|Continuous Deflection Separation, Fuzzy Filter, and UV Treatment of SSO-Type Wastewater: Pilot Study Results (EPA/600/R-02/100) December 2002
This report describes a demonstration project that first entailed operation of a continuous deflection separation (CDS) unit to treat raw wastewater, which is similar to sanitary sewer overflow and combined sewer overflow in solids characteristics.
Two screens with 1200-micron and 600-micron apertures were evaluated. These technologies were substantially smaller than the 2400-micron CDS technology typically used for floatables removal. Total suspended solids (TSS) removal averaged 10 and 30 percent for the two screen sizes, respectively. The smaller screen was observed to bind at its surfaces, while the 1200-micron retained the desired self-cleaning capability characteristic of this technology.
Other technologies were also tested at the same time as the CDS units. A fiber-based media, high-rate filter (the Fuzzy Filter) was operated downstream of the CDS unit. At loadings between 400 and 600 liters per minute per square meter (10 and 15 gallons per minute per square foot), it was capable of achieving approximately 40 percent TSS removal. The process was found to effectively remove particles greater than 50-microns, which benefited the performance of downstream ultraviolet (UV) disinfection processes.
Three different UV configurations were operated downstream of the CDS and Fuzzy Filter processes. One used low-pressure, high-output lamps, while the other two used medium-pressure lamps. The medium-pressure units comprised a close-chamber and an open-channel unit. In addition to operating the pilot units, collimated-beam, dose-response testing was conducted on the primary wastewaters.
The results of the study suggest that 2-log reductions can be consistently accomplished at doses on the order of 30 millijoules per square meter, with minimal removal of particles. These reductions can be increased to between 2.3 and 2.8 with removal of particles greater than approximately 50 microns. These results are based on enumeration of blended samples. If the exposed samples are not blended, the apparent reductions will be between 2.5 and 3.5 logs.
You will need Adobe Reader to view some of the files on this page.