Sampling and analysis of Waste Printed Circuit Boards (WPCBs) become of high importance because of the strong increase in the annual production of such waste and equally because of the value and criticality of the contained metals. As a result of human production, we could expect to know exactly the heterogeneity of constitution of WPCBs, but this is still far away from reality, and there is no effective data available due to lack of documentation and rapid obsolescence of such equipment, the lot composition itself being highly variable over time.

In the frame of a project focused on the WPCBs recycling, BRGM has conducted a set of experiments with the double objective to define an appropriate sampling and analytical procedure and to estimate the variability of such procedures for the determination of the content of six metals. Different operating conditions have been tested: three grinding sizes and three sample masses. For each case, the procedure has been repeated several times sufficient to consider statistics. A theoretical study has then been performed to deconvolute sampling and analytical errors from these statistics. The paper presents this approach and the techniques used to go back to the heterogeneity of such material. The analytical error is calculated for the different protocols allowing the experimental identification of the last stage total sampling error. Discussion is made about the possibility to estimate the intrinsic heterogeneity parameter at the different stages of the sampling procedure.