Prediction of Bubble-point Pressure (Pb) and Formation Volume Factor (Bo) Using Group Method of Data Handling (GMDH) approach and the effect of reducing correlating parameters; a comparative study

year Petroleum Engineering undergraduate students at Universiti Teknologi PETRONAS (UTP) are required to complete a Final Year Project (FYP), as part of the graduating requirement. In this course, students are given opportunity to use their knowledge as well as problem-solving tools and methods that they have acquired throughout their study to independently carry out research or design work. The project supervision and evaluations are mainly done by lecturers although practicing engineers from the industries are invited to participate. The author’s FYP title is “The Prediction of Bubble Point Pressure and Oil Formation Volume Factor using Group Method of Data Handling (GMDH) and the effect of reducing correlating parameters; a comparative study”. A precise description of the reservoir fluid properties, hold importance essence in finding solution and solving petroleum reservoir engineering related problems. It is important for engineers to get ahold of the physical reservoir fluid properties, because they help in designing the best approach and strategies for the development of any oilfields. A broad literature review was done to assist the author to apprehend and establish the parameter, boundary, limitations of the existing correlation as well as the method he’s going to apply for his study. This paper will be evaluating published correlations aimed to predict bubble point pressure and oil formation volume factor, the author’s propose correlation which is his ambition to improve the current correlation’s accuracy used by the industry in predicting the reservoir’s bubble point pressure and oil formation volume factor through the method known as GMDH. Statistical analysis will be conducted to evaluate the model’s performance. This paper aims to generate two correlation models; bubble point pressure Pb, and Oil formation volume factor Bo, with high accuracy and less number of parameters. On the other hand, this paper will also discuss the effect of reducing parameters used in previously published correlation that took into account while developing them.