Abstract The reversibility of aggregates during flocculation was investigated. The stable diameter (d) and the power law slope coefficient of the particle size distribution (β) were applied to follow re-formation after breakage. A non-intrusive image-based technique was used for monitoring flocs. Aggregates were formed by adding alum [Al2(SO4)3·18H2O] by the sweep-coagulation mechanism to two synthetic waters, prepared from kaolin (Fluka) and humic acid (Aldrich Chemical). Velocity gradients (G) varied from 20 to 120 s-1 during experiments, and the rupture occurred under controlled conditions. After rupture, the initial condition was reinstated and reversibility analyzed. Results pointed out the irreversibility of breakage for both Al-kaolin and Al-humic flocs. The stable diameter of aggregates after breakage (d2) varied from 157 to 132 µm for Al-humic and from 233 to 123µm for Al-kaolin aggregates, using G from 20 to 120 s-1. β 2 values varied from 1.2 to 4.6 for Al-humic and from 0.6 to 7.7 for Al-kaolin.

RESUMO: Este artigo apresenta algumas considerações sobre a floculação a partir da análise clássica e da avaliação direta da distribuição de tamanho de partículas. Para tal, foram realizados experimentos em escala de bancada. Os resultados foram avaliados à luz da abordagem clássica, dada pelo adimensional número de Camp (GT) e pela cinética de agregação e ruptura, e da distribuição de tamanhos de partículas. Verificou-se que o emprego isolado do produto GT não é adequado para representar as variações de desempenho na floculação e que o modelo clássico de agregação e ruptura pode ser influenciado pela medida indireta de turbidez, bem como pelo uso da sedimentação como etapa intermediária. Por outro lado, a avaliação direta de desempenho por meio da distribuição de tamanhos de partículas, representada pela função contínua linearizada, pode constituir ferramenta promissora na avaliação dedicada da floculação.

ABSTRACT: This article presents some considerations on the flocculation from the classical analysis and direct evaluation of the particle size distribution. For this purpose, essays in bench scale were performed. The results were evaluated in the light of classical approach, given by the dimensionless number Camp (GT) and the kinetics of aggregation and breakage, as well as particle size distribution. It has been found that the use of GT isolated product is not suitable to represent the variations in flocculation performance and the classical model of aggregation and break-up can be biased by an indirect measurement of turbidity and sedimentation as an intermediate step. On the other hand, the direct evaluation of performance by particle size distribution, represented by linear continuous function, can be a promising tool for the evaluation of flocculation.