RESUMEN
This study proposes the was to evaluate the stability and methane production with organic load differents in an upflow anaerobic sludge blanket reactor (UASB) treating swine wastewater by methods of multivariate analysis. Four organic loads were used with average hydraulic holding times of one day. The methods of data analysis of linear regression, Pearson correlation, principal component analysis and hierarchical clustering analysis were used for understanding stability and methane production in the reactor. The highest concentrations of bicarbonate alkalinity of 683 mg L-1 CaCO3 and total volatile acids of 1418 mg L-1 HAc with maximum organic loading applied were obtained. The optimal stability conditions occurred at an intermediate and partial alkalinity ratio between 0.24 and 0.25 observed in initial phases with a chemical oxygen demand (COD) removal of 47-57%. Maximum methane production was 9.0 L CH4 d-1 observed with linear regression positive and occurred at the highest applied organic load, corresponding to the highest COD removal efficiency and increased microbial biomass. Positive and negative correlation between functional stability in anaerobic digestion showed regular activity between acids, alkalinity and organic matter removal. This fact was also proven by the analysis of principal components that showed three components responsible for explaining 83.2% of the data variability, and the alkalinity, organic matter influent and organic acids had the greatest effects on the stability of the UASB reactor. Hierarchical clusters detected the formation of five groupings with a similarity of 50.1%, indicating that temperature and pH were variables with unitary influences on data dimensionality.
RESUMEN
The present work presents a methodology for modeling the quality of surface water, aiming at the management of hydrographic basins and the best allocation of resources in the treatment of sanitary sewage, considering the predominant uses of water and microbiological quality. The QUAL-UFMG model was used, composed of Excel spreadsheets, where visual basic for applications (VBA) routines were implemented, enabling deterministic and probabilistic modeling through Monte Carlo simulations. The proposed methodology was applied to a Brazilian hydrographic basin, called the GD2 Planning Unit (Grande River Slopes), considering the discharges of sanitary sewage from 30 municipal seats and approximately 740,000 inhabitants. Four scenarios were studied: the current situation (C-01), the trend for the year 2033 (C-02), compliance with environmental legislation (C-03) and compliance with the main uses of the basin (C-04). The results showed that for C-01 and C-02, the water quality, in terms of thermotolerant coliforms, in most stretches does not meet the defined uses. Even complying with the provisions of environmental legislation (C-03), which do not provide for disinfection, only the largest watercourses would have adequate quality for use. Complete service would only be achieved in C-04, which provides for universal sewage treatment with disinfection for the vast majority of municipalities.
