Perpustakaan Digital ITB

In a hydraulic fracturing project, it is imperative to consider a lot of parameters such as fracture growth history, proppant distribution, fluid properties, design treatment, geological aspects and reservoir characteristics in order to achieve an optimized treatment. A common measurement of the success ratio of a hydraulic fracturing treatment is the Fold of Increase (FOI), which is a comparison of the productivity index before and after the treatment. Nonetheless, there is no particular function that associates the parameters above to yield the value of FOI. To overcome this difficulty, Artificial Intelligence (AI) brings an alternative solution to this certain kind of complex problem where analytical calculation and classic statistics fail to provide satisfactory solutions. This paper introduces the utilization of AI, especially Adaptive Neuro Fuzzy Inference System (ANFIS) in recognizing the pattern of input data and determining FOI as the output, with a particular value of error. Processing and building the ANFIS model is the key to achieve the smallest error, which leads to an accurate estimation of FOI. This study presents a simple and quick calculating technique to predict an accurate FOI for a hydraulically fractured well. Consequently, it gives additional preferences for engineers prior to starting a hydraulic fracturing stimulation in a field. The result of this study is contenting since ANFIS successfully predicts the FOI with an average error of 1,319%. It is expected that the ANFIS model should be applicable for any hydraulic fractured well anywhere in the basin.