COD Removal Prediction of DAF Unit Refinery Wastewater by Using Neuro- Fuzzy Systems (ANFIS) (Short Communication)

Abstract

In this study the Dissolved Air Flotation (DAF) system in oil refinery was investigated for the treatment of refinery wastewater. In order to investigate sytem a labratory scale rig was built. The aim is to remove some of the wastewater pollutant materials and data modeling of COD test.The effect of several parameters on flotation efficiency namely, saturator pressure, and coagulant dose, on COD removal was examined experimentally. Experiments were done by using poly aluminum chloride coagulant (PAC) at pressures 2bar up to 5bar and in three doses 15mg/l، 20mg/l and 25mg/l. After final examination's removal efficiency obtained is close to the performance achieved by the refinery. The data obtained from COD experiments using neuro-fuzzy systems have been modeled. The correlation coefficient (R2), root mean square error (RMSE) and sum of square error (SSE) of predicted values by using neuro-fuzzy systems are obtained 0.9991, 6.35×10-3and 4.04×10-5 respectively, which shows the high accuracy of neuro-fuzzy systems.

Keywords

• Adaptive neuro-fuzzy inference system
• COD
• Dissolved air flotation
• Refinery wastewater