Hydrologic and metal removal potential of filtering swales for stormwater control in cold climates

Runoff from urban areas carries a wide range of pollutants both particle-bound and dissolved. Dissolved pollutants are generally more mobile and bioavailable as well as more difficult to remove by conventional infiltration based systems. Three adsorbent amended filters were studied for suitability in filtering swales for stormwater runoff management in cold climates, with special attention to the capture of dissolved metals. These alternative filters were composed of clean homogeneous sand and one adsorbent. The adsorbents used along the course of this thesis were granulated activated charcoal, granulated olivine Blueguard® G1-3, and pine bark Pinus Sylvestris. These adsorbents were selected in a previous multi-criteria study in which factors such as sorption capacity with respect to dissolved zinc (Zn), copper (Cu), nickel (Ni), and lead (Pb), cost,availability, end-of-life disposal, and others were considered. The investigations were divided into the following 4 parts 1. An urban runoff sampling campaign over the course of 17 months in the city of Trondheim in order to characterize runoff and understand local patterns. 2. A column test to study the retention performance of these filters towards toxic metals detrimental for receiving water bodies. 3. A column test to study the infiltration response of these filters under cold climate conditions. 4. A full scale experimental study with two swales composed of selected filters and one swale composed of traditional soil for bioinfiltration in order to explore and compare their hydrological balance during runoff events. The results from the characterization of urban runoff showed strong seasonal and spatial variations in all studied parameters. Metals associated to colloids were practically nonexistent, and Cu was the only studied metal that exceeded Norwegian, European, and American water quality standards for fresh water bodies. Overall the metals were mostly particle-bound (> 99 % for Pb, > 96% for Ni, > 90 % for Cu, and > 91 % for Ni), and this trend was followed over the course of the year. Lack of cleaning and maintenance in low annual average daily traffic streets can yield pollutant concentration similar as transited roads. In addition, boundaries associated to cold climate such as use of studded tires, deicing salts, and low temperatures were more likely reasons for the seasonal variations rather than long accumulation times on snow piles. With regard to the retention performance of the filters, pine bark and olivine amended filter showed the best affinity to the target metals both under high inflow rates (> 97 % retention) and large inflow volumes (no exhaustion was reached in these filters), which helped to decide the candidate filters to test in full scale. In addition, all filters except the charcoal amended filter for Ni showed little leaching of metals (< 2.8 %) after the addition of sodium chloride (NaCl), being a common de-icing salt to avoid snow accumulation on roads. The pine bark amended filter showed the longest estimated service life, which will vary according to the target metal as well as the size of the tributary. For example, a filter composed of sand and pine bark receiving polluted runoff from a tributary 50 times larger (than the filter) is expected to last 9, 13 and 46 years for Zn, Cu and Ni, respectively. With respect to infiltration capacity under cold climate conditions, snowmelt runoff will take more than 21 hours to be infiltrated through a partially frozen filter. This means that a ponding capacity is required if “shock” pollutant loads from snowmelt must be captured in the system. However, after the first 24 hours from the observation of the first discharge, infiltration rates will be higher than 1.27 cm/h, a value typically recommended for bioinfiltration. In addition, some of the studied adsorbents showed high unfrozen water content as well as being highly porous. These properties make the infiltration response to negative temperatures better due to positive impacts on the ratio ice content / porosity as well as on the heat exchange between incoming runoff and ice in the filter media. In the swale setup, two swales composed of adsorbent amended filters (filtering swales) were compared with a traditional grass swale, and among each other. The experiments were performed under different boundary conditions that have been shown to influence the hydrological balance; type of media (olivine amended filter, pine bark amended filter, and soil traditionally used for bioinfiltration practices), inflow rate, swale slope, and degree of saturation of the media. Results showed that the infiltration performance of the filtering swales was considerably better and only dependent on the inflow rate for the studied range of slopes, up to 4 %. An equation that relates infiltration performance, storm intensity, and size of the drainage area is included to help planners to design filtering swales presented in this thesis. As an example, these swales can capture 90% of the runoff generated by a tributary 40 times larger (than the swale) during a storm event with a peak intensity of 12.2 mm/h. All in all, the results of this thesis show that the environmental and hydrological benefits provided by the adsorbent amended filters are promising for further field implementation in cold climates. However, this thesis has not considered any economic feasibility aspects, which would be of interest for an overall assessment