Oil Production Forecasting with Uncertainty Description Using Data Driven Proxy Model

The petroleum industry operates under great uncertainty. Achieving an efficient approach to quantify uncertainty in oil production models is of key importance supporting decision-makers find suitable strategies for mitigating risks and maximizing profit. Uncertainty quantification commonly performed based on the Monte Carlo this a very time-consuming process by using physics-based developed reservoir simulators. To solve challenge, data-driven proxy which are less complex computationally can be used as alternative. This paper aims investigate functionality ANN method developing from advanced wells. investigation conducted through case study assessment cumulative water productions long horizontal well with ICD completion zonal isolation synthetic 10 years. In study, Eclipse® simulator base model it coupled MATLAB® generating required data sets train test model. According obtained results, trained predict wells accurately mean error than 4%. Besides, 150 times faster Eclipse challenge quantification.