Theory of sampling - a mini seminar under the NKS project SAMPSTRAT

At an emergency situation a large number of matrixes can be contaminated and samples of these different matrixes will be colleted. These sample matrixes might be or often certainly are heterogeneous and in general more unevenly distributed than from nuclear test fallout or even the Chernobyl accident.

On basis of the reported data conclusions and remedial actions causing social and economical costs for the society are taken. Therefore the number of samples from each site, their size and further homogenisations is of great importance. In the case of an emergency situation the activities are generally high and the errors due to counting statistics are small.

We could also imagine a situation when a certain nuclear enterprise/activity has to close down or being prosecuted, based on sampling and analysis, for not following directives of discarding radioactivity in the environment. We therefore organized a seminar focusing on the above mentioned problems. The seminar covered several important topics such as an introduction to Theory of sampling (TOS), Lot heterogeneity and sampling in practice, Statistics for sampling in analytical chemistry, Representative mass reduction in sampling.

Case studies were presented such as Sampling of heterogeneous bottom ash from municipal waste-incineration plants and Sampling and inventories at Thule Greenland, which also illustrated the difficulties with Plutonium Inventory Calculations in Sediments when Hot Particles were present.