Principal component analysis as a criterion for monitoring variable organic load of swine wastewater in integrated biological reactors UASB, SABF and HSSF-CW

• Published in: Journal of environmental management 2020-05, Vol.262, p.110386-110386, Article 110386
• Main Authors: Oliveira, Jacineumo Falcão de, Fia, Ronaldo, Fia, Fátima Resende Luiz, Rodrigues, Fernando Neris, Matos, Mateus Pimentel de, Siniscalchi, Luciene Alves Batista
• Format: Article
• Language: English
• Subjects: Agroindustrial wastewater; Anaerobiosis; Animals; Biological treatment; Bioreactors; Multivariate analysis; Organic stabilization; Principal Component Analysis; Sewage; Swine; Waste Disposal, Fluid; Waste Water; Wetlands
• ISSN: 0301-4797; 1095-8630
• DOI: 10.1016/j.jenvman.2020.110386

The multivariate analysis to optimize the parameters of wastewater is essential to reduce costs. The aim of this study was to evaluate the use of multivariate and conventional analysis in biological system composed by upflow anaerobic sludge blanket (UASB), submerged aerated biological filters (SABF) and horizontal subsurface flow constructed wetland (HSSF-CW) reactors in the organic stabilization of swine wastewater (SW). Four loads were used in the system with alteration by COD concentration of untreated SW, and the data were evaluated by principal components (PCA). The average efficiency of COD and BOD removal increased from 45% in phase I to 67% in phase IV in the UASB, SABF and HSSF-CW reactors. The principal component analysis promoted the reduction of 13 original variables to 5, 8 and 5 principal components in the UASB, SABF and HSSF-CW reactors, respectively, optimizing the dynamics of interpretation of the data that influenced the most the stability of the wastewater system across the four phases. There was a strong negative effect of oxygen concentrations in the SABF reactor in relation to organic variables, optimizing the biological mechanisms of the HSSF-CW and, therefore, enabling better decision making and cost reduction with analysis at treatment plants. [Display omitted] •Comparison between analysis of swine wastewater quality data by multivariate and conventional methods.•Principal components analysis and linear correlations represented 64.51% of the original variables in a treatment plant.•Integrated and sequential biological reactors optimize the removal of organic loads.•The use of HSSF-CW enhances SW polishing and possible the reduction of monitoring costs with multivariate statistical.