Approach to rapid detection of chemical oxygen demand in livestock wastewater based on spectroscopy technology

Chemical oxygen demand(COD) is an important index in evaluating the pollution degree of the wastewater. Traditional methods cannot satisfy the demand of the wastewater detection for its high cost, long time, and second pollution. A rapid measurement method was put forward in this paper by using spectroscopy techniques. A total of 18 wastewater samples with different concentrations (1000 mL, 2000 mL) were measured using the ASD Handheld FieldSpec spectrometer. The results from both spectral absorbance and COD values in wastewater show that there are high correlation coefficients between spectral features and COD values. The partial least square regression models were therefore developed. The r² values between measured and predicted values of COD in wastewater with containers of 1000 mL and 2000 mL were 0.9895 and 0.9985, the standard errors of calibration were 14.2 and 11.5, and standard errors of prediction were 22 and 32, respectively. The relative standard deviations for unknown samples were below 10%. The tested precision can meet the requirements of the application in agriculture engineering. Compared with standard chemical method, it indicates that the spectroscopy technique can be applied in the quantitative analysis of COD in livestock wastewater, and the predictive model for COD provides a basis for developing potable instrument to detect water quality on line.