Development of probabilistic models for forecasting the presence of reservoirs in horizontal wells

Background. The article is devoted to the problem of forecasting the distribution of reservoirs along the borehole of a horizontal section when planning the drilling of horizontal wells in the Perm Krai. The forecast of reservoir fraction along the length of the horizontal section taking into account only geological modeling does not in all cases allow to reliably determine the necessary value. Objective. To increase the accuracy of reservoir penetration prediction by means of building multilevel multivariate probabilistic statistical models. Materials and methods. The object of the study was geological and geophysical data of horizontal wells drilled in the northern part of the Volga-Ural oil and gas province. The task was solved by attracting additional geological and geophysical information that determines the presence of a reservoir in the horizontal wellbore. Results. A total of 301 horizontal wells drilled in the Perm Krai using LWD systems and geosteering technology were analyzed in detail. The main geological indicators that determine the presence of a reservoir in a horizontal wellbore were identified, such as: sandiness coefficient, dissection coefficient and effective formation thickness. These indicators are determined from the available data from the point of well entry into the horizontal section and are further used in the development of multivariate models. Conclusions. With the help of the constructed models, it is possible to determine the values of the probability of reservoir distribution and, consequently, to estimate correctly the planned penetration of the reservoir, which, in turn, for specific oil and gas areas, will provide a more solid basis for justification of drilling horizontal wells, and also will allow to determine the optimal sequence in case of drilling several horizontal wells.