Abstract
The suitability of poultry droppings as a biostimulant for the remediation of crude oil-contaminated soil has been investigated. Four equal-sized containers, each containing 400g of soil were each contaminated 20 ml of crude oil after which they were thoroughly mixed. To first three containers, A,B and C were respectively added 20 g, 60 and 100 g of poultry droppings which were previously dried and pulverised. To the fourth container which served as control, was added no poultry droppings. The degradation of oil in the samples were monitored for 7 weeks by observing the residual hydrocarbon content (RHC) and pH of samples. The RHC of the samples over the seven-week period were then modelled using artificial neural network (ANN) in the MATLAB Neural Network Toolbox. The RHC values decreased for all samples, with the highest reduction of 100 mg/kg obtained in sample C and the least reduction of 1000 mg/kg in sample D. The PH values were observed to increase slightly from the acidic region of 5.5 to a range of 7.8 to 8.3. The best and most suitable training algorithm of RHC of the samples was TRAINSCG since it had the least mean square error (MSE) value of 0.00183 as well as the highest R-squared value of 0.99808.
Highlights
The suitability of poultry droppings as a biostimulant for the remediation of crude oilcontaminated soil has been investigated
Chicken droppings were employed as biostimulants of indigenous microorganisms for the purpose of treating crude oilcontaminated soil
Residual hydrocarbon contents (RHC) and pH of the samples were used as indicators of bioremediation
Summary
Collection of Poultry Droppings and Soil sample: Loamy soil samples collected were air-dried and sieved using a 2 mm mesh sieve and stored in polythene bags. The poultry droppings were obtained from the University of Benin agricultural farm They were sun-dried for three days and thereafter pulverized before mixing with the soil. Thereafter, 20 g, 60 g and 100 g of the prepared poultry droppings were respectively mixed the oil-contaminated soil samples in the container A, B and C. The first 20 sets of data were used for training of the ANN models, and the last 8 sets of data were employed for testing. Residual hydrocarbon contents (RHC) and pH of the samples were used as indicators of bioremediation. These parameters were monitored for a period of seven weeks. The network was trained by the following learning algorithms: TRAINBFG (BFGS Quasi-Newton), TRAINBF, TRAINCGB (Conjugate Gradient with Powell/Beale Restarts), TRAINCGP (Polak-Ribiere Conjugate Gradient), TRAINCGF
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
