In his April “Product Safety” column in LD+A, Jerry Plank argues that trustworthy data on product performance starts with good lab procedures. An excerpt follows.
There was a time when anyone with a little electrical knowledge could conduct a temperature test on a luminaire and determine whether or not the potential for creating a fire and/or electrical shock hazard existed. In today’s litigious and overly regulated society, operating a laboratory for product safety testing has become extremely complicated and wrought with pitfalls.
Adding to the obvious technical complexity of conducting product safety tests, an international standard ISO/IEC 17025 has been adopted by all Nationally Recognized Testing Laboratories (NTRLs) titled, The General requirements for the competence of testing laboratories and calibration laboratories. The rationale for implementing ISO/IEC 17025 is to create an international structure whereby data collected globally can be relied upon as being accurate and reproducible, regardless of where it was collected.
The World Trade Organization (WTO) has promoted ISO/IEC 17025 as a means to remove technical barriers to trade; however, ISO/IEC 17025 is only valid if all laboratories collecting data comply with all applicable sections of the standard.
For a product testing lab to be fully ISO/IEC 17025-compliant and accredited, it must be audited by an accredited firm that is a signatory to the International Laboratory Accreditation Cooperation (ILAC). ILAC signatories include firms such as A2LA, ANAB, NVLAP and others.
A loophole does exist where NRTLs (who are not accredited or a signatory to ILAC) are authorized by OSHA to assess third-party or in-house manufacturer’s laboratories to ISO 17025; however, there is no oversight by an ILAC signatory that all ISO/IEC 17025 clauses are in compliance. Unless a product safety laboratory has been assessed to ISO/IEC 17025 by an ILAC signatory, an inequity exists, and there’s the potential for data that is not universally accepted.
CHECKING THE DATA
Let’s look at some laboratory basics that everyone reviewing test data should know. Data that is used to determine compliance to an UL/ANSI safety standard must irrefutably show that the product under evaluation meets with the intent of the appropriate clause.
The data recorded for a product safety program shall be accurate, which means that the equipment, facility, power systems, employees and sample preparation have written guides and specifications, such that data taken will be the same every time the test is run regardless of the laboratory conducting the test.
It is a recommended practice to conduct “round robin” or “audit” testing between different laboratories to ensure that the data collected is within accepted limits. Differences noted during a round robin or audit exercise will help a laboratory pinpoint why the differences in data existed and what corrective action is required.
Another key detail when collecting data is to conduct a test for which the data is reproducible. Reproducible data means that at any time in the future the test can be rerun on another or the same product sample, and the data collected will be identical or within accepted tolerances. Differences in data on round robin or audit tests may expose different levels of measurement uncertainty where the total of all tolerances of each aspect of a test could contribute to different data levels.
Due to ISO/IEC 17025 being implemented in the U.S., accredited laboratories can provide data worldwide without reservation or a fear that the data will be rejected. For firms that distribute globally, having an appropriately accredited laboratory by ILAC will prove to be a powerful asset.