Optimization of biohydrogen production from food processing wastewater

by Van Ginkel, Steven W.

The ideas of dwindling fossil fuel reserves, global warming, and the need for energy efficiency in our nation’s infrastructure inspired this thesis. One area vital to our nation’s well being is wastewater treatment. The production of hydrogen gas from wastewater using anaerobic treatment processes makes wastewater treatment more economical. Although hydrogen gas can be produced from any organic wastestream, the production of hydrogen from high strength food processing wastewater makes the most economic sense. The objectives of this thesis were to convert food processing wastewaters to hydrogen and to maximize hydrogen yields from glucose at concentrations typically found in the food processing industry. Batch and continuous reactor experiments were conducted to determine conditions for maximum biological H2 production yields and rates. A method used to increase H2 production yields involved decreasing the H2 concentration in the reactor vessel to reduce H2 partial pressure inhibition. This method was accomplished in two different ways using both batch and continuous reactor tests. In batch tests, H2 gas produced was released either continuously using a respirometer or intermittently using a glass syringe. H2 yields increased 43% when the continuous release system was used reaching a maximum of 0.92 mol-H2/mol-glucose. In continuous reactor tests, low glucose concentrations were used to reduce the H2 production rate and reduce H2 partial pressure inhibition. Continuous reactors were operated at glucose concentrations of 2.5 to 10 g COD/L at hydraulic retention times of 1 to 10 hours (2L reactor volume). H2 yields increased with decreasing glucose loading rate reaching a high of 2.6±2 mol H2 / mol glucose at the lowest glucose loading rates (0.5 – 1.9 g glucose/hr). High yields of hydrogen were consistent with a high molar acetate:butyrate ratio of 1.08:1 as more hydrogen is produced with acetate as a product (4 mol-H2/mol-acetate) than with butyrate (2 mol-H2/mol- iv butyrate). Flocculation was also an important factor in the performance of the reactors. The flocculant nature of the biomass allowed reactor operation at low HRTs with steady H2 production and > 90% glucose removal. In addition to inhibition due to hydrogen gas, undissociated acids also cause inhibition. In continuous reactor tests, the effect of the undissociated form of acetic and butyric acids on H2 production yields was tested by varying the pH, by operating reactors at high glucose concentrations, and by adding these acids directly to the influent of the reactors. Overall, total undissociated acid (p-value = 0.02) and undissociated butyric acid concentrations (p-value = 0.06) in the reactor (pH 5.5) were observed to decrease H2 yields while acetic acid had a lesser effect on H2 yields (p-value = 0.89). At influent glucose concentrations of 10 to 30 g/L, H2 yields were fairly constant at 50±2%. At a glucose concentration of 40 g/L, H2 yields were the lowest of all conditions tested at 1.6±0.1 mol-H2 / mol-glucose where a switch to solventogenesis occurred. It was concluded that a self-produced total undissociated acid concentration of > 19 mM is the threshold concentration that significantly decreased H2 yields and initiated solventogenesis under the conditions tested. In more applied tests, domestic and five different food processing wastewaters (apple, two potato wastewaters, and two confectioner wastewaters) were used as the substrate in batch tests. Gas produced from the domestic wastewater sample (concentrated 25×) contained only 23±8% hydrogen, resulting in an estimated maximum production of only 0.01 L/L for the original, non-diluted wastewater. COD removals from the food processing wastewaters as a result of hydrogen gas production were generally in the range of 5-11%. Overall hydrogen gas conversions were 0.7-0.9 L-H2/L-wastewater for the apple wastewater, 0.1 to 2.0 L/L for the confectioner wastewaters, and 2.1-2.8 L/L for the potato wastewater. v