Uncertainty from sample preparation in the laboratory on the example of various feeds

Abstract

A way of uncertainty calculation from test sample preparation in the laboratory was presented on the example of feeds. The essence of the proposal lies in separating two components of results’ variability expressed as coefficient of variation CVp: analytical variability CVa (repeatability) and technical variability CVh corresponding with the inhomogeneity of a component. The law of Gauss’s error propagation was used. Analytical variability CVa was calculated from the range of duplicate analyses, following Nordtest Handbook. It was assumed that the coefficient of technical variability CVh is the measure of the standard uncertainty from sample preparation in the uncertainty budget of a method us+h, and expanded uncertainty Us+h for k = 2 (P = 95 %) can be easily calculated as Us+h = 2 us+h. Calculated uncertainties with uncertainties from sample preparation for loose feed and no ground premixture have increased from 4 to 125 % and were higher than analytical uncertainties calculated acc. to GUM during validation of methods. In the case of granulated feed mixture, the obtained uncertainties were similar. Grinding the premix results in lower uncertainties. Uncertainty from sample preparation should be taken into account in the uncertainty budget of a test method, especially in the case of inhomogeneity of tested materials.