Assembly meeting, which was highlighted by the signing of the ILAC mutual recognition arrangement (ILAC Arrangement).

The ILAC Arrangement, which enters into effect January 31, 2001, means that a product tested by a laboratory accredited by a signatory to the Arrangement in one country will be accepted and promoted by all the other signatories in their home countries. For companies and other organizations involved in the export of products subject to testing, this is a major step towards reducing or eliminating the need for retesting of products when they arrive in an importing country.

"For many years, the retesting of goods by an importing country has been considered a major technical barrier to trade," noted Dr. Belinda Collins, Chair of ILAC and Director of the Office of Standards Services at the National Institute of Standards and Technology (NIST).

"Since the mid-1970s, the World Trade Organization has identified such technical barriers as a major concern to world trade. Such barriers can not only add significantly to the cost of goods entering a country, but they can delay and in some cases prevent the goods from being accepted by foreign markets."

"The ILAC Arrangement will provide a technical underpinning to international trade by promoting cross-border stakeholder confidence and acceptance of accredited laboratory data," emphasized Peter S. Unger, President of the American Association for Laboratory Accreditation (A2LA) and Chairman of the ILAC Arrangement Management Committee, who signed the Arrangement on behalf of A2LA. "Until now, there has been no international mutual recognition agreement in lab accreditation, which has been a hindrance for some types of international trade."

"ILAC has been working to overcome these technical barriers for the last two decades by encouraging the development of regional recognition arrangements culminating in today’s global recognition arrangement among representative bodies in each country," added Dr. Collins. "This will facilitate the acceptance of goods already tested by an accredited laboratory. Thus, goods tested in one country should enjoy easier access to foreign markets participating in the Arrangement."

What the Arrangement Means

The key to the Arrangement is the developing global network of accredited testing and calibration laboratories that are evaluated and recognized as being competent by specific authorities known as laboratory accreditation bodies. Many of these bodies, which are located in many economies, participate in ILAC.

The 36 signatories listed below in Table 1 are members of ILAC, which consists of accreditation bodies that have been peer-reviewed and shown to meet ILAC’s criteria for competence. A goal is to have all ILAC members become signatories to the Arrangement, as well as to provide outreach to those bodies that are in the process of becoming ILAC members.

The ILAC Arrangement is based on ILAC guidance for ISO/IEC 17025 implementation that was finalized in the ILAC General Assembly meeting on November 2, 2000. Under the Arrangement, all laboratories accredited to ISO/IEC Guide 25:1990 must be operating by the end of 2002 in conformance with the requirements of ISO/IEC 17025:1999, General requirements for the competence of testing and calibration laboratories, which replaces Guide 25. Accredited labs must have been through a surveillance, assessment or re-assessment by their accreditation bodies to confirm compliance (for more information about ISO/IEC 17025, see "ISO/IEC 17025: The Standard for Laboratory Competence", THE OUTLOOK, October 2000).

The use of ISO/IEC 17025 as the basis for laboratory accreditation also eliminates the use of several standards and guides in the marketplace. Until ISO/IEC 17025 was finalized, the United States and several other countries accepted use of Guide 25, but only to a limited degree, while the European Union (EU) and other countries made use of a European Norm (EN) that provided calibration and testing laboratory competency requirements. National standards were applied in a few other countries, which made acceptance of laboratory results more difficult when products moved between countries where different standards or guides were in use.

A major development was the simultaneous adoption of ISO/IEC 17025 as an EN standard when the EU member states voted to approve the laboratory requirements standard in late 1999. There is also movement to adopt 17025 as a US National Standard.

On October 30, 2000, a formal proposal was posted on the Quality Management, Environment Management, Dependability and Statistics (QEDS) web site calling for the members of the American Society for Quality’s Z1 Subcommittee on Quality Management to vote to adopt ISO/IEC 17025 as an American National Standard. The proposal is likely to receive approval, particularly after the signing of the ILAC Arrangement by three US accreditation bodies.

The ILAC guidance was strengthened during the conference, which was attended by more than 350 laboratory professionals from laboratories, accreditation bodies and government agencies in more than 60 countries worldwide. The guidance received consensus approval in the General Assembly meeting. Agreement was reached by the General Assembly on language addressing concerns with the guidelines for appeals of ILAC decisions not to grant Arrangement signatory status to a laboratory accreditation body and requirements for the impartiality of accreditation bodies.

Impartiality Without Obstruction

A critical issue that needed to be resolved was how to prevent the impartiality requirements from hampering national governments that may have one set of operations providing lab accreditation services and another set that provides high-level calibration and testing measurement services, such as official measurements for metric and English units for weight, length and volume.

The impartiality guidelines were revised to make clear that the laboratory accreditation operations of an organization must not engage in providing calibration and testing services that would put the organization in competition with the laboratories it accredits–a conflict-of-interest scenario–but permitting separate operations of the same organization to provide both accreditation and laboratory services when they are independent of each other and thus impartial.

