Publications

Steam flood injection is a thermal recovery method that involves injecting steam, or specific water at particular temperatures, through special injection wells. In this research, the Kasai heavy oil field, which has high viscosity, was injected with steam flood. However, preliminary prioritization of steam flood injection must be done at certain stages in a particular area of this field. This area consists of two sub-area, A, B, and C in the north section and A, B, C, D, E, F, G and H in the south section, which have different number of wells and subsurface conditions. The provision to prioritize injection in certain areas, consist of monthly well production data, well location data, and reservoir data. Machine learning techniques, specifically random forest regression, are used on the reservoir data to generate variable importance scores, which rank the reservoir properties variables in terms of their importance for determining the preliminary prioritization. This method was validated by comparing its accuracy with high level modelling Decisson Tree (DT) and correlating the results with the original state of the on-field development and reservoir properties. The method involves data cleansing, fitting models to data assessing the quality of fit, generating decision trees, and identifying key variables. Programming languages are used to perform these tasks in a systematic and structured way. The result of this method shows that the reservoir properties such as porosity, saturation, and permeability are important variables, in form of OOB score calculation. The final result of the prioritization shows that Kasai D, E and F in south section should be prioritized for steam flood injection first.