Millipore Sigma Vibrant Logo

Guideline for calculating your working tolerance

According to ISO 17025, every lab needs to calculate its working tolerance. Here are some hints for this procedure.

How can the user determine their own specific working tolerances? Here are a few tips...

To help determine the individual measurement-uncertainty estimation, every laboratory that works according to the ISO 17025 standard should prepare a control chart for each reference material and for every lab operator. When it comes to defining the lab’s working tolerance, the recommended procedure is to: 

  • Calculate the standard deviation of the measurements of the standard
  • Define the lab’s own error tolerances that are to be strived for.
Here the user should consider the con¬fidence in¬ter¬val that is usually used. Two-fold standard deviation yields a 95% confidence in¬terval, and three-fold standard deviation a 99% confidence interval. The defined confidence interval should be marked on the control chart as the upper and lower limit.

The standard should be measured regularly, if not on a day-to-day basis. The instructions can be taken from the corresponding standards or the internal specifications. Wherever applicable, in the case that the method covers a larger measuring range, two or more standards in the range of the normal results should also be determined. 

In the Spectroquant test kits, we state the accuracy of the respective method on the Certificate of Quality and in the package leaflet. This can be used to facilitate the estimation of the working tolerance. The working tolerance naturally also depends on the optical path length of the cells that are used.
The accuracy is calculated on the basis of the mean value of the 95% confidence interval of the re-spective test; this has been calculated over many years of experience along with the specified blank error. The blank error is stated in the batch certificate of each batch of the respective test kit for the reference cell.

These reference materials can be used to check all photometric methods (both standards and test kits). The reference materials can however, also be used for non-photometric methods. In this case, the user must perform their own measurement-uncertainty estimation. Guidelines or working tolerances of the type that we offer for our test kits are not available for other methods. The same applies for test kits supplied by other manufacturers.