Metrology is the science of measurement and is is divided into:
- The definition of internationally accepted measurement units.
- The use of these units in practice.
- Traceability that links measurements with accepted reference standards.
- Quantifying unknowns.
Since traceability and quantifying unknowns are particularly important to understand for our purposes, let’s turn to the dictionary: the International Vocabulary of Basic and General Terms in Metrology on page 45 defines metrological traceability as the “property of a measurement result whereby the result can be related to a reference through a documented unbroken chain of calibrations [through an established calibration hierarchy], each contributing to the measurement uncertainty”.
Traceability then, is what allows a measurement result taken in one place and time to be compared with a measurement taken at another place and time.
What begins as a gravimetric pipette calibration from a certified test balance in a controlled laboratory environment, can be traced, through proper certification of standards, equipment and test environment, all the way back to the International Standard (IS), through N.I.S.T. (National Institutes of Standards & Technology). When this calibration traceability chain is compromised (which is generally the case when onsite pipette service providers transport their equipment) the results are by definition invalid and unreliable.
Every measurement known to man has a level of uncertainty associated with it. The lower the uncertainty, the better the measurement quality, and the better the accuracy and reliability of the measurement process. Conversely, high measurement uncertainty signifies a poor measurement process, with a high “doubt” about the measurement result.
For example, if you weighed yourself on a truck scale, which is designed to weigh heavy objects up to 80,000 pounds in increments of 500 pounds or so, you would expect a high measurement uncertainty because the system is simply not geared to be accurate at weights in the 150-200 pound range.
Furthermore, imagine a poll on two political candidates running for office: candidate 1 has a 55% to 45% lead over candidate 2, but there is a ±5% “margin of error” associated with the results (this is the poll’s “measurement uncertainty”). In other words, the opinions measured in this poll may be off by as much as 5% in either direction. When you do the math, you realize this poll could reflect anything from a dead heat to a 20% landslide!
Because of the measurement uncertainty, neither your weight on this truck scale nor the results of this poll can be trusted.
Measurement uncertainty has important consequences for determining a pipette’s calibration accuracy, particularly when that uncertainty is unknown, inaccurate or known but high. In a controlled, metrology-driven calibration laboratory like that of TTE Laboratories, measurement uncertainties are identified and controlled and as a result, are among the lowest in the industry.
A key aspect for achieving meaningful and reliable pipette calibrations is to isolate measurements from external influences that might increase variability and effect measurement accuracy. For pipettes, these influences include temperature, humidity, vibration, static, air currents, and even the effects of thermal transfer, where the pipette holder’s body heat is transferred to the pipette.
Another central element of a valid calibration service is in the detail of the calibration report or certificate. Such details include client information, the unit under test, test conditions, services provided calibration data, statistical analysis of results vs. tolerances, measurement uncertainty, our reference procedure and interpretation of results and calibration is supporting documentation for both tasks and administration. Tasks are documented with respect to measurement and calibration procedures, and to the data generated and collected during those processes. Administration is documented with respect to identification of equipment, certification of measurements and related uncertainties, and other informational elements.