Comparison of measurement uncertainty estimation results based on duplicate samples of groundwater quality monitoring according to different experimental designs

In accordance with EU and Polish legal regulations the obtained results of chemical analysis of groundwater monitoring should include the information about the measurement uncertainty. Sampling process, transportation and analysis procedure – all impact the final outcome of measurement. In this paper, on the basis of duplicated samples taken during groundwater monitoring, various calculations of measurement uncertainty were compared via simplified balanced, full balanced and unbalanced experimental designs. The aforementioned estimations differ in terms of number of chemical analysis, which is reflected in the cost of monitoring. The details of results and the possibility of assessing particular components influencing the value of uncertainty derives from the number of undertaken analysis. Calcium, magnesium, bicarbonate and aluminum were investigated. The calculations were made using RANOVA program, applying robust statistics method. Conclusively, the simplified balanced design fulfills the necessary legal requirements at the lowest possible costs. The major drawback associated with this method is the total measurement uncertainty due to the lack of values of components. The full balanced experimental design provides sufficient information on certain components of measurement uncertainty. However, significant financial input is necessary. The unbalanced experimental design, which allows to estimate the contribution of particular components in the total measurement uncertainty, is more affordable and only one extra chemical analysis (for the pair of samples) is needed