Análise de desempenho e confiabilidade de estações de tratamento de esgotos

Abstract

The research presents results from a comprehensive evaluation of 208 full-scale WWTPs, located in Southeast Brazil (latitudes 20 to 22o South), comprising 21 treatment processes. The evaluation focused on the performance of these treatment technologies, in terms of achieving compliance with several effluent standards, and the influence of operational conditions and the reliability of the systems. Over 50000 plant operational data were analyzed and the results showed a great variability in terms of sampling frequency, monitoring period and measured parameters. The influent flow range varied from 0.05 Ls-1 to 6,838 Ls-1. The descriptive statistics of the influent concentration data for BOD, COD, TSS, TN, TP and FC (thermotolerant coliforms) were determined, and a comparison was made between the observed concentrations and the typical values reported by the literature on prevailingly domestic raw wastewater. The results showed that, in general, the raw wastewater presented higher concentrations than usually reported by the literature. The treated wastewaters were evaluated for six frequently used treatmentprocesses, and for which there was a higher data availability: septic tank + anaerobic filter, facultative pond, anaerobic pond + facultative pond, activated sludge, UASB reactors alone, UASB reactors followed by post-treatment. For these treatment technologies, the effluent quality was evaluated, in terms of effluent concentration and removal efficiencies, compliance with discharge standards, and influence of operational conditions. The best theoretical probability distribution to describe the behavior of the variables present in raw and treated wastewater was investigated, and by this investigation the coefficients of reliability weredetermined, in terms of the compliance of effluent BOD, COD, TSS, TN, TP and FC with discharge standards or targets. In general, the average performance of the treatment technologies, in terms of all statisticalanalyses used, was lower than expected. However, good performances were obtained by individual plants from all technologies, showing that the poor results could not be attributed to the treatment processes per se but possibly to design, operational and maintenance problems.