Traceability, Measurement Uncertainty & Technical Evaluation

The concerning policy's are for Traceability, Uncertainty of Measurement and Technical Competence of Laboratory Technicians. The changes presented in the policies are in place to better the accreditation and better the laboratories through continuous improvement.

Policy 001 Traceability

In previous revisions, Policy 001 accounted for traceability and uncertainty and it was presented by the L-A-B Technical Advisory Group that we split the areas into 2 different policies and rightfully so. There is equal importance for both areas to address the requirements of ISO/IEC 17025, NACLA, APLAC and ILAC, not to mention the importance when making a measurement.

In the past several years, L-A-B has attended conferences, seminars and meetings and there appears to be some misconception on how traceability is to be identified and reported. We sat through a conference talk from the Director of the BIPM and he went through some misleading information from the exhibitors, the brochures and exhibit booths displayed at the conference. He went through some statistics and it was kind of alarming in that the exhibitors identified their "Traceable to NIST" services. Whether the exhibitors were selling equipment or selling a service of calibration or test. In most cases the exhibitors understand the traceability chain but their customers were not properly educated about the importance or the meaning of the service that was being provided.

The past 30 years, the service requested was that a calibration be traceable to NIST which is partially correct but not totally. In most cases the traceability chain does not stop with NIST because from their defined periods their equipment is to be realized by the BIPM. Which means the traceability chain does not stop with NIST it works through NIST and ends at the SI unit. The following was found on NIST's website (http://ts.nist.gov/Traceability/SupplMatls/suppl_matls_for_nist_policy_r...) pertaining to the evidence of traceability "Test report numbers issued by NIST are used solely for administrative purposes. Although they often uniquely identify documents that bear evidence of traceability, test report numbers themselves do not address the issues listed in II.B.1 above, and should not be considered as the sole evidence of traceability." Further II.B.1 identifies Per the VIM definition, traceability is established through "an unbroken chain of comparisons, all having stated uncertainties." In practical terms, this means having in place, for each link in the chain:

• a clearly defined particular quantity that has been measured
• a complete description of the measurement system or working standard used to perform the measurement
• a stated measurement result or value, with a documented uncertainty
• a complete specification of the stated reference at the time the measurement system or working standard was compared to it
• an ‘internal measurement assurance' program for establishing the status of the measurement system or working standard at all times pertinent to the claim of traceability
• an ‘internal measurement assurance' program for establishing the status of the stated reference at the time that the measurement system or working standard was compared to it.

With that as a base for the Technical Advisory Group to elaborate on, the following information was revised within L-A-B Policy 001

• Definitions added for traceability, traceability chain, uncertainty of measurement, calibration according to the VIM
• 5.3 - L-A-B prospective clients and accredited labs may use a State Weights and Measure lab that is recognized by the NIST Office of Weights and Measures (OWM) State Laboratory Program and has a current "Certificate of Measurement Traceability" issued by the Weights and Measures Division of NIST for the calibration of mass standards, including test carts. This certificate must be available during the assessment of the laboratory.
• 6.2 A statement of traceability to be on a certificate / report to consist of the following verbiage:
• 6.2.1 The calibrations / test within the certificate / report are traceable through NIST or another National Metrology Institute to the International System of Units (SI units).
• 6.3 The certificate or report is to provide evidence of actual traceability from the provider of the calibration i.e. calibration certificate number or unique identification of the certificate or report of all of the equipment utilized in the testing or calibration process.

Policy 001.1 Uncertainty of Measurement *New Policy*

Policy 001.1 is a new policy that stems from the previous revision of L-A-B Policy 001 which identifies the requirements for uncertainty of measurement. There are a few changes that have been made for the calibration, dimensional inspection and testing laboratories. For the calibration, dimensional inspection and testing laboratories a couple of changes have been made elaborating on the requirements of ISO/IEC 17025. The new policy identifies that the laboratory calculating the uncertainty of measurement shall apply a procedure for calculating the uncertainty for the measurement being made. Within ISO/IEC 17025 5.4.6.1 it identifies that the calibration or 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.

The Technical Advisory Group looked at this requirement and the experiences that they have went, whether from the standpoint of assessors or laboratories, and decided that this should address all measurements. More times than not, laboratories will send their equipment out for calibration than perform the calibrations internally. There are a couple of good reasons for this and it was designed to help the laboratories year in and year out when calculating uncertainty. In a majority of assessments we see that the laboratories may address their uncertainty once a year when their standards have been calibrated. And more times than not a lot of head scratching takes place and trying to remember what they did in the previous year to come up with final answer. The policy was put in place to help the laboratories and guide them through the calculations for the future reviews that take place.

