## Electronic Theses and Dissertations

1996

Dissertation

Ph.D.

#### Department

Civil and Environmental Engineering

Bewtra, J. K.,

#### Keywords

Engineering, Chemical.

CC BY-NC-ND 4.0

#### Abstract

A biological process employing green sulfur bacteria was investigated to remove sulfide (S$\sp{=}$) from industrial wastewaters and convert it to elemental sulfur. This research is unique in that dissolved sulfide was present in the liquid influent fed into a continuous-flow photosynthetic bioreactor. A suspended-growth once-through continuous-flow stirred-tank bioreactor was successfully operated under five different experimental conditions. For the first three experiments, concentrated nutrient solution and sulfide stock solution were pumped separately into a 13.7 L reactor at a hydraulic retention time of 45 hours and S$\sp{=}$ loading rates of 2.1, 4.4, and 5.6 mg/h$\cdot$L. At the lowest loading rate, nearly all influent S$\sp{=}$ was oxidized to sulfate. The middle loading rate resulted in complete conversion of S$\sp{=}$ to S$\sp0$. Steady state conditions were not achieved at the highest loading rate, resulting in the accumulation of S$\sp{=}$ in the bioreactor. In two more experiments, nutrient medium and S$\sp{=}$ stock solution were separately fed into a 12.0 L bioreactor at S$\sp{=}$ loading rates of 3.2 and 2.7 mg/h$\cdot$L, and hydraulic retention times of 173 and 99 hours respectively. In these trials, the loading rates were adjusted to maintain a residual of S$\sp{=}$ in the bioreactor, and consequently, there was nearly complete conversion of the consumed S$\sp{=}$ to S$\sp0$. A parameter was developed to relate the experiments of this dissertation with those reported in the literature, where smaller reactors and higher bacterial concentrations were used in batch reactors fed with $\rm H\sb2S\sb{(g)}$. This parameter described the capacity of the bioreactor to consume S$\sp{=}$, and was calculated as the product of the radiant flux per unit reactor volume and the bacteriochlorophyll concentration. Three predictive models were developed for the bioreactor. In the yield-based model, a yield coefficient was used to link the increase in bacteriochlorophyll with the consumption of S$\sp{=}$. Poor correlations between the rates of reaction and the concentrations of the reactant sulfur species led to the conclusion that a reaction pathway-based model was not appropriate for this system. An empirical model was proposed to relate the reactor volume, S$\sp{=}$ loading rate, reactor irradiation and bacteriochlorophyll concentration.Dept. of Civil and Environmental Engineering. Paper copy at Leddy Library: Theses & Major Papers - Basement, West Bldg. / Call Number: Thesis1995 .H465. Source: Dissertation Abstracts International, Volume: 57-07, Section: B, page: 4554. Adviser: J. K. Bewtra. Thesis (Ph.D.)--University of Windsor (Canada), 1996.

COinS