Online high-resolution real-time monitoring techniques for anions in river water.
Julia Arndt, Anna-Lena Gerloff, Alex Zavarsky, Michael P Schlüsener, Arne Wick, Lars Duester
Abstract
Open AccessHigh-resolution real-time monitoring gains increasing importance in environmental monitoring. We employed and evaluated three real-time monitoring techniques to monitor anions at the river Rhine in Koblenz, Germany. These are a sensor for nitrate in situ and ex situ, a colorimetric device for nitrate and nitrite, and an online ion chromatography (IC) for fluoride, bromide, chloride, nitrate, nitrite, phosphate, and sulfate. The optical sensor performed well ex situ while in situ, a higher maintenance effort was needed due to fouling processes. Colorimetry failed to analyze and notify when solutions fouled and required a very high amount of maintenance to reliably analyze nitrate and nitrite. The online-IC was feasible for analyzing seven anions with a low amount of maintenance once the required infrastructure for the instrument was provided. The online high-resolution IC data was compared with conventional monitoring data with, e.g., a routine interval of two weeks, which was shown to fail to capture peak concentrations during high discharge events. Additionally, to monitor short-term events, like the shortest registered peaks of 5 h, monitoring intervals of at least 2 h at the river Rhine are needed. By extending the IC time series to 2.5 years supported by principal component analysis, a dendrogram clustering, and event analysis by QC-diagrams, two major groups of anions were visible. The first group includes discharge-driven anions such as chloride, bromide, sulfate, and fluoride, with fluoride being additionally influenced by the equilibrium precipitation of fluorite. The second group consists of seasonal-driven anions like nitrate, nitrite, and phosphate, with phosphate being also affected by short-term events and hence, remains to be an important indicator in online monitoring setups. To get a complete picture, anion analyses can be combined with, e.g., element or non-target analyses.