Skip to main content
SLU publication database (SLUpub)

Report2010Open access

Comparing filtered and unfiltered metal concentrations in some swedish surface waters

Köhler, Stephan

Abstract

A number of metals (vanadium, chromium, manganese, iron, cobalt, nickel, copper, zinc, arsenic, molybdenum, cadmium and lead) were analyzed in a large number of Swedish surface waters in both filtered and unfiltered samples. Concentrations of most metals are not significantly affected by the method of sampling, sample separation and preservation technique commonly used within the Swedish National Environmental Monitoring program. Copper, zinc, cadmium, chromium, nickel, arsenic and vanadium are in a majority of the samples in the so called dissolved form (94%+/-154% ) and the fraction of filtered to total metal indicates no trend with potential driving variables such as iron, pH or total organic carbon for these metals. Lead, vanadium and cobalt have concentrations that are much lower (up to a factor of four) in filtered as compared to unfiltered samples (70%+/-48%). For all three metals the relative fraction is related to the aforementioned driving variables with both TOC and iron leading to increased fraction of filtered metal. The effect with pH is the opposite. In 10% of all samples lead and cobalt are more than 60% in the particulate form. As the dissolved part is of interest only, correction factors of around 4 would need to be applied when defining the availability of lead and cobalt in the studied Swedish surface waters. Filtered metal concentrations were lower at around 40% of the sites that had elevated unfiltered metal concentrations. This indicates that it might be worthwhile starting site specific sampling campaigns at individual sites with the aim of identifying sites with elevated filtered concentrations before taking legal action. The following answers to the posed questions may be formulated: - There is no significant bias in metal concentration if the samples are treated according to the Swedish procedure or acidified and then filtered. - There is no significant bias in metal concentration if the samples are filtered in the field or in the lab but are otherwise treated similarly. The few differences that occurred may all be traced down to 2-3 sampling sites that were known to have temporary problems with metal contamination. - Except for chromium, were slight changes in concentration occurred; there is no significant bias in concentration if the time the samples are left acidified during the sedimentation step is varied. - A large bias in metal concentration (up to a factor of four) between samples that are acidified left for sedimentation and then decanted between those that are first filtered and then acidified as described in the guidance document No 19 occurs for the metals lead, vanadium and cobalt. - For all other metals, and considering the lower 25% of the sample population, the difference is on average below 35%. This is within the range of temporal variation observed at all sites and thus needs no further study. Prolonged sampling at selected sites at least during another 6 months encompassing both spring flood and part of the summer is still necessary for a more thorough analysis of the controlling factors of the ratio of filtered to total metal concentration, especially at situation with more extreme flow conditions. Also, given the good results for transferring unfiltered to filtered concentrations in the case of lead, we recommend to study potential methods for estimating filtered metal concentrations for other metals including filtered and unfiltered absorbance, filtered organic carbon (DOC) absorbance spectra and other parameters if possible

Keywords

metaller; filtration; ytvatten; lead; cadmium; copper; zink; arsenate; organic carbon

Published in

Rapport / Sveriges lantbruksuniversitet, Institutionen för vatten och miljö
2010, number: 2010:04
Publisher: Sveriges lantbruksuniversitet, Institutionen för vatten och miljö