TAIWO OLUWASEUN

The accurate prediction of the Well roductivity Index (PI) is critical for reservoir management, production optimization, and forecasting. Traditional methods, such as analytical correlations and decline curve analysis, are often limited by simplifying assumptions that fail to capture the complexities of heterogeneous reservoirs like those in the Niger Delta. This research addresses
this limitation by developing and evaluating a data-driven framework for PI prediction using machine learning (ML) on historical production data. The study implements and compares three advanced ensemble regression algorithms—Random Forest, GBoost, and CatBoost—to predict PI from daily records of oil, gas, and water production rates and downhole pressures. A dataset approximately 7,000 daily records from five Niger Delta wells was utilized, with the PI target variable calculated using a proxy for reservoir pressure drawdown. A clear performance hierarchy was established among the models. Random Forest yielded the weakest performance (R² = 0.18, MAE = 65.62), while XGBoost showed substantial improvement (R² = 0.78, MAE = 34.14). CatBoost erged as the superior model, achieving exceptional predictive accuracy with an R² of 0.95, a Mean Absolute Error (MAE) of 18.96, and a Root Mean Squared Error (RMSE) of 21.02. Residual and temporal analyses confirmed that CatBoost produced unbiased, homoscedastic errors and effectively tracked the dynamic PI trends of individual wells over time. Interpretability analyses revealed that production rates (oil, gas, and water) were the most influential predictors, a finding consistent with reservoir engineering principles. However, this also highlights a methodological caveat regarding the mathematical coupling between the model's inputs and the PI target. The study concludes that CatBoost provides a robust and highly accurate model for PI prediction from routine field data, offering a significant advantage over traditional methods for well performance monitoring and screening in the Niger Delta context

The accurate prediction of the Well Productivity Index (PI) is critical for reservoir management, production optimization, and forecasting. Traditional methods, such as analytical correlations and decline curve analysis, are often limited by simplifying assumptions that fail to capture the complexities of heterogeneous reservoirs like those in the Niger Delta. This research addresses this limitation by developing and evaluating a data-driven framework for PI prediction using machine learning (ML) on historical production data. The study implements and compares three advanced ensemble regression algorithms—Random Forest, XGBoost, and CatBoost—to predict PI from daily records of oil, gas, and water production rates and downhole pressures. A dataset of approximately 7,000 daily records from five Niger Delta wells was utilized, with the PI target variable calculated using a proxy for reservoir pressure drawdown. A clear performance hierarchy was established among the models. Random Forest yielded the weakest performance (R² = 0.18, MAE = 65.62), while XGBoost showed substantial improvement (R² = 0.78, MAE = 34.14). CatBoost emerged as the superior model, achieving exceptional predictive accuracy with an R² of 0.95, a Mean Absolute Error (MAE) of 18.96, and a Root Mean Squared Error (RMSE) of 21.02. Residual and temporal analyses confirmed that CatBoost produced unbiased, homoscedastic errors and effectively tracked the dynamic PI trends of individual wells over time. Interpretability analyses revealed that production rates (oil, gas, and water) were the most influential predictors, a finding consistent with reservoir engineering principles. However, this also highlights a methodological caveat regarding the mathematical coupling between the model's inputs and the PI target. The study concludes that CatBoost provides a robust and highly accurate model for PI prediction from routine field data, offering a significant advantage over
traditional methods for well performance monitoring and screening in the Niger Delta context.

Hydrocarbon contamination of land, water, air, vegetation and human is a widespread global
environmental concern. The aim of this study was to evaluate the performance of goat manure
for the bioremediation of used engine oil polluted soil. 10kg soil sample was collected from a
site free of used engine oil contamination (from an agricultural land in The Department of
Petroleum Engineering, Faculty of engineering, University of Benin, Ugbowo campus, Benin
City, Edo State in Nigeria) using a 22-cm hand-dug soil auger and stored in labeled black
polythene bag. The sample was air dried, grinded and sieved through 2mm mesh before use. Before contamination, the soil sample was subjected to chemical digestion using 1:1 ratio of
0.25M hydrochloric acid and Nitric acid. Thereafter, it was characterize to determine the physio- chemical properties. The physio-chemical properties determined include; Total Heterotrophic
bacterial, Moisture content Soil, pH, Electrical conductivity, Total hydrocarbon content (THC), Total organic carbon, Total nitrogen content in addition to the soil composition including percent
sand, Total Phosphorus, Lead (Pb) and Iron (Fe). The used engine oil was added gradually into
the bowl containing the unpolluted sieved soil sample and was properly mixed. The used engine
oil was to serve as the pollutant. The soil samples were left for 4days for stabilization before the
commencement of treatment process. The experiment was monitored for a period of eight (8)
weeks under which appreciable level of remediation had been obtained. Result obtained shows
that there was a gradual increase in pH, Electrical conductivity(EC) and Total Heterotrophic
bacterial(THB), and also a gradual decrease in total nitrogen content(TNC), total organic
carbon(TOC), total phosphorus(TP), lead(Pb), Iron(Fe) and total hydrocarbon content(THC). The result explicitly showed that goat manure is a good substrate for bioremediation of used
engine oil polluted site with calculated engine oil removal efficiency of 62.67%. The kinetic
IV
modeling shows that the experimental data fitted well with pseudo-second order kinetic model. On predicting the rate of hydrocarbon loss with time the non-linear regression model gave higher
coefficient of determination of 0.9874 compared to the linear regression model that gave 0.9665.