Alignment of terminology: calibration and verification
"The more critical the decision, the more critical the data. The more critical the data, the more critical the measurement.
–NASA Reference Publication 1342
ISO 17025 was initially released in 1999. In 2005, it underwent minor revision to align better to ISO 9000:2000. In 2010, the standard was up for review again and rather than voting for revision, the vote was to confirm the standard as-is for another five years. In attending the 2012 ILAC joint annual meetings, it is becoming my opinion that ISO 17025 needs to be revised. This is the first of a series of blog posts that I will be making to demonstrate why the revision needs to be made.
Some definitions in ISO 17025 contradict the VIM
In order to understand and interpret a requirements document correctly, it is important to know pertinent definitions of terms and concepts. ISO 17025 was written referencing definitions in the International Vocabulary of Metrology (VIM). Since 1999, the VIM has undergone significant change and refinement in its definition of terms and concepts associated with metrology. The VIM was been revised in 2007, and had minor corrections applied in 2012. As a result, there are many terms that are used in ISO 17025 that have different and contradictory definitions than are currently understood in the VIM.
Calibration
The first term to consider is calibration itself. The VIM defines calibration as “operation that, under specified conditions, in a first step, establishes a relation between the quantity values with measurement uncertainties provided by measurement standards and corresponding indications with associated measurement uncertainties and, in a second step, uses this information to establish a relation for obtaining a measurement result from an indication.”
This is a very long way to say that the process of calibration is to compare a known reference to a device under test. Every test and measurement instrument can be calibrated using this definition. It is possible to calibrate a standard resistor, a gage block, or a digital multimeter using this definition. When ISO 17025 was written, it was this basic definition of calibration that was largely used, where for each test point a measured value and corresponding uncertainty of measurement was given.
There are two very important things that the act of calibration does not address: The first is adjustment, which is referred to as “adjustment of a measurement system” in the VIM and means just what it sounds like. Many people confuse calibration for adjustment. It is possible to calibrate all test instruments, it is not possible to adjust all test instruments.
For example, it is not possible to adjust a standard resistor, but it is possible to compare a standard resistor to a reference standard resistor to determine a measurement value and associated uncertainty. Therefore a standard resistor can be calibrated, but not adjusted. For the case of a digital multimeter, it is possible to calibrate it as per the VIM definition.
Verification
The second thing that calibration does not address is that for most customers, this is only performing half the job. Most of our customers need to know if the multimeter is meeting a technical requirement such as whether all of the measured values obtained during calibration are within the published specifications for the instrument.
The act of providing objective evidence that a given item fulfills specified requirements is defined in the VIM as verification. Again, for the case of a standard resistor there are usually no published specifications, one uses the measured value and associated uncertainty of measurement, so it is possible to calibrate a standard resistor, but not adjust a standard resistor or to perform verification.
Section 5.10.4.2 of ISO 17025 discusses making a statement of compliance with a specification. Again, by the VIM definitions, this is not calibration, but is a second step of using the calibration data to perform a verification. ISO 17025 does not differentiate what should occur between these two very important processes. The result? We have to provide calibration reports with measured values and associated uncertainty and largely avoid the process of verification. But many of our customers may not care about the calibration data at all, and really want to verify that the instrument meets a particular specification.
The Solution
The solution for this is to re-write ISO 17025 to address the process of calibration and have an additional section to address verification when a customer requests this service during the contract review process. How is this accomplished? We all need to start letting our accreditation bodies, our National Measurement Institutes, and national standards writing bodies (example: ANSI in the USA) know that it is time to revise ISO 17025.