Determination of MBR fouling and chemical cleaning interval using statistical methods applied on dynamic index data

Membrane fouling is a serious problem in membrane bioreactor (MBR) processes which limits the widespread application of MBRs. Therefore, importance of prediction of MBR fouling and determination of membrane chemical cleaning interval has been increased in recent years. In this study, procedures for (1) prediction of MBR fouling using dynamic index data obtained by recursive least squares (RLS) method and (2) determination of membrane chemical cleaning interval using statistical quality control (SQC) methods applied on the dynamic index data are proposed. Traditional techniques to determine the membrane chemical cleaning interval cannot capture the dynamics of membrane filtration and fouling as there are developed based on steady-state assumption of MBRs. The newly proposed methods based on the recursive least-squares (RLS) and SQC are expected to overcome the limitations posed by traditional techniques, since these techniques can capture the dynamics of MBR filtration. Proposed methods have been tested with the data obtained from pilot-scale MBR plant. In this study, intercept (b) and coefficient (κ) of the traditional fouling model are updated using RLS method with respect to the MBR operational time, and are proposed as new indexes for determining membrane chemical-cleaning interval.

► For MBR fouling, the intercept (b) and coefficient (κ) of the linearized fouling mechanism are suggested as new parameters for membrane chemical cleaning.
► The prediction of MBR fouling and dynamic determination of chemical cleaning interval by new linearized parameters and statistical quality control are successfully carried out in a bench-scale MBR plant.
► The experimental results closely agree with those obtained from the proposed prediction method with cleaning interval determination.