Volume calibration of a 100 microliter micropipette

Project Description

The calibration of micropipettes was at that time a relatively recent capability for the majority of NMIs, so this lack of long term experience contributed to the observed dispersion in the results. The number of participating laboratories (6) was small and provided a small population for the statistical significance of the results.This was not very helpful for the needed conclusions.

Some lessons were learned from the experience gained and it was recommended at the time to repeat this comparison in the near future with more participants, in order to gather more information and experience about the calibration procedures.

During the EURAMET TC "Flow" meeting, held in Scotland in March 2010, it was agreed to start a comparison of a micropipette in order to verify if this experience were gained by the participants and it was proposed to use a more rigorous and detailed protocol in order to reduce possible errors in liquid handling.

Final Report 2012-05-10

The project has been completed and the report can be downloaded here>>

This comparison involved 10 laboratories at all, and lasted one year.
The initial and final results obtained by both IPQ and FORCE are consistent with each other. This demonstrates that the transfer standard had a stable volume during the entire comparison.
The values are very similar but since the dispersion is large (four laboratories with inconsitsent results) the median was the chosen estimator to calculate the reference value.
Inconsistency in some laboratories could be attributed (at least partially) on lack of experience in handling the micropipette. The comparison could be repeated with a new very detailed technical protocol which includes handling instructions and techniques.  Applying a detailed common and very detailed procedure could improve dramatically the performance of the laboratories.
The presented uncertainty budgets are very different but for the majority of the participants the two uncertainty components that have a major contribution to the final uncertainty, are the repeatability and the mass/balance.