OCENA WPŁYWU ZMIANY PRÓBOBIORCY NA NIEPEWNOŚĆ ZWIĄZANĄ Z OPRÓBOWANIEM W MONITORINGU WÓD PODZIEMNYCH

Abstract

ESTIMATION OF SAMPLER INFLUENCE ON UNCERTAINTY ASSOCIATED WITH SAMPLING IN GROUNDWATER MONITORING

Abstract. According to European Union directives (RDW, 2000; DWP, 2006) the identification and determination of uncertainty – also this associated with sampling, preservation and transport of samples – is an important part of the overall groundwater monitoring effort. In this work, the influence of sampler on the uncertainty estimation is taken into account. Two independent people (sampler J and K) were taken normal and duplicate samples from Królewski spring using the same sampling protocols. The same analytical method was used – ICP-MS (PN-EN ISO 17294-2:2006, PN-EN ISO 17294-1:2007). All samples were analyzed in hydrogeochemical laboratory which has implement extensive quality assurance and analytical quality control programs. Calculation of uncertainty and its components for this two operators has been shown in details for iron, Fe (limit of determination: 0.01 mg/l). To identificate the outliers, scatter plot, chart of differences and range chart were used (according to NORDTEST, 2006). To calculate the total variance separated into three parts (geochemical, sampling and analytical variance), standard (u), expanded (U) and relative (U’) uncertainty for each of these components, ROBAN software was used. In both sampling capaigns the geochemical variance is dominating in the monitored well. For J campaign, sampling uncertainty amounts 13% (with average iron concentration 775.69 μg/l) and total uncertainty is 50.42%. (± 391.10 μg/l). For K campaign, sampling uncertainty is smaller – 4%(with average iron concentration 890.18 μg/l) and total uncertainty too – 30.32% (± 269.90 μg/l). In both cases, estimated total uncertainty has not influence on estimation of groundwater chemical status of Królewski spring.