Multi Objectives Optimization to Gas Flaring Reduction from Oil Production

Abstract

In recent years the emission results from the gas flaring become cause big problem of emission the environment. Therefore, this paper focuses on present a novel prediction tool based on developing MARS data mining technique through replace it kernel by multi objective optimization function. The objectives of this paper are: (i) Give a specific definition of Gas Flaring Reduction (GFR) from oil production with determined the main limitations and hypotheses of that problem. (ii) Study the nature of datasets related to oil extraction, pre-processing that datasets, then determined the main features for each dataset. (iii) Set the constrictions for each sub-optimization problem. (iv) Design a novel predictor based on develop MARS data mining technique through replace their kernel by multi objective optimization function based on their construction. (v) Calculate the rate of Gas flaring based on three main gases (CO2, CH4, N2O) in parallel at the same time that increase the performance and reduce the time used to find an optimal result. (vi) Compute the gas to oil rate (GOR) for traditional predictor “MARS” and Develop Predictor”. Finally, evaluate and analyze the results of a proposed methodology.