Preliminary investigation and neural network modeling of palm oil mill effluent as a potential bio-stimulating organic co-substrate in hydrocarbon degradation

Abstract

The extraction processes of palm oil from palm fruit bunch for industrial and domestic applications generate the palm oil mill effluent (POME) and it is considered a huge environmental challenge. The present study investigates the application of the POME as a suitable bio-stimulating organic nutrient in hydrocarbon contaminated soil (HCCS). The wet POME (WPOME) and dried POME (DPOME) were investigated in this study. The physiochemical and microbial characteristics in the HCCS showed the inadequacies of the required organic nutrients in the HCCS while its microbial population indicated well-acclimatized microorganisms. The DPOME was able to stimulate the microbial population and caused a reduction from the initial HC concentration of 4,248mg/kg to 1,600.35mg/kg while the WPOME reduced the initial HC concentration from 4,248mg/kg to 2,600.56mg/kg. The organic nutrient levels and pH in POME need to be adjusted to be within the range for optimum HC degradation and microbial activity. Data from the first-order kinetics interpretation confirmed that the DPOME was more beneficial to the HCCS as an organic nutrient. Performance evaluation of the artificial neural network (ANN) model through the root mean square error (RMSE), mean square error (MSE), and correlation coefficient (R²) confirmed that the DPOME with lower RMSE, MSE, and a higher R² performed better in the HC degradation process. Finally, this study showed that the nutrient present in POME could potentially stimulate microbial growth and serve as an effective organic nutrient in HCCS degradation.