Date of Award


Document Type


Degree Name

Master of Science (MS)


Civil and Environmental Engineering

First Advisor

Rizzo, Donna


Pilot-scale constructed wetlands (CW), with electric arc furnace steel slag phosphorus removing filters were built and implemented for treating dairy wastewater on the Paul Miller Dairy Farm at the University of Vermont. Two distinct CW types were used in the study, hybrid and integrated. Hybrid CWs consist of more than one CW in-series, and in this research three hybrid CW systems were tested. Two hybrid CWs used a vertical saturated flow (VF) CW followed in-series by a horizontal saturated flow (HF) CW. One hybrid CW used a HF CW followed in series by a second HF CW. In this study, three integrated CW systems were implemented which consisted of a CW followed by a phosphorus removing slag filter. Two integrated CW systems consisted of a VF CW followed in-series by a saturated horizontal flow slag filter. One integrated CW system consisted of a HF CW followed by a horizontal saturated flow slag filter. All individual CWs and slag filters had the same dimensions; a length, width, and height of 1.7m, 1.1 m, and 0.5 m, respectively. CWs were filled with 2 cm diameter gravel, topped with 3 cm of compost, and planted with river bulrush (Schoenoplectus fluviatilis). Slag filters were filled with 2-5 cm diameter slag, having a porosity of 0.42. From August to December of 2007 the six CW systems were fed with a pulse flow of dairy wastewater with a hydraulic loading rate of 1.9 cm/day. This gave a nominal retention time of ~10 days for each CW system. From May to September of 2008 the flow was changed to a continuous inflow, and resulted in a hydraulic loading rate of 3.9 cm/day. A nominal retention time of ~5 days for each CW system was calculated. Weekly monitoring was carried out for five-day biochemical oxygen demand (BOD5), total suspended solids (TSS), ammonium (NH4+), dissolved reactive phosphorus (DRP), and pH. Integrated CWs and hybrid CWs were monitored for their ability to treat dairy wastewater. The analyses focused on determining if there were important differences in pollutant treatment efficiencies between the two distinct systems. Time series temporal semivariogram analysis of the measured water parameters illustrated that different treatment efficiencies existed in the beginning of the 2008 summer (early summer) compared to the end of the 2008 summer (late summer). Furthermore, the CWs were found to have significantly different (p-value < 0.0001) treatment performances, in terms of TSS and BOD5 removal, from early to late summer 2008. Integrated CWs remove significantly more DRP than hybrid CWs (p-value < 0.05). During the late summer of 2008, the integrated CWs removed significantly more ammonium (p-value < 0.05) then all other CW systems. Hybrid and integrated CWs were both efficient in removing organics, but the hybrid systems were significantly (p-value < 0.05) more efficient during the period of highest macrophyte biomass. Mechanisms of pollutant removal in these CW systems were further analyzed. Ultimate BOD laboratory experiments were used to determine the maximum amount of biologically available organic matter and the corresponding rate constants for the removal kinetics of organic matter in the dairy wastewater. Geochemical modeling of the minerals that form on steel slag show that hydroxyapatite controls the activities of phosphate and calcium ions. The geochemical modeling results show a similar finding to field results, because as the wastewater:slag ratio decreases the pH increases (pH > 11) and phosphorus is removed.