Content area
Abstract
Road de-icing salt is one contaminant of concern in stormwater runoff, as it has been shown to have negative effects on plant and animal species, decrease biodiversity, and degrade environmental quality. It has been assumed that road de-icing salt would wash through watersheds with spring rains, as road salt (usually NaCl) is soluble, and chloride (Cl-) has long been considered a conservative tracer. However, many recent studies suggest that significant proportions of chloride mass may be retained within a watershed and that chloride levels resulting from winter salting activities may remain elevated late into summer months. Stormwater control measures (SCMs) have been praised for both volume reduction and improved water quality, but recent studies are showing that certain SCMs may increase the negative effects of road salting on the surrounding environment, such as contamination of groundwater, trace metal leaching, stratification in ponds, toxic effects, and reduced biodiversity. Because chloride poses a possible threat to downstream waters, a study was performed to study the fate and transport of chloride through one SCM, a constructed stormwater wetland (CSW), on Villanova’s campus.
The study had three main goals: i.) to determine if effluent concentrations of chloride from the CSW meet recommended EPA standards for both chronic and acute criteria; ii.) to compare total dissolved solids (TDS), conductivity, and chloride concentration data in order to further validate the data sets and also quantify correlations between each parameter; and iii.) to perform a mass balance of chloride to study the fate and transport of chloride upstream, within, and downstream of the CSW. Chloride concentrations and TDS/conductivity/chloride correlations were analyzed over a period of four years, from December 2011 - November 2015 and a mass balance was done with flow data for 2013 and 2014. High chloride spikes were observed during the Salt Application period, between December and May, and heightened baseflow chloride concentrations lasted through summer months, even when storm event concentrations were low approaching > 10 mg/L. In using EPA freshwater quality criteria of chronic (230 mg/L) and acute (860 mg/L) toxicity as benchmarks, it was found that the CSW effluent concentration was below chronic only 33% of sampled baseflow events (10% exceeded acute) and 77% of sampled storm events (12% exceeded acute). On a yearly basis, it was estimated that the CSW effluent is below chronic criteria only 39% of the year, exceeds the chronic criterion 61% of the year, and exceeds the acute criterion 12% of the year. Because this study was unable to assess the precise EPA criteria definitions (based on 1 and 4-day averages), this serves only as an estimate of the CSW’s performance.
Chloride/TDS/conductivity correlations were created between each parameter and the Villanova CSW had a TDS/conductivity k constant of 0.64 (mg/L)/(μS/cm) for TDS < 1,000 mg/L and a k constant of 0.54 (mg/L)/(μS/cm) for values above 1,000 mg/L TDS. A comparison to a bio-infiltration basin also on campus showed that correlations were similar, even for subsurface samples at that location.
A mass balance of the Villanova CSW showed that about 40-50% of volume is reduced between the inlet and the outlet, and an equal, if not slightly greater percentage of chloride mass is retained within the CSW as well. In comparison to road de-icing salt application estimates, there is evidence that there is a greater amount of chloride mass entering the CSW than from impervious surfaces alone. This leads to questioning the possibility that groundwater flow interacts with the CSW and carries additional chloride mass from upstream locations. Analysis of discharge and chloride concentration data of Mill Creek, the natural water body immediately downstream of the CSW, showed that a large amount of chloride mass is added to Mill Creek after the CSW. This indicates that the CSW is not the only source of chloride loading to Mill Creek and that it is only a small percentage of the total chloride loading throughout the creek. Of the available dataset, Mill Creek chloride concentrations were below 230 mg/L for all sampling events but one that occurred in mid-winter of 2014.