Data clustering analysis in the assessment of wastes using in the sewage filtration

Abstract

The paper discusses the use of multiclustering statistical analysis in the assessment of domestic wastewater filtration effectiveness. Calculations included data collected over four months of experiments with using waste as filling material of vertical flow filters for domestic sewage treatment. The effectiveness of pollutants removal was analysed in case of mechanically shredded waste in the form of PET flakes, PUR foam trims, shredded rubber tires and wadding. The organic compounds (CODcr, BOD5) removal, suspend solids, biogens (as NH4+, PO43−ions) and oxygen saturation changing compared with sand filling was analysed. Multiclustering statistical analysis allowed to divide pollutants removal efficiency of analysed materials into 3 clusters, depending on the hydraulic loading. The first group consisted in quality parameters of treated sewage: the highest reduction of BOD5 and NH4-N. It included the values of quality parameters and indicators for the filtrates obtained at the lowest hydraulic load from columns filled with 60 cm of rubber tires or sand. The second group comprised the results for fillings containing foam, PET and rubber tires (the other hydraulic loads). It featured the highest reduction of total suspended solids and PO43−. Removal of easily biodegradable organic compounds was at a similar level in both cluster groups. The filter filled with polyester waste (wadding), which was as effective as 30 cm layer of sand, and the filters filled with 60 cm of sand working at the highest hydraulic load. Third group showed the lowest values of parameters and indicators for analysed filtrates.

Keywords

• cluster analysis
• PET
• polyester
• polyurethane foam
• rubber tires
• sewage filtration