Uncertainties due to stormwater sampling techniques (2013)
The aim of this study is to examine the influence of automated sampling on indicator bacteria results in the field.
Since stormwater sampling can be done manually or by automated sampling, often the latter is prioritized, implying lower operational costs and avoiding the need of staff at the sampling sites during rain. Due to the high variability and special dynamics characterizing bacteria, the issue of sampling line contamination was raised impacting bacteria concentrations in the sampler.
The study is designed for both field and laboratory conditions. In Östersund, stormwater samples collected during rain events are compared by parallel sampling of the same spot manually and by automatic samplers. In the microbiological laboratory Hjortens AB in Östersund, several tests are conducted with automatic samplers under controlled conditions focusing on the impact of typical high bacteria concentrations on lower concentration samples in artificial stormwater as well as the impact of varying sampling line length (1-5m) on a possible cross-contamination between samples.
The study indicates extremely high uncertainties related to sampling methods for some of the indicator bacteria, especially coliforms and the hypotheses of cross-contamination between typical high and low concentration in stormwater samples was proven. The findings are presented and published on international conferences (Novatech 2013), journals, respectively, and the study in general provides a better understanding of sampling related biases and the importance of choosing the right sampling techniques for different stormwater constituents when assessing indicator bacteria variation in stormwater.
The project is conducted by the Urban Water Research Group in collaboration with Östersund municipality and financed by Åke och Greta Lissheds Foundation, J. Gust. Richert Foundation and Luleå University of Technology.
Helen Galfi
Updated: