ODOR CONTROL PILOT PLANT TESTING AND SYSTEM DESIGN FOR A FERMENTATION PLANT

A pharmaceutical manufacturer who produces animal antibiotics by fermentation was causing offsite odor impacts. Suspected odor sources included fermenter tank exhaust vents, rotary dryers that dry the final product, and the facility's wastewater treatment plant. In response to the offsite odor impacts, a conceptual design study was conducted which included characterization of the odor sources, and bench-scale screening tests. Odor sources were characterized through dilution-to-threshold (D/T) measurements conducted by the Modified Syringe Dilution Method, and through speciation by gas chromatography-mass spectrometry (GC/MS). The D/T measurements provided the more meaningful results. GC/MS results provided minimal insight into the constituents of the odor sources because most compounds were below the method's analytical detection limit. The bench-scale tests employed a three-stage impinger train to evaluate the effectiveness of different chemicals in removing the odors. Chemicals tested were sulfuric acid to scrub basic compounds; and sodium hydroxide with sodium hypochlorite, and hydrogen peroxide, to scrub acid gases. Impinger performance was measured with D/T measurements. The results of the impinger bench tests and an evaluation of other technologies indicated that chemical wet scrubbing was the most cost-effective control technology. As a result, a pilot-scale study was conducted which consisted of a two-stage packed bed system designed to operate the first stage at low pH through the addition of an acid, with the second stage designed to operate at an elevated pH through the addition of sodium hydroxide, and with the addition of sodium hypochlorite. Pilot test performance was measured in terms of D/T and chemical oxygen demand (COD) of the scrubbant solution. Gas chromatography and an Orsat analyzer measured inlet and outlet carbon dioxide concentrations to the pilot unit. An equilibrium gas-liquid relationship between D/T and the COD of the scrubbant solution was employed to determine the minimum required blowdown rate of the two-stage scrubber system. Pilot data was used to scaleup to a full-scale design. Dispersion modeling was conducted to determine the packing depth required to reduce offsite impacts and to evaluate alternative exhaust stack heights.