인공지능기법을 이용한 다생산정 위치선정 연구

Abstract

Well placement is one of the most important steps in conventional or unconventional field development. Reservoir simulation has been frequently used to determine the optimal locations in well-placement problems. The ideal method is to conduct simulations for all possible drilling locations, called as the exhaustive run, to obtain a true global solution wherein the cumulative production or the economic value is a maximum. However, the exhaustive run takes too much computational time. Therefore, studies using artificial neural networks have been carried out to reduce the required time and cost. In previous studies to investigate multi-wells placement problem, artificial neural network was applied to scenarios selected by expert's experience and intuition. However, this method does not take into account all cases in which the multi-wells can be located in the entire reservoir, and thus there is possibility of not obtaining the true global solution. In this study, a new algorithm is proposed to find the optimal locations of multi-wells by using sequentially artificial neural networks in hierarchical grid systems. By applying the sequential artificial neural network (ANN) for each hierarchical grid system, a new search space is identified, where a more detailed grid system is used for the next sequential ANN. In this methodology it is possible to identify the true global solution with relatively small number of reservoir simulations. The proposed method was applied to optimize multi-well placement in a coalbed methane reservoir. The result proved that the proposed method provided a practical, cost-effective and robust tool for field production management in helping petroleum engineers determine locations for infill drilling wells.