Everyone in the quality profession has heard the term “NIST traceable.” Having calibration traceability to the National Institute of Standards and Technology (NIST) is desirable for most measurement devices. It is also enshrined as a requirement in some regulatory documents. Unfortunately, NIST traceability does not ensure measurement quality. Here’s why.
NIST, a U.S. government organization based in Maryland, is the official keeper of the flame for the highest level of measurement accuracy. If you want to know how good your thermometer is, you can send it to NIST, and it will compare it to the very best temperature standards (“calibration” is simply the comparison of one device to another) and send it back with a report. Of course this isn’t practical; NIST cannot calibrate the millions of thermometers that are in daily use, and the service is expensive.
This is where the concept of traceability comes in. If you have one thermometer that has been calibrated at NIST, it can become your “standard,” and you can calibrate all of your other thermometers with it. All of these thermometers will have NIST-traceable calibration. This batch of thermometers can be used to calibrate other thermometers, which can in turn be used to calibrate still more thermometers. With proper documentation, all of these thermometers will have NIST-traceable calibration. Traceability requires an unbroken chain of calibrations that goes all the way back to NIST. The chain could be several calibrations long or a hundred calibrations long; it doesn’t matter, as long as the chain is unbroken.
Now for the bad news: NIST traceability does not necessarily provide a reliable and high-quality measurement. Remember, calibration is only a comparison between two devices. It’s possible to calibrate a $2 thermometer against the best thermometer in the world, but it’s still a $2 thermometer.
Imagine this scenario: You are responsible for the proper operation of a stability chamber that is actively in use. After 12 months of operation, you remove the temperature sensor that controls the chamber environment and have it calibrated. The calibration report shows the thermometer to be out of specification. Now what? In the best case, the deviation is small, and a backup thermometer functioned properly for 12 months, providing quantitative data regarding the actual chamber temperature. In the worst case, someone will have to analyze the potential effect of “out of control” temperature on the products inside the chamber. If the effect is significant or cannot be determined, this can be costly, setting back a stability study by several months.
Now for the good news: You can prevent measurement disasters. If you are responsible for a temperature measurement, start by selecting the right thermometer for the job. If you’re not sure how to do this, consult a metrologist in your organization, or contact a reputable vendor of thermometers for technical assistance. Set up a calibration program for this thermometer (and make sure your calibrations are traceable to NIST). Finally, consider using a second, independent thermometer for a backup measurement.
Returning to NIST traceability, there are other issues to consider. First, all calibration documents should be in a format that is generally acceptable. ISO/IEC 17025 defines the key elements of an acceptable calibration document. In fact, this is a global standard for quality systems of testing and calibration laboratories. If you are outsourcing calibration, consider using a vendor that is accredited to this standard. Visit the sites of the American Association for Laboratory Accreditation (A2LA) and the National Voluntary Laboratory Accrediation Program (NVLAP) to learn more and to find firms that have accreditation.
As you might imagine, other countries also have national metrology institutes (NMIs) that serve similar functions to NIST. In Canada the equivalent organization is NRC. In Mexico it is CENAM. It is entirely possible that organizations located in different countries might request calibration traceability to their own NMI. This can be particularly challenging if your own organization has operations in more than one country. But there is more good news. Many NMIs “recognize” each other. This recognition is formally organized by the Bureau International des Poids et Measures (BIPM). It does not necessarily mean that your company or your customers will freely accept calibration documents traceable to any of the BIPM signatories. However, if you see any logistical advantages, you may be able to modify your quality system to allow for this.
I have used the example of a thermometer for simplicity. NIST is the keeper of many standards—pressure, time, frequency, voltage, humidity, and many more. All traceability issues remain the same, regardless of the measurement parameter.