The following information was revised from the previous L-A-B Policy 001

• Definitions added for uncertainty of measurement and calibration according to the VIM
• 4.1 Accredited laboratories of L-A-B that are performing calibration(s), dimensional inspection(s), and necessary tests whether commercially, captively, legal for trade or internally are required to apply a procedure for the estimation of the uncertainty of measurement.
• 5.4 When a calibration certificate or test report contains expanded uncertainty of measurement, a statement of the measurement result and the associated uncertainty must be accompanied by an explanation of the meaning of the uncertainty statement. ......An example of an explanation might be "The reported expanded uncertainty of measurement is stated as the standard uncertainty of measurement multiplied by the coverage factor k = 2, which for a normal distribution corresponds to a coverage probability of approximately 95.45%".
• 6.6 The uncertainty of measurement reported on a calibration / dimensional inspection certificate shall not be numerically lower than what is reported on the current scope of accreditation.
• Testing Laboratory requirements for uncertainty
• 7.4 The laboratory shall do a Needs Assessment for all tests within their Scope of Accreditation. The Laboratory shall define the categories and what actions will be necessary with regard to the reporting of uncertainty. The Needs Assessment shall group the tests into four (4) categories as defined below.
• 7.4.1 Type A Test - Qualitative or semi-quantitative tests are not numerical or are not based on numerical data (e.g., pass/fail, positive/negative, or based on visual or tactile or other qualitative examinations) estimates of uncertainty or other variability are not required.
• 7.4.2 Type B Test - A test performed to well-recognized test methods that specify limits to the values of the major sources of uncertainty of measurement and specifies the form of presentation of calculated results. These are defined in ISO/IEC 17025 Clause 5.4.6.2 Note 2.
• 7.4.3 Type C Test - Test methods based on published in international, national, regional or a technically reputable organization for which uncertainty is not defined in the method. For these types of test, uncertainty can be estimated using the standard deviation of laboratory control samples for more than 50 points. (This does not include laboratory-developed methods that require validation and are covered below).
• 7.4.4 Type D Test - Identify contributors of uncertainty and detailed uncertainty of measurement budgets calculated in accordance with published methods that are consistent with the ISO "Guide to the Expression of Uncertainty of Measurements".

Policy 010 Technical Competence Evaluation for Laboratories

This policy was revised to account for the laboratories ensuring that they cover the technical competence as part of their internal audit. Another reason why the policy was revised is to account for the laboratories performing onsite calibrations, dimensional inspections or testing. In conversations about the revision of the policy the Technical Advisory Group brought into consideration what other areas can this change help the laboratories improve the management system. A lot of ideas were talked about but one stuck out above the rest. Going through assessments, laboratories may struggle to address the requirements of performing evaluations of their personnel concerning effectiveness of the training. The change to this policy should help the laboratories address that particular area of the ISO/IEC 17025.

As an Accreditation Body, we have requirements to ensure that the laboratories performing onsite activities are covered during the assessment process. As we go through the preparation of the assessment, we work with the laboratory's to perform an onsite assessment of the technicians and that the assessor will have the opportunity to witness the calibrations at a customer's facility and a couple of days prior to the assessment or the day of the assessment something comes up that may out of the hands of the Accreditation Body or the laboratory being assessed, preventing the onsite witnessing. This particular circumstance puts the accreditation body in a bind attempting to meet the requirements of ISO/IEC 17011, NACLA, APLAC and ILAC.

To assist laboratories in meeting the requirements of this policy, they may use the same form that L-A-B assessors use. The Form 205.1 covers the requirements of section 5 and 4.13 of ISO/IEC 17025 and L-A-B policies for traceability, uncertainty of measurement and proficiency testing. The form can be supplied for the L-A-B accredited laboratories upon request to the L-A-B office.

The following was revised in Policy 010

• 4.1 All scope parameters performed onsite by the laboratory must be technically witnessed during an internal audit and onsite by the L-A-B technical assessor during the accreditation cycle. The laboratory is responsible for setting a schedule to assure the technical assessor is able to witness the given parameters onsite.
• 4.1.1 This schedule must be submitted to L-A-B for review.
• 4.1.2 Laboratories may prove through their internal audit that the onsite activities are covered and properly witnessed. With the necessary information of the internal audit, L-A-B may waive the need to perform an onsite witnessing of certain parameters on the Scope of Accreditation.
• 4.1.3 L-A-B reserves the right to require witnessing of any parameter whether in-house or onsite to confirm the technical competence of the laboratory at any time prior to, during and after the assessment
• 4.6 Laboratories that fail to meet the requirements of having the onsite parameters technically witnessed by the L-A-B Assessor onsite during an assessment or as part of the internal audit, the note from the Scope of Accreditation will be removed. The requirements of L-A-B Policy 012 will be enforced if the certificate / report are issued bearing the L-A-B Accredited Symbol is issued for the non identified onsite capability.
• 5.1.3. Laboratories are required to complete, and keep current, the L-A-B Form 001 - Traceability Tracking prior to the assessment taking place.
• 5.1.4. For all instruments calibrated in-house the following must be in place:

o Appropriate Environment
o Trained Personnel
o Reference Standards
o Procedure for the Calibration
o Procedure for uncertainty

Special Note on: Proficiency Testing

ISO 17043 has just been released; replacing ISO Guide 43 parts 1 and 2, and is now available for purchase from ISO and many other sources.
The complete review of our Policy 002 along with the associated documents and forms for compliance to this new standard is expected by the end of February. L-A-B will update client laboratories as needed of any changes affecting proficiency testing requirements and L-A-B policies and forms.