Prøvetakingsstrategi – prøvetaking av masser langs vei

In Norway, sampling along a road has traditionally been done by collecting discrete samples. This is not a scientifically based method and will most likely give a poor image of the contamination in the given area. A more suitable method is called multi incremental sampling (MIS) or incremental sampling method (ISM). This method involves collecting a large amount of samples/increments (>30) which are gathered to a composite sample. The samples are collected from a given area where an average concentration of one or more substances is needed (decision unit, DU). When comparing uncertainty associated with sampling method, sample treatment before analysis, and the analysis itself, the sampling method represents the biggest error. Choosing the right sampling strategy is therefore important. MIS/ISM is based on the sampling strategy developed by Pierre Gy. Most acknowledged sampling strategies are based on Gy’s theory. Gy’s theory mostly concerns obtaining a representative sample by for instance collecting large enough amount and mass of samples, and using the correct equipment. To test the hypothesis that MIS/ISM produces a better image of the contamination than discrete sampling, the methods were compared by using them to describe the contamination along a highly trafficked stretch of road. In accordance with MIS/ISM the area was divided into three DUs. In each DU, 60 incremental samples were collected to one composite sample (three replicates). For discrete sampling, 10 samples were collected from each DU. The analysis performed on the samples was a standard package consisting of benzene, toluene, ethylbenzene and xylenes (BTEX), aliphatics, aromatics, polycyclic aromatic hydrocarbons (PAH) and metals. The results from the experiment showed that the analytical results from the composite samples (ISM/MIS) had lower standard deviation than the discrete samples. This shows that ISM/MIS gives a more precise result than discrete sampling. When collecting samples from an area, it is more likely that an ISM/MIS sample will give an accurate result than several discrete samples. ISM/MIS is also a more economic method, as it requires fewer samples to be analyzed to provide a satisfying average of a contaminant over an area than using the discrete sampling method. It is concluded in this report that ISM/MIS is the recommended method for soil sampling along roads.