Research on a spectral recognition method for on-line measurement of COD in dyeing wastewater based on SIMCA

Abstract

To overcome the relevant disadvantages of the standard method and the single wavelength spectroscopic method for on-line measurement of chemical oxygen demand (COD) in dyeing wastewater after sewage treatment, a method using multi-wavelength UV-Vis spectra combined with a pattern recognition method based on soft independent modelling by class analogy (SIMCA) was developed and evaluated. The UV-Vis spectra were measured on-line by a self-developed UV-Vis optical fiber sensor system. Through experiments using binary mixtures of dyes, the results show that the SIMCA model can offer greater recognition and accuracy for COD detection. Since it is likely that multiple models are required in reality, the complexity of the COD monitoring system will be greatly reduced in this case as the SIMCA model can include several species of dyes. Thus, without requiring a process of digestion, as a preliminary qualitative analysis method, UV-Vis spectroscopy with SIMCA showed good potential to be used for on-line water quality monitoring.