Non-point sources (NPS) of pollution are non-discernable, diffuse sources of pollution that are often difficult to localize and in turn mitigate. NPS can include stormwater runoff, agricultural/aquaculture wastes and wastes from small decentralized wastewater treatment systems, such as conventional septic systems. The mitigation of these NPS is imperative to reduce their potential detrimental effects on the water environment. This dissertation addresses novel treatment technologies for the mitigation of nutrients, particularly nitrogen, in Recirculating Aquaculture Systems (RAS) and onsite wastewater treatment systems (OWTS). The removal of trace organics limiting RAS production and water reuse were also investigated.
The first question this dissertation addressed is: Can the application of a UV-TiO2 reactor reduce the concentration of off-flavor compounds in RAS? In the UV-TiO2 reactor, spray-coated TiO2 plates were placed in an aluminum reactor and exposed to UV light. The process was applied in both a full-scale sturgeon RAS and a bench-scale RAS for the degradation of Geosmin (GSM) and 2-methylisoborneol (MIB). Improved performance on the removal of GSM and MIB was observed when the UV-TiO2 was applied as a batch reactor since it allowed for a longer treatment time without the effect of constant production of the compounds in the biological treatment processes. Treatment performance of UV-TiO2 was affected by GSM and MIB concentrations and dissolved oxygen. No harmful effects were observed on other water quality parameters when the UV-TiO2 reactor was operated as a batch or side stream process.
The second question this dissertation addressed is: Does the application of Tire-Sulfur Hybrid Adsorption Denitrification (T-SHAD) in RAS improve nutrient and off-flavor compound removal when compared to conventional heterotrophic denitrification? T-SHAD combines tire mulch as an adsorbent and sulfur oxidizing denitrification for the removal of NO3--N from the aquaculture waters. Adsorption studies showed the tire has significant adsorption capacity for the off-flavor compounds GSM and MIB but can be limited by contact time and, possibly, the presence of competing organic matter in RAS. The application of T-SHAD as an effluent polishing step in RAS with a high empty bed contact time (EBCT) of 720 min removed 96.6% of NO3--N and 69.6% of GSM. The application of T-SHAD within RAS as denitrification side treatment for NO3--N removal resulted in lower EBCT (185 min) that limited NO3--N removal to 21% and showed no significant removal of off-flavor compounds. The comparison between T-SHAD and a molasses fed heterotrophic upflow packed bed reactor (UPBR), showed no significant differences in N species concentrations as well as off-flavor compound removal. However, high production of SO42- resulted from sulfur oxidizing denitrification (SOD) processes was noted.
Hybrid Adsorption and Biological Treatment Systems (HABiTS), is composed of two biofilters in series employing ion exchange (IX) and nitrification for removal of NH4+ and tire scrap coupled with sulfur chips and oyster shells for both adsorption and SOD of NO3-. The third question addressed in this dissertation is: What IX/adsorption media best balances both ammonium removal and cost effectiveness for application in OWTS? Adsorption isotherms performed with different media materials showed that the zeolite material, clinoptilolite, was the best medium for the nitrification stage of HABiTS due to its high IX capacity for NH4+and cost. An adsorption capacity of 11.69 mg g-1 NH4+-N when in competition with other cations present in septic tank effluents was determined by the IX model fit to the data.
The cost of clinoptilolite is significantly higher than the other media materials tested. However, the high adsorption capacity would allow for low dosages that can be combined with non-adsorptive material reducing overall costs.
The fourth question this dissertation addressed is: How is the BNR process within HABiTS affected by IX? Results from side-by-side biofilter studies with HABiTS and a conventional nitrification/denitrification biofilter showed that the combined IX and nitrification in HABiTS can allow for faster startup, sustain variable loading, and achieve over 80% removal of NH4+ at a hydraulic loading rate of 0.34 m3 m-2-d-1 when compared to the conventional biofilter with 73% removal. Under lower loading rates the biological treatment was enhanced and dominated the NH4+ removal processes in both columns. The addition of a denitrification stage decreased Total Inorganic Nitrogen (TIN) by 53.54% and 40.97%, for the HABiTS treatment and the control treatment, respectively, under loading rates of 0.21 m3 m-2-d-1. Further decrease of NH4+-N loading rates results in high desorption of exchanged NH4+ in the clinoptilolite, resulting in lower TIN removal efficiencies (28.7%) when compared to the conventional control treatment (62%).
The final question addressed in this dissertation is: Does the proposed hybrid system enhance the removal of TIN in OWTS under transient loading conditions? Further studies with HABiTS and the conventional biofilter were performed to determine N removal performance on an hourly basis. It was found that the performance of HABiTS varies with daily and hourly loads, particularly when recovering from periods of very low loading to high loadings and vice versa. If recovering from low loading periods, IX is observed for HABiTS and the biofilter outperforms the conventional treatment in overall TIN removal. However, recovery from a high loading period results in release of NH4+-N stored in the clinoptilolite and increased production of NO3--N that could affect the performance of the denitrification stage.
| Identifer | oai:union.ndltd.org:USF/oai:scholarcommons.usf.edu:etd-8138 |
| Date | 06 July 2017 |
| Creators | Rodriguez-Gonzalez, Laura C. |
| Publisher | Scholar Commons |
| Source Sets | University of South Flordia |
| Detected Language | English |
| Type | text |
| Format | application/pdf |
| Source | Graduate Theses and Dissertations |
Page generated in 0.0026 seconds