Why laboratory duplicates often fail to match

The ratio between particles of different sizes and densities in a soil sample has a strong influence on the resulting contaminant concentration.

The implication of this microscale heterogeneity is that the concentration of any soil sample analysis depends on the ratio between the small and large particles in the analytical subsample. Unless measures are taken to prevent it, two analytical subsamples from the same sample (e.g., laboratory duplicates) will vary in their proportions of larger particles carrying lower contaminant loadings vs. smaller particles with higher loadings, causing different subsamples to have different concentration results. The situation is exacerbated by what happens to soil samples during collection, shipment to the laboratory, laboratory handling, and subsampling. All these activities promote segregation or stratification of soil samples by particle size and density, making it likely that subsamples are biased for or against certain particles, biasing the concentration results away from the true mean for the sample.