Lassen, P. & Larsen, M.M. 2013. Report on the HELCOM PLC6 intercalibration. Aarhus University, DCE – Danish Centre for Environment and Energy, 114 pp. Technical Report from DCE – Danish Centre for Environment and Energy No. 27 http://dce2.au.dk/pub/TR27.pdf
This report presents results from the PLC6 intercalibration on metal and nutrients in fresh water and waste water. The intercalibration was performed in ordre to evaluate the analytical quality of results reported to HELCOM. 17 laboratories participated in the intercalibration.
In general the analytical quality is good and comparable between the laboratories with a few exceptions which can be explained by other means. See below.
NH4-N has high coefficients of total variation (CV(R)) for both fresh water and waste water. This does not necessarily reflect the analytical quality but is more likely to be caused by instability of the component which makes it difficult to include this component in intercalibrations like this. The instability was verified for fresh water but not for waste water in the control analysis (see section 2.2). However, if NH4 should be included in be included in future intercalibrations it would probably be necessary to prepare NH4 samples separately as a synthetic sample.
For PO4-P and P-total in fresh water, the coefficients of total variation (CV(R)) are high but this is an artefact to low concentrations. The corresponding absolute values, reproducibility, S(R), are 0,009-0,014 mg/L which is acceptable. For fresh water sample C the recovery of PO4-P is far too high. An error in the spiking has been considered but as P-total not has the same high recovery it is considered likely that the reference material used for spiking total-P consists of a not insignificant amount of PO4-P.
For fresh water Hg has a high deviation for sample A/B and low recovery in sample C. This could be caused by the use of PE bottles instead of glass bottles. Another possibility could be that Hg is more analytical difficult to analyse compared to the rest of the components due to the low concentration levels.
A comparison between data and the analytical methods/apparatus used in the different laboratories was performed. The components were analysed with two to six different methods/apparatus. This comparison showed no bias for specific analytical methods/apparatus. The data is not shown as this could compromise the anonymity of the single laboratories.