Many countries have national metrology institutes (NMIs) that serve as "the calibration laboratory for the calibration laboratories in a given country", and the NMIs are government-operated entities that co-exist with government-operated accreditation services. The Arrangement provides guidance that will permit these co-existing entities while ensuring that the operations are independent and impartial.

The ILAC Arrangement builds upon existing or developing regional arrangements established in the Americas, the Asia-Pacific region, Europe and Southern Africa. The bodies participating in these regional arrangements are responsible for maintaining the necessary confidence in accreditation bodies from their region that are signatories to the new ILAC Arrangement.

"This is an important change for the laboratories and their customers, who faced a patchwork of arrangements involving some countries but not others," affirmed David Stanger, a member of the Board of Governors of the Union Internationale des Laboratoires Independ-endants (UILI), head of its delegation to ILAC 2000 and Chairman of the ILAC Laboratory Liaison Committee.

"Indeed, it comes down to the fact that manufacturers and exporters using labs in one country can look forward to greater ease in trading, since the accreditation bodies in countries importing products tested in the export country labs will accept the test results as equivalent to theirs and will encourage companies in their countries to also accept the results."

"Now that the Arrangement is in place, the next crucial step is for governments to take advantage of this Arrangement by using it to further develop or enhance trade agreements," explained Mike Peet, Chief Executive Officer of the South African National Accreditation System and Chair of the ILAC committee that developed the new Arrangement.

"There is now a firm foundation in place for manufacturers and exporters that have their goods tested by accredited laboratories to enjoy greater market access, less costs associated with retesting and greater overall competitiveness in global markets."

Another important step that is already underway involves government acceptance of the results from accredited laboratories. "Regulatory agencies around the world, including in the United States, are beginning to accept the results from testing and calibration laboratories that are accredited by bodies such as the ILAC members without direct government review, including results from labs in other countries," said Jeffrey Horlick, a Physicist with the National Voluntary Laboratory Accreditation Program (NVLAP) at NIST. He gave as an example acceptance by regulatory agencies in Australia and New Zealand of results from laboratories accredited by NVLAP.

The Issue of Measurement Uncertainty

A new issue with ISO/IEC 17025 for both testing laboratories that used Guide 25 or the EN standard and those that are new to laboratory accreditation are the requirements relating to "measurement uncertainty".

Particularly among testing laboratories, where several different pieces of testing equipment might be used to conduct a single measure and the tests for each product may involve different configurations of equipment, what this means is itself uncertain. It is one of the reasons that many laboratory accreditation bodies are recommending a "pragmatic, incremental approach to measurement uncertainty", according to Unger.

The requirements for estimating measurement uncertainty are contained in Subclause 5.4.6, Technical Requirements–Test and Calibration Methods and Method Validation, Estimation of Uncertainty of Measurement, which is reprinted below as Figure 1. One of four technical sessions of the ILAC conference was devoted solely to these requirements, which arose in other sessions as well.

At first glance, the requirements of 5.4.6 may indeed seem overwhelming for laboratories that have never had procedures to estimate the uncertainty of their measurements, because the assumption might be that a laboratory will need to produce detailed calculations of the amount of uncertainty there is in every type of measurement the lab provides to its clients.

However, it is important for laboratories and the accreditation bodies to consider that 5.4.6 calls for labs to estimate measurement uncertainty, with the level of the calculations dependent on the nature of the test method involved and whether those methods can be measured for uncertainty in a valid way.

In a presentation in the technical session on measurement uncertainty, Philip Stein, Principal Scientist and President of P.G. Stein Consultants, indicated that laboratories should not become overly concerned with the measurement uncertainty requirements of ISO/IEC 17025, because 5.4.6 does not specify a high degree of specificity in reporting measurement uncertainty, only an estimation.

"At a high level of calculation, you can get a pretty precise measure of uncertainty, but that isn’t necessary for lab accreditation," Stein explained. "You simply need to understand the range of uncertainty your measures can produce and be able to express that range. For ISO/IEC 17025 conformance, you may be able to get by with limited measurements for testing. All you–and the accreditation body–need to know is how good, how variable, your measurement is. If you can do this without extensive calculations, do so."

Stein advocated the use of control charts for a testing laboratory’s repeated measurements, with the variation on a control chart providing the uncertainty measurement the lab needs.

Horlick told THE OUTLOOK that part of the problem is in the use of the term "measurement of uncertainty", which has a misleading connotation of inaccuracy when most measurements are unlikely to be highly precise except when required (e.g., for parts ordered by NASA for a space shuttle). "What you are really doing is measuring how precise a measurement is going to be when performed repeatedly, not occasional variations," remarked Horlick.

Horlick gave as an example the measuring of 10-pound sacks of potatoes. "A company that sells potatoes needs to make sure that the potatoes in each bag weigh at least 10 pounds to avoid Federal Trade Commission penalties, but the company wants to avoid having much more than 10 pounds in each bag, since that reduces revenue and increases overhead costs (e.g., transportation).

