An Improved Algorithm for Calculating Flow Paths of Injection-Production in Single Sand Body in Old Oilfields Under the Background of Big Data

This paper proposes an accelerated algorithm for quickly and automatically identifying flow paths of injection-production in single sand body by computer. For old oilfields, which are characterized by large availability of data, long production history, and frequent stimulations (such as well pattern infilling, reperforating and zonal isolating), great efforts and much time will be spent in calculating injection-production relationship with respect to dates, sand bodies and well clusters since initial production of these oilfields. To solve these problems, the point robot shortest path algorithm based on traditional visibility graphs is improved. By analyzing the features of vertices at boundary of each single sand body in the free space, the number of boundary vertices for constructing visibility graphs in single sand body is reduced, and then the number of boundaries for constructing visibility graphs can be reduced. This new algorithm greatly improves in the speed for calculating the shortest flow path of injection-production. It has been proved fast and efficient in applications - 20 times faster than traditional algorithms. This algorithm has been programmed and applied for finely analyzing production performance of reservoirs in more than 80 old oilfields of PetroChina. It significantly enhances the efficiency and accuracy of analyzing the control effect of well patterns on single sand body.