A Model for Platform Location Optimization in Shale Gas with Learning Effect

Shale gas in the mountain area is exploited in well factory mode. Learning effect due to well factory mode significantly affect the drilling cost, which has a powerful effect on platform location. The learning index, which is quantitative assessment of learning effect in the process of shale gas exploitment, is made by adjusted cosine similarity in this paper. The learning index, of which data comes from adjacent well, takes drilling cost, the well length and drilling time into account. The platform location optimization model, which considers learning effect, maximum number of wells one platform allowed and well trajectory, is established. The genetic algorithm is applied to solve the optimization model and the genetic operator is improved base on shale gas exploitation in mountain area. All the calculation procedure of genetic algorithm is performed in this work. The case study indicates that the optimization model can reduce the platform amount in a given area and increase the well amount one platform drills, namely, reduce the drilling cost by optimizing the platform location. The study demonstrates that the platform location optimization model established in this paper can both effectively quantify learning effect due to the well factory mode drilling in mountain area and decrease the drilling cost.