A Needed Paradigm Change for Environmental Soil Sampling at Urban Sites

Heterogeneous distribution of solid contaminant residues in surface soils creates an unique challenge for collecting soil samples yielding representative and reproducible results. Research over the past decade involving energetic and metallic residues indicates conventional grab sampling yields unreproducible, biased, results with poor precision non-representative of site conditions. Spatially, grab sampling yields information only about the point sampled. However, risk assessment and remedial decisions are based on contaminant distribution over an area. Multiple grab samples are collected for spatial coverage with the number often based on "expert opinion". Recent research suggests new field sampling and laboratory processing procedures encompassed in the Incremental Sampling Methodology (ISM) and incorporated into United States Environmental Protection (USEPA) Method 8330B for energetics and the update to Method 3050B for metals are necessary to overcome heterogeneity issues. The ISM approach is different from conventional grab sampling due to the focus on a spatially averaged result over a decision unit (DU) obtained by the collection of many increments using a systematic random sampling approach to form a single sample. Total precision error using ISM is typically <30%, whereas relative errors > 100% are observed between field splits and laboratory replicates with conventional grab sampling. Field splitting often employed prior to laboratory sample processing resulted in poor precision and is not recommended. Milling or grinding of soil samples is necessary to reduce the total error of the sample and to obtain acceptable precision, especially in situations where the mean contaminant concentration is expected to be near the action level or other regulatory value.