Developing a reliable predictive model for the biodegradability index in industrial complex effluent

The interaction between chemical oxygen demand (COD) and biological oxygen demand (BOD5) in wastewater from Tehran's Paytakht and Nasirabad Industrial Parks is investigated in this work. Monitoring platforms of industrial parks were the base frame of monthly collection data for laboratory measurements (for BOD5 and COD) and in-situ measurements (for DO, EC and Temperature-T°C) with a frequency of 4-hour samples/day. Backward elimination regression analysis was employed as an integrated procedure to find out effective model removing ineffective independent variables. Multivariate Regression analysis showed a relatively strong linear relationship between COD and BOD, with independent variables with R²=0.64 and R²=0.59, respectively. A prediction model for BOD based on COD was found by analyzing important effluent quality variables using simple linear regression and a strong linear association (BOD = 0.433COD + 222) with R² = 0.94, MSE = 38,829, RMSE = 197.05 was obtained. In all of these regression analyses, model accuracy was assessed by conducting statistical tests on the residuals. To verify and improve the reliability and practicability of model, it is applied of industrial parks' wastewater records of countries around the world such as Egypt, France, India, Pakistan and Malaysia. The extracted model applied on some of the mentioned countries' records and the results of BOD prediction was matched by observations in 95% of reliability domain. Variation of BOD-COD ratio was least affected by pH and temperature; the results underline the requirement of localized validation resulting from industry-specific differences and promote cost-effective, quick wastewater evaluation, hence lowering reliance on laboratory-based BOD₅ testing. It defiantly provides the opportunity of analytical and applied researches in south countries toward sustainable industrial wastewater management.