Multivariate assessment of activated sludge stability in lab-scale experiments

• Assessment of activated sludge stability is crucial for reproducible results.
• This paper proposes a novel multivariate stability assessment technique.
• It is a combination of PCA and the squared prediction error statistic.
• Validation shows clear discrimination between the transient and stable periods.

In wastewater research, the transfer of activated sludge from large-scale plants to lab-scale reactors induces a transient period. It is crucial to quantitatively assess the stability of activated sludge before starting any experimental procedure. Otherwise, the transient interferes with the experimental results, jeopardizing reproducibility and accuracy. This paper presents a novel multivariate technique to assess activated sludge stability based on the combination of principal component analysis (PCA) with the squared prediction error (SPE) statistic. The proposed method allows for a more accurate estimation of activated sludge stability than existing univariate methods and also eliminates the need to establish thresholds for every single variable. The procedure is validated on experimental data obtained by Van den Broeck et al. [1].