"However, it is impossible to get exactly 10 pounds of potatoes in every bag even if you weighed each bag and made corrections when there is variance, so the company is going to have variation that could range from plus-or-minus half an ounce to more than an ounce. If the company can get a rough estimate of what it can do to make sure each bag has at least 10 pounds of potatoes but with minimal variation above 10 pounds, that estimate represents its measurement uncertainty.

"While testing differs from measuring potatoes, the concept is the same: Do you know how close to expected results your measurements of a product or material are going to be?"

In another presentation at the uncertainty measurement session, Dr. Steve Ellison, head of the Statistics and Quality Systems group at the UK’s Laboratory of the Government Chemist, described uncertainty as "a possible range of predicted outcomes. The intent of estimating the uncertainty of your measurements is to offer some sense of assurance of accuracy to a laboratory’s work. With accuracy, the question you are asking is, ‘Are we getting close to the measurement we expect.’"

THE OUTLOOK will provide further coverage of the measurement uncertainty issue as accreditations begin to take place, providing better information on how laboratories are satisfying the ISO/IEC 17025 requirements. For further details on the ILAC Arrangement and about ILAC, visit ILAC’s web site (http://www.ilac.org).

If you missed the ILAC 2000 Conference–the Technical Conference for Testing and Laboratory Professionals–you can still obtain a copy of the ILAC 2000 Proceedings, which is a collection of papers from speakers summarizing their presentations. For information on the ILAC 2000 Proceedings, contact INFORM, publisher of THE OUTLOOK, by phone (703-680-1436) or by e-mail (INFORMintl@erols.com). Buy extra copies so that others in your organization can have a personal reference.

Table 1.    Accreditation Bodies in ILAC Participating in the Arrangement

Country/ Econ	Accreditation Body	Scope
Australia	NATA	Testing & Calibration
Belgium	BELTEST and BKO/OBE	Testing & Calibration
Brazil	Directoria de Credenciamento e Qualidade/Instituto Testing & Calibration Nacional de Metrologia, Normalizacao e Qualidade Industrial (INMETRO)	Testing & Calibration
Canada	SCC	Testing & Calibration
China, People’s Republic of	CNACL	Testing & Calibration
Czech Republic	Czech Accreditation Institute, o.p.s. (CAI)	Testing & Calibration
Denmark	Danish Accreditation (DANAK)	Testing & Calibration
Finland	FINAS, Finnish Accreditation Service Centre for Metrology and Accreditation	Testing & Calibration
France	Comite Francais d’Accreditation	Testing & Calibration
Germany	Deutsches Akkreditierungssytem Prufwesen (DAP)	Testing
	Deutsche Akkreditierungsstelle (DACH)	Testing
	Physikalish-Technische Bundesandstalt (PTB)	Calibration
	Deutsche Akkreditierungsstelle für Technik (DATech)	Testing
Hong Kong, China	HKAS	Testing & Calibration
India	NABL	Testing & Calibration
Ireland	The Irish National Accreditation Board (NAB)	Testing & Calibration
Italy	Sistema Nazionale per l’Accreditamenta (SINAL)	Testing
	Servizio de Taratura en Italia (SIT)	Calibration
Japan	JAB	Testing
	JCSS	Calibration
	JNLA	Testing
South Korea	KOLAS	Testing
Netherlands	Dutch Accreditation Council (RvA)	Testing & Calibration
New Zealand	IANZ	Testing & Calibration
Norway	Norwegian Accreditation	Testing & Calibration
Singapore	SAC	Testing & Calibration
South Africa	South African National Accreditation System (SANAS)	Testing & Calibration
Spain	Entidad Nacional de Acreditacion (ENAC)	Testing & Calibration
Sweden	Swedish Board for Accreditation and Conformity Assessment (SWEDAC)	Testing & Calibration
Switzerland	Swiss Accreditation Services (SAS)	Testing & Calibration
Chinese Taipei	CNLA	Testing & Calibration
UK	United Kingdom Accreditation Service (UKAS)	Testing & Calibration
USA	A2LA	Testing & Calibration
	NVLAP	Testing & Calibration
	ICBO-ES	Testing
Vietnam	VILAS/STAMEQ	Testing & Calibration

Back

Figure 1.    5.4.6, Estimation of Uncertainty of Measurement

5.4.6.1    A calibration laboratory, or a testing laboratory performing its own calibrations, shall have and shall apply a procedure to estimate the uncertainty of measurement for all calibrations and types of calibrations.

5.4.6.2    Testing laboratories shall have and shall apply procedures for estimating uncertainty of measurement. In certain cases the nature of the test method may preclude rigorous, metrologically and statistically valid, calculation of uncertainty of measurement. In these cases the laboratory shall at least attempt to identify all the components of uncertainty and make a reasonable estimation, and shall ensure that the form of reporting of the result does not give a wrong impression of the uncertainty. Reasonable estimation shall be based on knowledge of the performance of the method and on the measurement scope and shall make use of, for example, previous experience and validation data.

5.4.6.3    When estimating the uncertainty of measurement, all uncertainty components which are of importance in the given situation shall be taken into account using appropriate methods of analysis.

Back

If you liked this article, subscribe now.