Tove Wium-Andersen

Corrosion of concrete sewers—The kinetics of hydrogen sulfide oxidation

Hydrogen sulfide absorption and oxidation by corroding concrete surfaces was quantified in a test rig consisting of 6 concrete pipes operated under sewer conditions. The test rig was placed in an underground sewer monitoring station with access to fresh wastewater. Hydrogen sulfide gas was injected into the pipe every 2nd hour to peak concentrations around 1000 ppm. After some months of operation, the hydrogen sulfide became rapidly oxidized by the corroding concrete surfaces. At hydrogen sulfide concentrations of 1000 ppm, oxidation rates as high as 1 mg S m(-2) s(-1) were observed. The oxidation process followed simple nth order kinetics with a process order of 0.45-0.75. Extrapolating the results to gravity sewer systems showed that hydrogen sulfide oxidation by corroding concrete is a fast process compared to the release of hydrogen sulfide from the bulk water, resulting in low gas concentrations compared with equilibrium. Balancing hydrogen sulfide release with hydrogen sulfide oxidation at steady state conditions demonstrated that significant corrosion rates--several millimeters of concrete per year--can potentially occur at hydrogen sulfide gas phase concentrations well below 5-10 ppm. The results obtained in the study advances the knowledge on prediction of sewer concrete corrosion and the extent of odor problems.

Influence of pipe material and surfaces on sulfide related odor and corrosion in sewers

Hydrogen sulfide oxidation on sewer pipe surfaces was investigated in a pilot scale experimental setup. The experiments were aimed at replicating conditions in a gravity sewer located immediately downstream of a force main where sulfide related concrete corrosion and odor is often observed. During the experiments, hydrogen sulfide gas was injected intermittently into the headspace of partially filled concrete and plastic (PVC and HDPE) sewer pipes in concentrations of approximately 1,000 ppm(v). Between each injection, the hydrogen sulfide concentration was monitored while it decreased because of adsorption and subsequent oxidation on the pipe surfaces. The experiments showed that the rate of hydrogen sulfide oxidation was approximately two orders of magnitude faster on the concrete pipe surfaces than on the plastic pipe surfaces. Removal of the layer of reaction (corrosion) products from the concrete pipes was found to reduce the rate of hydrogen sulfide oxidation significantly. However, the rate of sulfide oxidation was restored to its background level within 10-20 days. A similar treatment had no observable effect on hydrogen sulfide removal in the plastic pipe reactors. The experimental results were used to model hydrogen sulfide oxidation under field conditions. This showed that the gas-phase hydrogen sulfide concentration in concrete sewers would typically amount to a few percent of the equilibrium concentration calculated from Henrys law. In the plastic pipe sewers, significantly higher concentrations were predicted because of the slower adsorption and oxidation kinetics on such surfaces.

Improved urban stormwater treatment and pollutant removal pathways in amended wet detention ponds

Dissolved and colloidal bound pollutants are generally poorly removed from stormwater in wet detention ponds. These fractions are, however, the most bio-available, and therefore three wet detention ponds were amended with planted sand filters, sorption filters and addition of precipitation chemicals to enhance the removal of dissolved pollutants and pollutants associated with fine particles and colloids. The three systems treated runoff from industrial, residential and combined (residential and highway) catchments and had permanent volumes of 1,990, 6,900 and 2,680 m3, respectively. The treatment performance of the ponds for elimination of total suspended solids (TSS), total nitrogen (Tot-N), total phosphorous (Tot-P), PO4-P, Pb, Zn, Cd, Ni, Cr, Cu, Hg were within the range typically reported for wet detention ponds, but the concentrations of most of the pollutants were efficiently reduced by the planted sand filters at the outlets. The sorption filters contributed to further decrease the concentration of PO4-P from 0.04 ± 0.05 to 0.01 ± 0.01 mg L−1 and were also efficient in removing heavy metals. Dosing of iron sulphate to enrich the bottom sediment with iron and dosing of aluminium salts to the inlet water resulted in less growth of phytoplankton, but treatment performance was not significantly affected. Heavy metals (Pb, Zn, Cd, Ni, Cr and Cu) accumulated in the sediment of the ponds. The concentrations of Zn, Ni, Cu and Pb in the roots of the wetland plants were generally correlated to the concentrations in the sediments. Among 13 plant species investigated, Rumex hydrolapathum accumulated the highest concentrations of heavy metals in the roots (Concentration Factor (CF) of 4.5 and 5.9 for Zn and Ni, respectively) and Iris pseudacorus the lowest (CF < 1). The translocation of heavy metals from roots to the aboveground tissues of plants was low. Therefore the potential transfer of heavy metals from the metal-enriched sediment to the surrounding ecosystem via plant uptake and translocation is negligible.

Sorption media for stormwater treatment: A laboratory evaluation of five low-cost media for their ability to remove metals and phosphorous from artificial stormwater

Five sorption materials were studied with a focus on polishing pretreated stormwater: crushed limestone, shell-sand, zeolite, and two granulates of olivine. These materials are commercially available at comparatively low cost and have been subjected to a minimum of modification from their natural states. The sorbents were tested for phosphorus, arsenic, cadmium, chromium, copper, nickel, lead, and zinc at concentration and conditions relevant for typical stormwater. The materials were tested for sorption capacity and kinetics. Desorption was tested under neutral and alkaline conditions and in the presence of chloride. For most sorbent/sorbate combinations, significant sorption occurred within the first minutes of contact between sorbent and sorbate. Treatment to the low microgram per liter range could be achieved by contact times of less than 1 hour. The study indicated that sorption filters can be designed for long life expectancy at comparatively low cost by applying the materials tested.

Monitoring the startup of a wet detention pond equipped with sand filters and sorption filters.

The startup of a wet retention pond designed for extended stormwater treatment was monitored by more than one year of continual measurement of hydraulic parameters, nutrients and quality parameters in the pond itself (pH, temperature, dissolved oxygen, turbidity). The data revealed that photosynthesis played an important role for dissolved oxygen and pH for most of the year. Another important observation was that the pond behaved more like a completely mixed reactor than like a plug flow reactor--even though the length to width ratio was as high as 4.5:1. The pond was equipped with sand filters and sorption filters whereby very good nutrient removal efficiencies were achieved.

Reduction of Stormwater Runoff Toxicity by Wet Detention Ponds

Stormwater runoff contains a wide range of contaminants [1]. Pollutants like PAH’s, biocides, heavy metals and nutrients are often found in varying, but significant concentrations. Due to the content of various contaminants, stormwater runoff has been found to be toxic to the aquatic environment [2–4], with potential negative ecological impacts on receiving waters.

Modeling nutrient and pollutant removal in three wet detention ponds

Three wet detention ponds receiving stormwater runoff and located in Arhus, Silkeborg and Odense, Denmark were studied. The ponds were continuous monitored by measuring turbidity, temperature, pH, oxygen level and water level in the ponds as well as inflow and outflow. Water samples were collected from the inlets and from the downstream end of the ponds and analyzed for suspended solids, total and soluble phos phorous, nitrogen and seven heavy metals. Performance of the ponds was modeled by routing the measured flow though the ponds and simulating pollutant removal by 1st order kinetics with and without temperature adjustment.